Chaos Corona Forum
Chaos Corona for 3ds Max => [Max] General Discussion => Topic started by: Ondra on 2016-03-08, 10:57:58
-
Here is related blog post: https://corona-renderer.com/blog/corona-1-3-benchmark/
(https://i.imgur.com/2EXGtlJ.jpg)
We have finally release an updated Corona Benchmark. It features automatic upload and verification of results, new scene, and Corona 1.3 core. Download the benchmark and view results here: https://corona-renderer.com/benchmark
In addition you can use this thread to share and discuss results, and report any problems.
edit: I am attaching the high-res version of the scene used here as a bonus
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Core(TM) i7-4960X CPU @ 3.60GHz(up to 4.4 Ghz)
Time: 0:03:02, Rays/sec: 2 670 970
-
It's alive!
-
Hah, works well :- )
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU E5-2680 v2 @ 2.80GHz (x2) {it's 3.1Ghz all-core turbo though)
Time: 0:01:17, Rays/sec: 6,315,400
-
Finally!
Wonder who will be the first to go under one minute :]
-
There's still E7-46xx something guy who should be able to score less than 30 seconds :- )
-
Finally!
Wonder who will be the first to go under one second :]
(https://forum.corona-renderer.com/index.php?action=dlattach;topic=11243.0;attach=43886;image)
-
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Unable to install VS++ 2015 redistributables.
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
What now? Anything I can do about it?
-
Does the benchmark encompass adaptivity?
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU E5-1620 v2 @ 3.70GHz
Time: 0:05:37, Rays/sec: 1,443,980
windows10
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz
Time: 0:04:24, Rays/sec: 1.840.070
CPU Clock is not correctly reported from Windows, the correct Clock is 4.7Ghz
-
CPU Clock is not correctly reported from Windows, the correct Clock is 4.7Ghz
Yeah, this should have been somehow solved. For clarity. Or at least offer self-reporting.
-
Hi!
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Genuine Intel(R) CPU @ 2.50GHz (x2)
Time: 0:01:13, Rays/sec: 6 697 330
CPU Intel Xeon E5-2697v2 (ES)
-
Does the benchmark encompass adaptivity?
It's a 1.3 benchmark, not 1.4.
-
Does it detect overclocked cpu frequencies properly? Or it's just that nobody submitted OC'd results yet?
-
It does not.
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU E5-2640 v2 @ 2.00GHz (x2)
Time: 0:02:11, Rays/sec: 3 763 080
-
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Unable to install VS++ 2015 redistributables.
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
Can confirm, Corona benchmark tries to install 64bit VS++ redist, but fails.
The official 64bit VS++ exits with the following in the log:
[0128:0A88][2016-03-08T14:23:13]e000: Error 0x80240017: Failed to execute MSU package.
[0B20:06D4][2016-03-08T14:23:13]e000: Error 0x80240017: Failed to configure per-machine MSU package.
[0B20:06D4][2016-03-08T14:23:13]i319: Applied execute package: Windows81_x64, result: 0x80240017, restart: None
[0B20:06D4][2016-03-08T14:23:13]e000: Error 0x80240017: Failed to execute MSU package.
Know bullshit with the 2015 installer, this is Computer specific, I'll post a workaround in this thread in couple of minutes.
It's a 1.3 benchmark, not 1.4.
I thought that was the Trello entry submited earlier then 1.4 release...
See attached image.
In addition you can use this thread to share and discuss results, and report any problems.
The executable lacks a Version number, looks aweful with 0.0.0.0 in the explorer :D
see attached image
-
Does it detect overclocked cpu frequencies properly? Or it's just that nobody submitted OC'd results yet?
sometimes ;). We did our best, but it seems to be really hard to detect in windows
-
it was originally planned for 1.3, but it got postponed. We decided to name it 1.3 benchmark, as it does not use the 1.4 features - adaptivity (which currently causes non-determinism), denoising (which is not fully debugged yet), and new VFB.
-
I got this error :
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Unable to install VS++ 2015 redistributables.
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
any help would... help ;-)
-
sorry, but this test is not quite correct. The results are surprising:
http://joxi.ru/MAj78KWtvEQ8lr
-
Does it detect overclocked cpu frequencies properly? Or it's just that nobody submitted OC'd results yet?
in most cases - no
-
There is a long pause at the end of the render.
I'm getting inconsistent result with my E5-2699v3 - anywhere from 0:52 to 1:50 The rendering itself is finished at around 0:45.
-
ouch :/
@keymaster: could you do a Legacy version? with HUGE warning sign that it is legacy?
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Your processor doesn't support required features (SSE 4.1).
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
-
There is a long pause at the end of the render.
I'm getting inconsistent result with my E5-2699v3 - anywhere from 0:52 to 1:50 The rendering itself is finished at around 0:45.
does this happen also with regular Corona rendering?
-
(http://i.imgur.com/4aWuW1R.png)
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Core(TM) i7-4790K CPU @ 4.00GHz
Time: 0:03:58, Rays/sec: 2 042 270
....
-
sorry, but this test is not quite correct. The results are surprising:
http://joxi.ru/MAj78KWtvEQ8lr
I can easily make up to 20% difference in render time on my machine depending on various factor, like how fresh OS install is, how many programs are running in the background, how buisy they are and so on. Nothing surprising here.
-
Yep, that i7 5930k is my benchmark, and I may have been watching some funny cat fail compilations on youtube while benchmark ran in the background :)
-
I've run it 2 times, the second time closing all other programs and I got this result.
(http://i.imgur.com/YBlcO28.jpg)
The first run was at 1:16. After it finished rendering it just kept going for about half a minute, but reported the results with the time it actually finished rendering.
-
In my testing of my home PC (i7-3930K) I've got five different result between 4;38 and 5;35. The "hang" after rendering differed from 4 sec. to over 20 sec.
-
!!! see my results: !!!
3ddd.ru/forum/thread/show/tiestiruiem_protsy_max_corona/ (http://3ddd.ru/forum/thread/show/tiestiruiem_protsy_max_corona/)
-
My PC is warning me about this site.
-
There is a long pause at the end of the render.
I'm getting inconsistent result with my E5-2699v3 - anywhere from 0:52 to 1:50 The rendering itself is finished at around 0:45.
does this happen also with regular Corona rendering?
No, I haven't noticed.
Also, looks like i'm not the only one:
In my testing of my home PC (i7-3930K) I've got five different result between 4;38 and 5;35. The "hang" after rendering differed from 4 sec. to over 20 sec.
-
we will tomorrow probably release updated version of the benchmark that reports consistent times in the popup window (the one sent to web seems OK), and runs even if MSVS fails to install
-
Would be also great to increase the resolution to get higher render times. All high-end xeons get about the same result with the current test.
-
I have feature request for benchmark score table - make it able to sort records by other columns, not only render time. Ability to sort by CPU would be most welcome as it would allow to estimate average performance of particular CPU.
I saw i7 4771 entry with 7:16 time, while mine does it in about 5 minutes. Without sorting list by CPU, it's impossible to hunt down all entries of certain PC configuration and thus it makes this benchmark less effective than it could be.
-
is it posible to see all row in one page like benchwell? they have 1490 row and thats easy to read and find any cpu.
-
we will tomorrow probably release updated version of the benchmark that reports consistent times in the popup window (the one sent to web seems OK), and runs even if MSVS fails to install
heh - thanks.
I have 3 systems, all three for some awesome reason post VS++ 2015 install errors. Win7 64bit, 2x 8.1 64bit
all updated windows.
I found the culprit on one system, but not the other 2.
All redists of c++ 2015 fail to install, culprit on one being a broken security update, that you have to force install through Deployment Image Servicing and Management. I can't get that update installed on the other two, every fix for this is PC specific and can have a multitude of reasons, ranging from too small reserved system partition (deleted it on one PC for programming reasons on my part) to random shit like Cisco VPN clients conflicting.
There are pages upon pages of this update's clusterfuck and it forbidding to install vs c++ 2015, whilst posting awesome and very descriptive "unspecified error" messages.
Basicly an exeption in the code, with no explanation or debug messages to help.
One system is freshly installed, 100% empty - no idea why there.
-
Added: https://corona-renderer.com/benchmark/?all
-
The 4-cpu system ranked #22 falls behind a 2-cpu system with the same cpu... wonder what went wrong here.
Looking forward to the updated benchmark, would really like to know how my machine compares.
-
Feature request: rays/s recorded in the table - it is proportional to time in this particular scene, but could serve as reference for other scenes (to see if there is some mistake in the scene or if cpu is performing as it should)
Bug report: splash screen is heavily JPG'd
-
The 4-cpu system ranked #22 falls behind a 2-cpu system with the same cpu... wonder what went wrong here.
Looking forward to the updated benchmark, would really like to know how my machine compares.
Those are E5-2xxx cpus, they don't work in 4-cpu configs. So probably some kind of bug. (ES cpus, too many cores, etc.)
-
please send optimal settings for 3930k P9X79 32gb2133mhz
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU E5-2650 v3 @ 2.30GHz (x2)
Time: 0:01:17, Rays/sec: 6,362,200
Still waiting for the xeon E7-E5 46xx guys to put us to shame :P
-
I wonder, why is a 2680v2 lower that the 2650v3. The 2650v3 doesnt even come close to the base clock of the 2680v2 with turbo boost on. When on full load I get close to 2.6ghz per core.
-
Might be perhaps because Corona is more memory bandwidth taxing that most multithreaded applications.
2680v2 is still stronger in all other benchmarks than 2650v3 (3.05Ghz all-core turbo vs 2.6Ghz all-core turbo cannot be nullified by Haswell vs Ivy Bridge improvement, which isn't so drastic).
On other hand, Haswell-E is on DDR4, most builds usually at 2166mhz, while most Ivy-Bridge-E are 1600Mhz (my board for some example won't even accept to run at such clock, so I am forced to run at 1333).
This can make a difference if Corona can use it well.
-
That looks to be the case. I am just curious about future upgrades and the difference between generation of cpus. There best benchmarks are on the software you will actually use and since corona is one in our case I was surprised to see that difference. (note: Juraj that was your test I saw). Very nice benchmark overall and easy to test across many machines. The alpha V4 gave me the higher results for a i7 3820 than the 2650v3's.
-
The V4 build basically ignored second CPU in dual-setup and even than it didn't scale well across many threads.
-
User-Submitted Results
#50 00:02:00 kklors Intel(R) Core(TM) i7-5960X CPU @ 3.00GHz 3 8 16 60 2016-03-08.
Decent performance from your i7-5960X, you must be OC'd to 4.5Ghz? You shaved :30 seconds off of a i7- 5960X OC'd to 4Ghz, tried 4.5 but too many bluescreens ;(
-
New version is up, try it!
- detect real CPU frequency better (win8+) and detects when it cannot detect it
- No more fails due to MSVC redistributables
- Displays consistent times in popup window and on web
- no freezes when rendering finishes (hopefully)
- Prints more info in the popup
Old unreliable frequencies were removed from results. I am considering resetting/archiving the scoreboard. We will continue to search for way how to detect CPU frequency on win7, but it is unbelievably hard
-
no freezes when rendering finishes (hopefully)
Ondra, thanks for trying to fix it. I hate to say it but it still freezes. But reports a correct time .
-
Ondra, thanks for trying to fix it. I hate to say it but it still freezes. But reports a correct time .
Hey,
damn... thanks for letting us know! Would it be possible to put together a build with some diagnostic information and give it to you to try? The problem has not occured to anyone here at the office.
Thanks,
Robin
-
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Application exited with errors.
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
-
no freezes when rendering finishes (hopefully)
Ondra, thanks for trying to fix it. I hate to say it but it still freezes. But reports a correct time .
the time reported in renderstamp is total time, including startup/cleanup. The popup and our website show only rendering time.
As for the freezing, could you do the "2. Reporting errors or freezes if Autodesk Error Report window is not available" here https://coronarenderer.freshdesk.com/support/solutions/articles/5000524006 and send us the minidump?
-
I added a 5k high-quality render of the scene in the first post as a bonus ;)
-
Ondra, thanks for trying to fix it. I hate to say it but it still freezes. But reports a correct time .
Hey,
damn... thanks for letting us know! Would it be possible to put together a build with some diagnostic information and give it to you to try? The problem has not occured to anyone here at the office.
Thanks,
Robin
Of course, please send me a build - I will try and report back
As for the freezing, could you do the "2. Reporting errors or freezes if Autodesk Error Report window is not available" here https://coronarenderer.freshdesk.com/support/solutions/articles/5000524006 and send us the minidump?
Not sure if that applies to the benchmark.
-
Not sure if that applies to the benchmark.
If Ondra says so... ;)
-
Got an Error after Rendering
see screenshot on further detail
-
Got an Error after Rendering
see screenshot on further detail
can you send us the minidump? https://coronarenderer.freshdesk.com/support/solutions/articles/5000524006
-
C++ 2015 32 and 64 bit succesfully installed.
I presume it's the windows security update, that's required to install the vs c++ 2015. I circumvented it and force installed it, worked great, but the programm still wouldn't launch.
On neither of 3 machines.
Far fetched, but German OEM licence preventing the update maybe? At this point I believe a higher power is trying to prevent me from running that piece of code.
-
The updated benchmark starts (and it seems to install the ms files), runs fine as far as I can tell but crashes upon finishing, the error reported is:
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Application exited with errors.Visual C++ 2015 Update 1 Redistributable (x64) Failed (ErrorCode =3010)
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
-
Seems to work now. My thinking is a windows updated that was pending. Now its all good.
No crashes
-
Hello guys
I`m using a dell workstation with double Xeon X5570 processor but when i render the final results comes as if I have only one processor !!!! can any one tells me whats wrong?
Thank you
-
Ran the benchmark twice on my desktop computer, each run took exactly the same time:
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Core(TM) i5-4460 CPU @ 3.20GHz
Real CPU Frequency [GHz]: Undetected
Render Time: 0:07:22, Rays/sec: 1,098,630
Tested on my laptop(i5-3230M @ 2.60GHz) as well, it took around 19 minutes to finish it.
-
Yesterday I decided to check how the benchmark performs on my old laptop for fun, and I am getting this super informative error while trying to submit the result. This probably isn't very important, but I will report it on mantis.
Btw, render time on this super old i3@2,4GHz is ~20min, while it's 5min on an i7-4770. That's pretty surprising. Only 4x difference. :)
-
i understand that phenom X6 not new CPU but it would be interesting if test will run on it - for now it give message:
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Your processor doesn't support required features (SSE 4.1).
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
ОК
---------------------------
-
i understand that phenom X6 not new CPU but it would be interesting if test will run on it - for now it give message:
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Your processor doesn't support required features (SSE 4.1).
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
ОК
---------------------------
The older benchmark supports it. You can check that and compare to those plattforms posted in the thread.
https://forum.corona-renderer.com/index.php/topic,559.0.html
Why a non SSE4.1 version was included is a mystery to me aswell.
SSE 4.1 was introduced only a generation later, with the bulldozer architechture, although ironically enough Phenom II to this day are faster with FPU performance, due IPS being higher, but core count lower.
You get a ~20% performance hit and do not support the embree kernel fully. Otherwise there is no difference.
edit: As far as I recall, one of my older machines, phenom II x6 1045T 2,7ghz was 1.2-1.3 mio rays/s on the old benchmark.
BTW, I know AVX didn't bring any really new improvments over SSE4.1 and using embree with AVX would screw over compatibility, but was corona ever tried out with AVX, the way embree was originally intended to be run? Was there a performance imporvement?
-
corona has SSE2 or SSE4.1 switch that you need to do during installation. Embree switches between SSE2, SSE4.2, AVX, and AVX2 in run time based on CPU type. It always picks the best option for you, regardless whether corona has legacy or regular build.
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Genuine Intel(R) CPU @ 2.50GHz (x2)
Real CPU Frequency [GHz]: 2.8
Render Time: 0:01:10, Rays/sec: 6 893 130
Intel Xeon E5-2697v2 (ES)
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU E5-2675 v3 @ 1.80GHz
Real CPU Frequency [GHz]: Undetected
Render Time: 0:01:54, Rays/sec: 4,239,430
-
Nice benchmark :)
Just wanted to say, if anybody cares, that the page count under the table is not "responsive" like the rest :) and it goes all the way to the right (screenshot)
-
we will fix it soon ;)
-
Why this processor reports different number of physical cores in benchmark? Intel(R) Xeon(R) CPU E5-2696 v3 @ 2.30GHz (×2)
-
Why this processor reports different number of physical cores in benchmark? Intel(R) Xeon(R) CPU E5-2696 v3 @ 2.30GHz (×2)
It is possible they disabled some cores so they have only one processor group (and legacy applications work with all cores)
-
fixed the pagination
-
sad :(
My quad Opteron (4x12/48 K10 cores@3 GHz) gives this message :(
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Your processor doesn't support required features (SSE 4.1).
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU E5-2660 0 @ 2.20GHz (x2)
Real CPU Frequency [GHz]: Undetected
Render Time: 0:01:48, Rays/sec: 4 489 470
-
hmm does that count?
Windows10 ... and dual xeon 8cores - so 16 physical in total.
-
"Debuggen" :D
I think you should be able to get a minidump using the method described here:
"2. Reporting errors or freezes if Autodesk Error Report window is not available"
https://coronarenderer.freshdesk.com/support/solutions/articles/5000524006
And then send it over to Mantis: https://corona-renderer.com/bugs/
-
Win10 Pro
i7 970 + 24gb ram
(http://dl1.joxi.net/drive/2016/04/10/0001/1614/120398/98/1b25dfab57.jpg)
(http://dl2.joxi.net/drive/2016/04/10/0001/1614/120398/98/854026fb0c.jpg)
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Core(TM) i7-5960X CPU @ 3.00GHz
Real CPU Frequency [GHz]: 3.7
Render Time: 0:02:14, Rays/sec: 3,625,700
-
Win10 Pro
i7 970 + 24gb ram
(http://dl1.joxi.net/drive/2016/04/10/0001/1614/120398/98/1b25dfab57.jpg)
Considering the fact, that you made it to 15 passes it implies a hardware failure. Faulted returns when over clocking usually make this behaviour.
-
it could be also from something just after the render, such as sending the results
-
Win10 Pro
i7 970 + 24gb ram
(http://dl1.joxi.net/drive/2016/04/10/0001/1614/120398/98/1b25dfab57.jpg)
(http://dl2.joxi.net/drive/2016/04/10/0001/1614/120398/98/854026fb0c.jpg)
Right click on the file -> Run as Administrator.
-
win10 pro
-
unexpected shutdown
-
I had the same experience like iNevestenko
OS X Crash-Report
http://pastebin.com/2qrteSJX
CPU: 2 x 2,66 GHz 6-Core Intel Xeon X5650
-
What CPU do you have?
-
Crashes after completing benchmark.Tried 2 times.
Macbook Pro
i5 2.4 Ghz
Crash report
http://pastebin.com/a0HpUc9X
-
Hi there not working for me ... same as iNevestenko and vladoviz
OSX Capitan v 10.11.4
Model: MacPro5,1, BootROM MP51.007F.B03, 12 processors, 6-Core Intel Xeon, 3.46 GHz, 96 GB
Graphics: NVIDIA GeForce GTX 980 Ti, NVIDIA GeForce GTX 980 Ti, PCIe, 6144 MB
-
No problems here. The benchmark ran OK. Results:
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Core(TM) i7-6700K CPU @ 4.00GHz
Real CPU Frequency [GHz]: Undetected
Render Time: 0:03:57, Rays/sec: 2,048,850
When I compare my rendertime with Windows machines that have the same processor and less RAM. My mac is sometimes a minute slower! I was running the benchmark with no other applications open on a freshly started iMac.
Regards,
-
Hi,
I am the developer of Corona Benchmark OS X port. First, I would like to thank you for testing the Benchmark, it is very helpful.
To find the cause of problems I created special testing build https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc). Please test it and let me know if there will be some error message or crash.
Thank you
PS:
This testing build is slower than normal one.
-
Again "Corona1.3.Benchmark.Testing" crashes right away.
Corona1.3.Benchmark.Testing Crash-Report
http://pastebin.com/6nTBTGB9
CPU: 2 x 2,66 GHz 6-Core Intel Xeon X5650
-
Please test it and let me know if there will be some error message or crash.
Still crashing as soon as benchmark finished.
Macbook Pro
i5 2.4 ghz
OSX 10.9.5
http://pastebin.com/caVtrr6W
-
Still crashing on opening for me
Mac pro 6,1 (late 2013)
3.5 GHz 6-Core Intel Xeon E5
16 GB 1866 MHz DDR3 ECC
Osx 10.10.5
-
Hi there,
Corona1.3.Benchmark.Testing.dmg still crashes over here...
Best regards
OSX Capitan v 10.11.4
-
The test build works on my machine. Strange.. I tried it at home with a slightly older iMac and it crashes there.
-
The test build works on my machine. Strange.. I tried it at home with a slightly older iMac and it crashes there.
We did another update, try it now...
-
But try normal build (https://corona-renderer.com/stuff/Corona1.3.Benchmark.dmg (https://corona-renderer.com/stuff/Corona1.3.Benchmark.dmg)) instead of testing now.
-
The latest version has a small improvement. Now the crash happens after the Splash-Screen shows up.
(http://i.imgur.com/Fm5uyzB.png)
-
Now its working fine.
OSX 10.9.5
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Core(TM) i5-4258U CPU @ 2.40GHz
Real CPU Frequency [GHz]: Undetected
Render Time: 0:13:13, Rays/sec: 612,560
-
Hi again,
same as vladoVIZ ... can't run it it crashes
-
ok, we will continue tweaking it until it is rock solid. Expect a new version soon
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Core(TM) i7-5820K CPU @ 3.30GHz
Render Time: 0:03:01, Rays/sec: 2,675,920
-
Corona1.3.Benchmark.dmg
-
It is working now!
Many thanks
-
How long it takes before the application becomes unresponsive or is the application unresponsive immediately after splash screen appears?
-
How long it takes before the application becomes unresponsive or is the application unresponsive immediately after splash screen appears?
On MacOSX
The splash screen appears right away, that it takes about 8-10 seconds until application becomes unresponsive
or (Corona 1.3 Benchmark not responding) showing up in the Activity Monitor.
-
Thank you for the quick reply. I probably know where is the problem but to be sure will the Benchmark start if you wait longer (few minutes ignoring the "not responding message")?
-
Thank you for the quick reply. I probably know where is the problem but to be sure will the Benchmark start if you wait longer (few minutes ignoring the "not responding message")?
If keep the Benchmark open after about 15-20 seconds the App quits and I get the "Corona1.3 Benchmark quit unexpectedly" message.
-
I download a Corona 1.3 Benchmark today. It run without any alert. When it finished, it crash and popped up a alert.
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Application exited with errors.Visual C++ 2015 Update 1 Redistributable (x64) Failed (ErrorCode =1638)
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
-
Thank you for the quick reply. I probably know where is the problem but to be sure will the Benchmark start if you wait longer (few minutes ignoring the "not responding message")?
If keep the Benchmark open after about 15-20 seconds the App quits and I get the "Corona1.3 Benchmark quit unexpectedly" message.
Can you please click on "Report..." button on "Corona1.3 Benchmark quit unexpectedly" message dialog, save the crash log to file and attach it to you reply? Thank you.
-
Thank you for the quick reply. I probably know where is the problem but to be sure will the Benchmark start if you wait longer (few minutes ignoring the "not responding message")?
If keep the Benchmark open after about 15-20 seconds the App quits and I get the "Corona1.3 Benchmark quit unexpectedly" message.
Can you please click on "Report..." button on "Corona1.3 Benchmark quit unexpectedly" message dialog, save the crash log to file and attach it to you reply? Thank you.
Here is the "Corona1.3 Benchmark quit unexpectedly" - Report
http://pastebin.com/5Tsiv08i
-
According to crash report it looks there is a bug in one of external libraries. I created another testing build with newer version of library (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc)). Please test it and let me know. Thank you.
-
According to crash report it looks there is a bug in one of external libraries. I created another testing build with newer version of library (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc)). Please test it and let me know. Thank you.
When I start The Benchmark, I get this error message
(http://i.imgur.com/WTBi8Uq.png)
Crash-Report
http://pastebin.com/yx49ysZt
-
Hello
Which configuration has your Mac?
-
Hello
Which configuration has your Mac?
Thanks rafpug - but I´m not sure that you can be helpful in this matter.
Else you would know that the information about the "Mac configuration" is in the Crash-Report.
-
When I start The Benchmark, I get this error message ...
Crash-Report
http://pastebin.com/yx49ysZt
Thank you for your patience Vlado. I uploaded new testing build with different configuration. Can you please test it (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc))?
-
Thank you for your patience Vlado. I uploaded new testing build with different configuration. Can you please test it (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc))?
It still crashes. The only difference is that the splash-screen doesn´t show at all.
I also run it on my macbook pro - the same result.
-
OSX benchmark working well here on my old iMac El Capitan 10.11.6 beta.
Not anything to write home about - 4.56, but interestingly I ran the benchmark on the same machine using Bootcamp (Windows 10) and got a slower speed of 5.37.
Adam
-
OSX benchmark working well here on my old iMac El Capitan 10.11.6 beta.
Not anything to write home about - 4.56, but interestingly I ran the benchmark on the same machine using Bootcamp (Windows 10) and got a slower speed of 5.37.
Adam
:- D
Now I am really curious if this is "fault" of the OSX Benchmark or what.
-
OSX benchmark working well here on my old iMac El Capitan 10.11.6 beta.
Not anything to write home about - 4.56, but interestingly I ran the benchmark on the same machine using Bootcamp (Windows 10) and got a slower speed of 5.37.
Adam
interesting, what is your exact CPU model?
-
It still crashes. The only difference is that the splash-screen doesn´t show at all.
I also run it on my macbook pro - the same result.
Mac Book Pro with the same result? Interesting. Would it be possible to connect to your computer remotely e.g. with TeamViewer that I can immediately test my testing builds to find the cause of the bugs?
Thank you
-
It's a mid 2011 iMac and the processor is (from the specs at Apple - I'm not at home to check right now) a 3.4GHz quad-core Intel Core i7.
It says Intel(R) Core(TM) i7-2600 CPU @ 3.40GHz on the benchmark details.
-
It still crashes. The only difference is that the splash-screen doesn´t show at all.
I also run it on my macbook pro - the same result.
Mac Book Pro with the same result? Interesting. Would it be possible to connect to your computer remotely e.g. with TeamViewer that I can immediately test my testing builds to find the cause of the bugs?
Thank you
Sorry, but I don´t feel comfortable to give to anyone a remote-access to my computer. You have to finde a different solution.
If your read the postings here you see the pattern that all Mac users that seem to be able to run the benchmark have a consumer iMac with the Intel i7. On the MacPro with the Intel Xeon-Processor the benchmark is crashing.
-
Sorry, but I don´t feel comfortable to give to anyone a remote-access to my computer. You have to finde a different solution.
If your read the postings here you see the pattern that all Mac users that seem to be able to run the benchmark have a consumer iMac with the Intel i7. On the MacPro with the Intel Xeon-Processor the benchmark is crashing.
I understand and I would like to thank you again. You have already helped us a lot. The problem is with some Xeon processors as you write. If there's anyone who has a problem with crashing benchmark and would not mind me connecting to his computer remotely please let me know.
PS
There is still testing build (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc) to download and test. Even though this build will crash it is very helpful to send me the Crash Report that I can see where exactly the build crashes so I can create another testing build with different configuration. After few iterations we can find the configuration which causes the crash as we found it for BatRM 6-Core Intel Xeon E5 processor.
-
Sorry, but I don´t feel comfortable to give to anyone a remote-access to my computer. You have to finde a different solution.
If your read the postings here you see the pattern that all Mac users that seem to be able to run the benchmark have a consumer iMac with the Intel i7. On the MacPro with the Intel Xeon-Processor the benchmark is crashing.
I understand and I would like to thank you again. You have already helped us a lot. The problem is with some Xeon processors as you write. If there's anyone who has a problem with crashing benchmark and would not mind me connecting to his computer remotely please let me know.
PS
There is still testing build (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc) to download and test. Even though this build will crash it is very helpful to send me the Crash Report that I can see where exactly the build crashes so I can create another testing build with different configuration. After few iterations we can find the configuration which causes the crash as we found it for BatRM 6-Core Intel Xeon E5 processor.
Here is the Crash-Report from this test-build
http://pastebin.com/cSZtZLey
-
I just uploaded new testing build (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc). The crashing should be fixed now.
-
I just uploaded new testing build (https://drive.google.com/open?id=0B7xyYS3daSJ7Z1d6UWpkZXVrMHc). The crashing should be fixed now.
Yes, I can confirm. No crash - the scene is rendering.
-
Uploaded fixed version of OS X Corona Benchmark to https://corona-renderer.com/benchmark/ (https://corona-renderer.com/benchmark/).
Thanks all for testing especially vladoVIZ.
-
Now it is working thank you for the effort.
Anxiously waiting for the osx releases
osx 10.11.4
2 x 3.46 GHz 6-Core Intel Xeon
96 GB 1333 MHz DDR3 ECC
2xNVIDIA GeForce GTX 980 Ti 6144 MB
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU X5690 @ 3.47GHz (x2)
Real CPU Frequency [GHz]: Undetected
Render Time: 0:01:57, Rays/sec: 4,122,290
-
Hello guys!!
I just got this error after few minutes running benchmark.
I already tried to re-instal Visual C++, but the error remains.
Thx!
-
Hello guys!!
I just got this error after few minutes running benchmark.
I already tried to re-instal Visual C++, but the error remains.
Thx!
Hello,
you must install vc_redist.x64.exe from https://www.microsoft.com/en-US/download/details.aspx?id=48145 (https://www.microsoft.com/en-US/download/details.aspx?id=48145). If it doesn't help try to restart computer after installation of redistributables.
-
Congrats to the person with 2679v4 OEM :- ) Nice score.
-
Hello guys!!
I just got this error after few minutes running benchmark.
I already tried to re-instal Visual C++, but the error remains.
Thx!
Hi, try this solution. It may sound as stupid as "reboot your PC", but has actually helped some users:
https://coronarenderer.freshdesk.com/support/solutions/articles/12000011862-material-editor-is-very-slow-freezing-in-corona-1-4
-
Uploaded fixed version of OS X Corona Benchmark to https://corona-renderer.com/benchmark/ (https://corona-renderer.com/benchmark/).
Thanks all for testing especially vladoVIZ.
nikolat thanks for your appreciation. Glad I could help and I´m looking forward for corona on OSX.
-
My results on OSX:
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU X5680 @ 3.33GHz (x2)
Real CPU Frequency [GHz]: Undetected
Render Time: 0:02:04, Rays/sec: 3,889,750
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU E5-2630 v4 @ 2.20GHz (x2)
Real CPU Frequency [GHz]: Undetected
Render Time: 0:01:18, Rays/sec: 6.178.260
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU E5-2690 v3 @ 2.60GHz (x2)
Real CPU Frequency [GHz]: Undetected
Render Time: 0:00:59, Rays/sec: 8 224 920
-
Damn, you Coroners Xeoners. 1 min benchmark, us i7-5950X don't get there....whaz i7-6950X doing ?
got it. looking like 1:30
-
looking like 1:30
Don't mistake overclocked results crowding the top places.
The 5960x (there is no 5950x) gets ~2:20m on full turbo @ stock freq.
The 6950x gets ~ 1:50m on full turbo @ stock freq.
Both have the same frequency, but the 6950x has two more cores.
-
I'm running the test and at the end it's crashing and gives me thie error: ---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Application exited with errors.
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
-
Did you try running it as admin?
-
I get this error after installing Wacom Drivers. I have tried re-installing VC Redistributable. Below is the error
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Application exited with errors.Visual C++ 2015 Update 1 Redistributable (x64) Failed (ErrorCode =1638)
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
any solution?
-
I would love a new benchmark based on Corona 1.5 or Corona 1.6 (once released).
-
we are not planning new benchmark that soon, there is really no point since the rendering core does not change that much. You will have to wait few versions more ;)
-
Ondra, I am planning to write CPU compare utility based on data from CoronaBenchmark.
Is there any way to get access to the database?
-
Ondra, I am planning to write CPU compare utility based on data from CoronaBenchmark.
Is there any way to get access to the database?
yes, goto https://corona-renderer.com/benchmark/?all , copy table to excel, done.
You can do all Excel stuff, average across CPUs, CPU vendors and all the statistics Jazz.
The 6811 entries are probably all there is.
-
That isn't really the solution.
First - data not up to date. But that could be solved by dynamic accessing html data.
Second - I have a feeling that original database is a little bit better sorted. For example, the model of the processor and number of processors should be separated.
-
We show the CPu name exactly as it is sent to us - just one string ;) We store some additional data not shown on the page, such as cache sizes, but they are probably not interesting to users (which is why they are not shown on the page). We could make you an API to access the data easier (json/xml), but if you dont mind I would rather first see your utility ;). We had too many people promising something and not delivering, I dont want to commit extra work to this beforehand. Hope you understand.
-
I understand perfectly.
So I'll try to make some mockup application based on data from the website.
My main purpose is to find easy way to compare true performance of processors. I don't find synthetic benchmarks satisfying.
My main PC uses are Max + Corona/Vray, and I just want to make good purchase decisions.
Sometimes you just get few percent more performance, for a lot more money. That's why I came up with the idea (currently building next workstation for my work).
-
My main purpose is to find easy way to compare true performance of processors. I don't find synthetic benchmarks satisfying.
Sounds super exciting, good luck!
-
Someone with Ryzen configuration? Which configuration do you use?
-
Someone with Ryzen configuration? Which configuration do you use?
There are 5 Pages of Benchmark results and sepcs (https://corona-renderer.com/benchmark/?cpu-type=Ryzen&submit=Search) by now ;]
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Core(TM) i7-3960X CPU @ 3.30GHz
Real CPU Frequency [GHz]: Undetected
Render Time: 0:02:56, Rays/sec: 2.752.430
PC Specs: http://goo.gl/WqiqBp
+Info: https://corona-renderer.com/benchmark/?cpu-type=i7-3960X&submit=Search
-
Did anyone solve this?
J.
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Application exited with errors.Visual C++ 2015 Update 1 Redistributable (x64) Failed (ErrorCode =1638)
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
-
Same error here !
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Application exited with errors.Visual C++ 2015 Update 1 Redistributable (x64) Failed (ErrorCode =1638)
(Press Ctrl + C to copy the message to clipboard.)
---------------------------
OK
---------------------------
Corona 1.7 im using.
Any idea how to solve?
-
Have you tried running Windows Update, installing or re-installing MS VC++?
A guide about this can be found here, check out point 1 - https://forum.corona-renderer.com/index.php?topic=11721.0
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU E3-1230 v3 @ 3.30GHz
Real CPU Frequency [GHz]: 3.3
Render Time: 0:05:43, Rays/sec: 1 413 210
-
Have you tried running Windows Update, installing or re-installing MS VC++?
A guide about this can be found here, check out point 1 - https://forum.corona-renderer.com/index.php?topic=11721.0
Running the benchmark as Admin has solved this for me
-
I would like to see more information about computer configurations to understand what needs to be improved in my computer.
For example, with my i5-3470 processor, there are 29 records, but I'm in 25th place 00:10:05.24, against 00:06:57.21 in the first place
-
I would like to see more information about computer configurations to understand what needs to be improved in my computer.
For example, with my i5-3470 processor, there are 29 records, but I'm in 25th place 00:10:05.24, against 00:06:57.21 in the first place
CPU is the most important factor, and specific model should get pretty consistent results. The faster ones are probably overclocked.
-
Therefore, you need to display the actual frequency of the processor, the system bus, and perhaps the model of the motherboard. A lot of talk on the network about what kind of configuration should be a computer for 3DMax. This table will help at least partially solve this problem.
I also want to see the diagram on this table (possibly with grouping by processors). 30 entries on the page are few to analyze the results.
For starters, it would be enough to download the data for self-analysis.
-
Hi guys, I recently bought a new PC and I'm in a bit of a struggle here. When I build it I benchmarked it in Cinebench and I got a score of 1977pts which was way too low. I then went in BIOS and activated the X.M.P. and then the score went up to 2650pts. After that I overclocked it @4.0 GHz and the score was around 2950pts. So, today I decided to benckmark it in Corona 1.3 benchmark and I'm posting a screenshot with my score. I know this has something to do with my computer and I'm asking you guys for any possible solution to this problem. Thank you!
AMD Threadripper 1950x OC@4.0Ghz
Gigabyte x399 Aorus Gaming 7
Enermax LIQTECH TR4 360 (water cooled)
Corsair Vengeance 2x16GB @3000MHz
Gigabyte Aorus GeForce GTX 1080 Ti
Samsung 960 PRO M2
PSU: Corsair HX1000i (1000W)
-
My new PC ;)
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Core(TM) i9-7980XE CPU @ 2.60GHz
Real CPU Frequency [GHz]: 2.6
Render Time: 0:00:57, Rays/sec: 8 404 450
-
time for a new benchmark scene soon?
this one is less than 2 years old and already not really useable as some machines are down to 14 seconds. How about for the next one the average machine takes more like 30 mins to complete? I reckon there are still enough of us nerds willing to wait that long so that you will still get a decent amount of data.
-
Hi guys, I recently bought a new PC and I'm in a bit of a struggle here. When I build it I benchmarked it in Cinebench and I got a score of 1977pts which was way too low. I then went in BIOS and activated the X.M.P. and then the score went up to 2650pts. After that I overclocked it @4.0 GHz and the score was around 2950pts. So, today I decided to benckmark it in Corona 1.3 benchmark and I'm posting a screenshot with my score. I know this has something to do with my computer and I'm asking you guys for any possible solution to this problem. Thank you!
AMD Threadripper 1950x OC@4.0Ghz
Gigabyte x399 Aorus Gaming 7
Enermax LIQTECH TR4 360 (water cooled)
Corsair Vengeance 2x16GB @3000MHz
Gigabyte Aorus GeForce GTX 1080 Ti
Samsung 960 PRO M2
PSU: Corsair HX1000i (1000W)
Using cpu-z or your bios check whether your ram is running in dual or single mode. You want dual (quad is even better but you only have 2 sticks.). If you're in single mode your sticks might be in the wrong spots. Seems like you are RAM bottlenecked.
-
time for a new benchmark scene soon?
this one is less than 2 years old and already not really useable as some machines are down to 14 seconds. How about for the next one the average machine takes more like 30 mins to complete? I reckon there are still enough of us nerds willing to wait that long so that you will still get a decent amount of data.
We are not planning to do new benchmark soon, we are currently understaffed as there is ridiculous stortage of programmers in czech republic, and we need to focus on more important projects. The benchmark is just a "nice to have" side project, compared to displacement, tonemapping, etc. improvements.
-
I've never managed to get the Benchmark scene to work properly. Well it renders but the second it is finished it crashes, so I'm unable to see the time or submit my score. This happened on both machines I try it on. I also tried starting it in Admin mode, but still crashes when finished. Giving me this error.
---------------------------
Corona 1.3 Benchmark Error
---------------------------
Could not run Corona 1.3 Benchmark: Application exited with errors.Visual C++ 2015 Update 1 Redistributable (x64) Failed (ErrorCode =1638)
Anyone has any suggestions?
-
Anyone has any suggestions?
What is your CPU and Windows version?
Please try installing our official version first: https://corona-renderer.com/download/
And then run the benchmark. The point is that the installer installs some additional system components which are required by Corona. It is possible that you are missing those.
If there are any problems with the installation of the Windows components, try this: https://coronarenderer.freshdesk.com/support/solutions/articles/12000019570
-
Anyone has any suggestions?
What is your CPU and Windows version?
Please try installing our official version first: https://corona-renderer.com/download/
And then run the benchmark. The point is that the installer installs some additional system components which are required by Corona. It is possible that you are missing those.
If there are any problems with the installation of the Windows components, try this: https://coronarenderer.freshdesk.com/support/solutions/articles/12000019570
I all ready have the corona renderer installed, without any errors and working perfectly in 3ds Max. On both this machine and the previous one I had Corona installed while testing the benchmark.
Both machines are running the latest windows 10 Creators update.
First machine I used and tried it on had Broadwell E 6850K.
Current/new which I tried has AMD Threadripper 1950x .
-
Is there a way to start the benchmark via command line?
I would like to add Corona 1.3 benchmark to our render test suite and it would be great to have this feature - start from command line and a result dump to a file or to stdout. Thanks!
-
Anyone has any suggestions?
If there are any problems with the installation of the Windows components, try this: https://coronarenderer.freshdesk.com/support/solutions/articles/12000019570
THis is the message I get every time the Corona Benchmark crashes, dunno if that makes it any more obvious what the problem is? I tried on 4 machines now, its the same crash every time.
https://www.dropbox.com/s/j34wnlzqqb5ggte/Screenshot_1.jpg
-
THis is the message I get every time the Corona Benchmark crashes, dunno if that makes it any more obvious what the problem is? I tried on 4 machines now, its the same crash every time.
https://www.dropbox.com/s/j34wnlzqqb5ggte/Screenshot_1.jpg
404 error
Can you re-send the error?
If you can see some error message, you should be able to generate a minidump file (see point 2): https://coronarenderer.freshdesk.com/support/solutions/articles/5000524006
-
Sorry for the slow reply I tend to forget turning on the notification.
I think you misunderstood, its not the plguin in max, Im talking about the benchmark.exe file, the standalone benchmark test. Every time I run it Im getting an error at completion. So no way of uploading results or anything.
See screenshots:
http://prntscr.com/j6ekg6
error message:
http://prntscr.com/j6ejlc
-
That would be an error in the Visual C++ Redistributable, and I'd guess those Windows components need installing or repairing. A google for the error gave me this link
https://support.microsoft.com/en-us/help/4092997/vc-2015-redistributable-install-error-1638-when-newer-version-present
(https://support.microsoft.com/en-us/help/4092997/vc-2015-redistributable-install-error-1638-when-newer-version-present)
Which did seem to have advice on how to repair the redistributable.
-
Hmm, yeah actually I dont have that installed on my system. I have a fresh install of windows 10. same with the other machines Ive tested on. ANd I have the c++ 2017 Redistributable installed. So I assume Corona benchmark requires the 2015 version then to work?
And apparently I cannot install the 2015 because I all ready have a newer one.
-
Is there a way to start the benchmark via command line?
I would like to add Corona 1.3 benchmark to our render test suite and it would be great to have this feature - start from command line and a result dump to a file or to stdout. Thanks!
Can I have a reply, please? Is it possible or not? Thank you!
-
Is there a way to start the benchmark via command line?
I would like to add Corona 1.3 benchmark to our render test suite and it would be great to have this feature - start from command line and a result dump to a file or to stdout. Thanks!
Can I have a reply, please? Is it possible or not? Thank you!
Sorry for the lack of reply.
I think it should be possible to start the benchmark app using a simple batch script, but not sure about saving the output/info. I'll try to find out asap.
-
@Toorop:
You can use this automatic benchmark app: https://www.dropbox.com/s/50g6f06zloptqzg/corona%20automatic%20benchmark.zip?dl=0
Usage instructions:
-in the _run.bat file you can change the number of loops (times the benchmark will be run), you can set it to 1 to have just one run
-run the _run.bat file
-wait for the benchmark to finish
-after that, the result will be saved to CoronaBenchmarkResult.txt in the following form:
Corona Benchmark Results
2018-04-23 12:59:41
MARCIN-I7-6700
Microsoft Windows 10 Pro (10.0.16299)
Intel(R) Core(TM) i7-6700 CPU @ 3.40GHz
32704MB of physical memory
Iteration 1
Render Time [ms]: 0:05:27
Rays/Second: 1,483,750
Average
Render Time [ms]: 0:05:27
Rays/Second: 1,483,750
Median
Render Time [ms]: 0:05:27
Rays/Second: 1,483,750
-
@maru:
Thanks a bunch, perfect!
-
there is no x5470 xeon with 8 cores/8 thread....... wrong cpu and of course the skor!!
please corrected or removed this post
thanks alot
(http://i214.photobucket.com/albums/cc242/ultraex2003/dfgdhffhgfgh.jpg)
-
Hi
i get this score
2m 30s on a amd ryzen 7 2700 + 32 gb ram
But there is something i don't understand.
Why there are pc with same processor but less ram (like 16) than can get faster result?
-
If the benchmark render fits into your memory, there is no difference in 16 or 32 or more RAM. More RAM doesn't result in faster render times.
The clock speed of your RAM however will have an impact on the result. (Would be nice to see the clock speed in the table as well)
Another thing is the speed of the CPU itself. Some CPU's run at stock speeds, some are overclocked.
-
Every now and then, when new processors get released i consult the benchmark results page to see if updating our rendernodes with these new CPUs would be a good choice.
Now with the Benchmark fixed on corona core 1.3 I just wonder if the performance diffenrence between two systems still accurately transfers to the newest release of corona. And if the answer is no: will there be a corona benchmark using a newer render core in the future?
-
If we released a new version of benchmark, there would certainly be some performance difference (I hope for the better), but the relative performance of 2 difrerent CPUs should be roughly still the same. We didn't do any specific CPU optimization for a long time, so anything which speeds up Corona on CPU A should (almost) equally speed it up on CPU B.
For this reason we don't have any immediate plans for releasing a new version of benchmark, unless we do some optimization targeting specific CPUs (or specific instruction sets) or we port Corona to a new platform.
-
Hi! I'm doing some test and I'd like to know if this render times are alligned with my specs.
I own a laptop Dell Inspiron 7559 (Intel Core i7 6700HQ, Nvidia GeForce GTX 960M, 16GB/1600Mhz, SSD 850EVO) and from benchmark tool I got 00:06:48.45, which is one of the worst results (first result was 00:04:51.49)
I've also converted an old scene done with V-Ray 2.x, where I used at the end almost only materials from the Corona Library and for one image of ~2K my laptop took 4 hours to clean up to 5%(+denoiser) at ~700K rays/s without "low on ram" alert.
Maybe is a conflictive scene because of GI and DOF, but I had read that normal values will be within 1,5M and 4M.
Thanks for reading
-
Laptops will be generally slow, and your result does seem to be within the range available for that processor. Mine uses the same processor (reports a clock speed of 3.13 on the Freq column of the benchmark which is not that different from yours; does report a higher rays/s at 1,532,620 though) and came in at 5m 17s. There are some at the 7 or 8 minute mark for that processor too.
-
So the same CPU in desktop workstation can perform up to 2X better?
They are the first times with a (slightly) biased solution, so to reduce render times I've to necessarily buy a dedicated workstation.
Thanks
-
I'm not sure you can get the same CPU in a desktop version - laptop CPUs are generally made specifically to be laptop CPUs (lower power consumption and other things to optimize them for that particular use). This means that in general laptops are not good rendering machines, compared to what you will get for the same price as a desktop.
-
I've used corona benchmark a lot recently to test some workstations with OC that would be used as rendernode, I've noticed a big difference in watt consumption and temperature of cores between the benchmark and the actual render with v2/3, to the point that I dropped the benchmark completely. I assume it's because the latter uses the new AVX instrunction and a more recent version of embree, among other changes. Are there any plans to release an updated version of the benchmark? I believe it would be helpful also to show the difference between the newer CPUs.
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) Platinum 8172M CPU @ 2.60GHz (x2)
Real CPU Frequency [GHz]: 3
Render Time: 0:00:22, Rays/sec: 21 272 300
-
Rendering took 67.064 seconds
CoronaCore::exiting renderFrame
Unique Primitives: 3253360
Primitives with instancing: 471064508
Avg samples per pixel: 16
Avg rays per sample: 51.4926
Rays/s: 7.24491e+06
Samples/s: 140719
xeon 2696 v4
64gbram
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
AMD Ryzen Threadripper 2950X 16-Core Processor (x2)
Real CPU Frequency [GHz]: 3.8
Render Time: 0:01:09, Rays/sec: 7,020,770
-
Hello there guys!
Right now I work with 5 pcs for render, using it in DR! I was wondering if it will ever be a Benchmark so see how well the network rendering perform.
Now I'm using an old Dual XEON + 3 i7 4790K + 1 i7 3820 and I'd love to know if it was worth upgrading all the machines or just sell everthing and by the best one on the market (in terms of Brazilian market of course... over here a XEON is worth at least two livers!)
Thanks in advance!
-
I've used corona benchmark a lot recently to test some workstations with OC that would be used as rendernode, I've noticed a big difference in watt consumption and temperature of cores between the benchmark and the actual render with v2/3, to the point that I dropped the benchmark completely. I assume it's because the latter uses the new AVX instrunction and a more recent version of embree, among other changes. Are there any plans to release an updated version of the benchmark? I believe it would be helpful also to show the difference between the newer CPUs.
Hi, I also think update to Corona Benchmark would be good. In our case overclocked i9 machines tested with Benchmark 1.3 were completely fine (while running Corona 2.0 on daily basis), but they all started to overheat once we updated renderers to Corona 3.0. (temperatures over 105C). I believe new benchmark could prevent such incidents, adjusting overclocking settings would be easier knowing Corona's behaviour in latest updates.
-
Hello all,
I remember I saw this around but can't find the right answer, so sorry for asking again. Why in the benchmark results many Ryzen Threadripper are x4? As long as I know you can't have double processors, but maybe I'm not well informed.
Anyone knows about this and can explain? Thanks
-
Anyone knows about this and can explain?
If you mean in general, multi socket boards have existed since forever.
In case of threadripper you are correct, there are no multi socket boards for thread ripper, only for the epyc line up.
The Corona benchmark accidentally counts the multiple dies as multiple CPUs. Whilst you could make the argument, that the way thread ripper is set up it's not too far away from that idea, it's a bug.
-
yes, the benchmark was released before ryzens/threadrippers, so we didnt know the detection will give incorrect results.
-
Just one question.
I have a Ryzen 7 2700.
Searching on the benchamark results this kind of cpu gives 1:40 min to rendering, in my pc instead is 2.40.
Why?
-
We just finished a new node, a dual 7742 ES and get very funny result of 36 sec. Cinebench R20 scales correct to 30k. Any insights on this?
Thanks,
H
-
We just finished a new node, a dual 7742 ES and get very funny result of 36 sec. Cinebench R20 scales correct to 30k. Any insights on this?
Thanks,
H
What's the Ray/s on production scene though?
Also, can you send me an email about this if you feel particularly Christmass-ly :- )
-
12k, email sent!
-
I'm also getting poor performance in corona too, cinebench r20 is around 30k so that is correct. Are there any limitations to core number or something in corona?
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
AMD Ryzen Threadripper 3970X 32-Core Processor
Real CPU Frequency [GHz]: 3.88
Render Time: 0:00:29, Rays/sec: 16,312,300
-
My new amazing Ryzen PC ;)
Corona 1.3 Benchmark
Ryzen 3950x CPU @ 4.200GHz all core
Render Time: 0:00:53, Rays/sec: 9 101 310
Best cpu for your money that sure !!!
-
Can not to run Benchmark...what can i do???
-
Can not to run Benchmark...what can i do???
Do you have internet connection?
-
Can not to run Benchmark...what can i do???
Do you have internet connection?
Of course......
-
Can not to run Benchmark...what can i do???
Do you have internet connection?
Can you send me a testing license?
liulei.n@gmail.com
Thank U!
-
Already solved.Thank U.
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
AMD Ryzen Threadripper 2970WX 24-Core Processor (x4)
Real CPU Frequency [GHz]: 3.7
Render Time: 0:00:52, Rays/sec: 9 219 870
-
40 seconds on the military vehicle...
Will continue to improve...
E5-2699 v3
-
Hi Guys, I assembled the PС with AMD Ryzen 9 3950X + Asus ROG B 500E, the render results are below.
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
AMD Ryzen 9 3950X 16-Core Processor
Real CPU Frequency [GHz]: 4
Render Time: 0:00:59, Rays / sec: 8 175 210
59 seconds is 17 seconds longer than the best result, what does it depend on and how can you bring the render speed to the ideal?
-
Nice benchmark, but it's too old.
Dev team, could you please make a new one with interior scene + caustics, it would be great to test the brand new features and compare the results.
If it will be issued after the release of Corona 7 (with all its feature), it would be awesome.
Thanks.
-
Nice benchmark, but it's too old.
Dev team, could you please make a new one with interior scene + caustics, it would be great to test the brand new features and compare the results.
If it will be issued after the release of Corona 7 (with all its feature), it would be awesome.
Thanks.
I still has same correlation. If something is faster here by 30% it will be faster similar amount in any scene and any setttings.
-
What we'd ideally really love is a test which is approx. 3-4x more intensive as the old one, as current high-end machines are shredding it in just 14 seconds or so. Plus it should have several intensive elements to denoise, which is really useful to check temps, as we know they climb significantly during denoising. Would it be an issue to start again with a CoronaBenchmark-v2 with a fresh set of results? Also one which doesn't pop up with the vdist thing too, ideally. And even more ideal would be the ability to have it run itself x-times in a row and take an average result.
-
Hi Alex,
We made some "advanced" command-line version of the benchmark which supports multi-run & easier averaging. Please find the link attached below to download & test it for yourself.
https://www.dropbox.com/s/50g6f06zloptqzg/corona%20automatic%20benchmark.zip?dl=0
I hope this helps.
Thanks.
Regards,
Avi
-
Hey Avi, that looks interesting. Is it possible to set a custom number of passes? And maybe it has denoising too? Or any plans to implement/update?
Cheers,
-
Hi,
In reply to your questions:
1. No, it is not possible to set a custom number of passes.
2. Currently, It does not have denoising implemented because this benchmark is from Corona 1.3, and denoising was implemented in 1.4
3. We want to implement a new benchmark at some point but at present, we don't have this planned but in case this changes, we will notify you.
Regards,
Avi
-
Thanks for the update, Avi.
-
Hello everyone!
Please tell me what can be done in such a situation?
Всем привет!
Подскажите пожалуйста, что можно сделать в такой ситуации?
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
AMD Ryzen 7 2700 Eight-Core Processor
Real CPU Frequency [GHz]: 3.1
Render Time: 0:03:16, Rays/sec: 2 467 250
-
At a guess, the CPU is thermal throttling due to insufficient cooling.
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel(R) Xeon(R) CPU E5-2660 0 @ 2.20GHz (x2)
Real CPU Frequency [GHz]: 2.7
Render Time: 0:01:55, Rays/sec: 4.209.790
-
Lower than expected results on 12th gen 12900K. Probably incompatible with Windows 11 or Hybrid architecture. Can't share the result due to NDA.
-
Lower than expected results on 12th gen 12900K. Probably incompatible with Windows 11 or Hybrid architecture. Can't share the result due to NDA.
I don't think is a Win 11 problem.
From other bechmarks online, is obvious that in multithreading the 12900K is not one of the best CPUs out there. An AMD Ryzen 5950X is still faster and in single thread is just a bit lower that the 12900K at the end. Excuse me for my comment but I don't see anything revolutionary from this CPU now...
-
Well, it's kinda hard to judge anything from that comment without knowing what the expected and the actual measured results are :)
-
Good evening everyone,
I post my first benchmark with Corona Render.
I am a CGI enthusiast who has not been in the business for a few years.
I am getting closer again and I have seen the potential of this CPU software.
My current workstation:
Xeon E52697v2
48Gb Ram
Nvidia Gtx 970
it is several years old but it is still in perfect working order and I haven't used it for at least 5 years.
I would love to hear from you taking into account that I currently use 3ds Max 2015 and Vray 3.6 and a "decent" 1920 * 1080 image with about two and a half hours I can get it.
Probably the best test would be to buy a month of software and try.
Thanks for any advice.
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
Intel (R) Xeon (R) CPU E5-2697 v2 @ 2.70GHz
Real CPU Frequency [GHz]: 2.6
Render Time: 0:02:20, Rays / sec: 3.447.620
Greetings,
Marco
-
Probably the best test would be to buy a month of software and try.
You don't need to buy Corona for this. We offer a 45-day demo license. All you need to do is just to download the installer from web, install Corona and then when you try to use Corona in 3ds Max, you'll be prompted to activate your license where you'll be given the choice to activate the demo.
-
Thank's for your reply
-
Hello everyone!
Cant run Corona renderer Benchmark with new clean OS win 7 x 64. Benchmark requests license every time. Does it need Corona's license for testing PC?
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
AMD Ryzen Threadripper 2990WX 32-Core Processor (x4)
Real CPU Frequency [GHz]: 3.3
Render Time: 0:00:43, Rays/sec: 11,285,800
....I think I have the best time around here.. I'm impressed.. and I was thinking of upgrading this PC to a Threadripper PRO 3975WX....
-
Dual EPYC 7T83 with 256 GB of old 2666 MHz RAM - 9 sec
-
I know this has all been requested years ago but I'd really love to see an updated benchmark that has the following:
- ability to run multiple tests automatically and take an average (and best) result
- something that's at least taking 5 or 10 x longer than currently on an average machine, to allow proper time for any throttling to take real effect etc.
- something that uses multiple render elements with denoising (which stresses the CPU a lot more and helps judge temps)
- outputs the result to desktop automatically
-
- something that's at least taking 5 or 10 x longer than currently on an average machine, to allow proper time for any throttling to take real effect etc.
No need to do this longer by x-times. It will be much better to run it for 1 or 2 minutes like in vray benchmark. This way it will be good for everyone.
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
11th Gen Intel(R) Core(TM) i7-11700F @ 2.50GHz
Real CPU Frequency [GHz]: 4.4
Render Time: 0:01:32, Rays/sec: 5 243 820
-
I know this has all been requested years ago but I'd really love to see an updated benchmark that has the following:
- ability to run multiple tests automatically and take an average (and best) result
- something that's at least taking 5 or 10 x longer than currently on an average machine, to allow proper time for any throttling to take real effect etc.
- something that uses multiple render elements with denoising (which stresses the CPU a lot more and helps judge temps)
- outputs the result to desktop automatically
We are working on new benchmark to be released during v10 cycle, that will have everything mentioned here, probably except for the denoising (unless we add it as a separate benchmark, but it is still tricky, as denoising is a finite workload. We want the benchmark to run for a fixed amount of time this time)
-
I know this has all been requested years ago but I'd really love to see an updated benchmark that has the following:
- ability to run multiple tests automatically and take an average (and best) result
- something that's at least taking 5 or 10 x longer than currently on an average machine, to allow proper time for any throttling to take real effect etc.
- something that uses multiple render elements with denoising (which stresses the CPU a lot more and helps judge temps)
- outputs the result to desktop automatically
We are working on new benchmark to be released during v10 cycle, that will have everything mentioned here, probably except for the denoising (unless we add it as a separate benchmark, but it is still tricky, as denoising is a finite workload. We want the benchmark to run for a fixed amount of time this time)
Awesome great to hear. I get the denoising issue. If you can please have it as a separate benchmark option that would be great - you could just prevent that one from submitting results to the website as they would skew the results. It's really more for internal testing/speed comparison/temps testing. Cheers!
-
IMO
for a longer, more intricate evaluation of a complex scene, instead of a still, an animation should do way better
in this regard, having options on start (to denoise and/or animate - select frames...) would be a must
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
AMD Ryzen 9 5900X 12-Core Processor
Real CPU Frequency [GHz]: 4.3
Render Time: 0:01:08, Rays/sec: 7,090,030
-
Hi everyone,
Does anyone know why my pc (pc_Tiziano) has a calculation time of 1:28 sec and makes 5 million beams against that of my colleague (pc_Marco with identical hardware) takes 46 sec and makes 10 million beams? When my pc performs the calculation, the cpu goes to 100%, what does it depend on?
This is the configuration of the 2 pcs:
Operating System
Windows 10 Pro 64-bit
CPU
AMD Ryzen Threadripper 2990WX 55 °C
Colfax 12nm Technology
RAM
128GB Dual-Channel Unknown @ 1064MHz (15-15-15-36)
Motherboard
ASUSTeK COMPUTER INC. PRIME X399-A (SP3r2) 59 °C
Graphics
4083MB NVIDIA GeForce RTX 3090 Ti (NVIDIA) 39 °C
Storage
1863GB Western Digital WDC WD20EFRX-68EUZN0 (SATA ) 27 °C
931GB Western Digital WDC WD10EFRX-68PJCN0 (SATA ) 27 °C
931GB Western Digital WD Blue SN570 1TB (Unknown (SSD))
-
Main suspects would be things like one is overheating and so throttling (cooler could be incorrectly attached), something else is running on the machine and limiting the availability of the CPU, something is badly seated like RAM and this is causing intermittent problems, or some Windows setting is different e.g. https://www.reddit.com/r/Amd/comments/9764si/windows_is_having_issues_with_2990wx/
-
Thanks so much Tom! I check everything!
Tiziano
-
You're welcome, if nothing shows up under those possibilities, let us know! Or let us know if something does show up too, I am interested to find out why this might be happening.
-
Hi Tom!
The computer technician solved the problem by updating and optimizing the motherboard bios and updating all drivers including windows, in fact it had never been done in 5 years of computers! :-)
Thanks for the support
-
Thanks for coming back to report the cause and solution! It's always useful to hear, and may help others in a similar situation. Glad all is resolved for you now!
-
Dual EPYC 7T83 with 256 GB of old 2666 MHz RAM - 9 sec
Just wondering how much Ram you have in your system and how is the RAM fitted in the system, 4 channel or 8 channel?
We just got a Dual EPYC 7773x and we're averaging 20 seconds on the benchmark which seems super slow.
-
Just wondering what do these numbers mean at the end of the CPU column?
-
Normally it means the number of physical CPUs. But it also means that sometimes our benchmark does not detect that number correctly. :)
We are working on a new benchmark where this (among other things) will be improved.
-
To be honest, even the latest Cinebench R23 gets some numbers and windows version all wrong. Might just be hard to get accurate reports from unusual platforms.
-
Modeling workstation:
i9 13900 ks, msi z690 ace, 128Go DDR5 5600MHz (4000MHz because xmp does not work with 4 sticks... another story...) , rtx 4090 suprim X, windows 11
It renders until the end but crashes when it populates results...
-> 44s, 11 185 291 rays
Notebook workstation (MSI RAIDER GE77HX)
i9-12900HX, 64Go DDR5 4800MHz, RTX 3080TI, win 11 (also crash at the end)
-> 1.13m, 6 308 106 rays
Rendering station (5 years +)
Dual xeon E5 2696 V4, 128 Go DDR3 2133MHz, GTX 1080 Rog Strix, Z10 PE-DW WS, windows 10
-> 48s, 10 710 948 rays
The result in cinebench R23 tells a different story... +38.000 points for I9 13900ks and +28.000 points for the xeons (quite huge as difference...)
It makes me questioning. Does bucket rendering takes more advantage of the processor power? Does corona can take full advantage of 12 or 13th gen proc architecture?
If we use blind maths the I9 should have been rendered in about 32s.
Just trying to understand... Any guess?
-
I've just run cinebench r23 on the notebook and the results seems to be +/- on the good ratio with 19.000 points.
So I just wondering why this ratio is so different with a 13th gen proc.
-
I would not pay too much heed to any result from current benchmark. Cinebench R23 is far newer, and Corona will eventually prepare new benchmark :- )
What I do to compare pure "multi-threaded" speed (but I rarely do that nowadays since I value smoothness of workflow more than brute speed) is using production scene, and only paying attention to Ray/s.
Don't worry about memory speed for both 13th gen Intel and Zen3+ platforms for work. Neither can support more than 4600 MT/s +/- when memory reaches total of 8-ranks. Which for DDR5 memory usually means 4 DIMMS. It doesn't affect almost any workstation task at all, I've tested this with dual-channel 64GB 5600/6400 on my HX12800 & HX 13800 laptops vs 128GB at 4000 on my Dell&MSI laptops. Since with lower MT/s you can also lower the timings, the latency is identical, you only lose bandwidth but that doesn't affect anything with current-gen hardware when it comes to most SPEC/Workstation tasks.
Funny, I had HX 12800 for 6 months and HX 13800 for 2 months and I didn't even bother to run Corona benchmark at all :- ).
-
AMD Ryzen Threadripper 3970x 32 core
64 Gb RAM
Render Time : 39 sec
12,365 M Rays/sec
-
AMD Ryzen Threadripper 3970x 32 core
64 Gb RAM
Render Time : 39 sec
12,365 M Rays/sec
Howdy! That seems a tad slow for a 3970x imho. Have you checked your temperature and power plans? Might be some extra performance you can get there. Typically a 3970x sits around the 30-32s mark.
-
The processor is in default settings without tuning Temperatures below 80 degrees. CPU frequency 3,9 - 4 Ghz.
Is it still worth investigating the cause of the slowness?
-
That is a good question but I guess it totally depends on whether you feel like it is worth investing time into it and whether that speed up is worth it to you when all is said and done. Setting up hardware to work as it should can sometimes take ages but if the speed up is worth it to you then I reckon it is worth it :)
-
Hi! I can see some people got 30 sec on this benchmark but i'm having 50sec.
Can anyone tell me what's the problem.
I'm using i9 13900k win 11 corona 7.
-
Hi! I can see some people got 30 sec on this benchmark but i'm having 50sec.
Can anyone tell me what's the problem.
I'm using i9 13900k win 11 corona 7.
Hi Perodj,
There may be a number of reasons behind the different rendering times among PCs with the same specs. Some of them might have already been mentioned in the thread.
- Some PCs might be overclocked.
I use a Xeon 2697 V2 on an Asus P9 X79 Pro motherboard. This CPU completes the benchmark in 2.17 in its stock settings. I have overclocked the BClk value to 114, which currently renders at around 1.51. The highest 1.49 rendering time in the mentioned processor's list is at the highest value that I could achieve - I don't want to push that high anyways. 1.51 is also quite a gain. I see some PCs on that list showing frequencies other than stock values which means they are overclocked however in my tests, processor speeds are reported as its non turbo value -2.7gHz, so don't rely on the reported frequencies.
Thermal Throttling (TT) might be causing a reduction in speed.
I was using an I7 3930K with this exact same motherboard when I first bought it. The stock fan provided by Intel was not adequately cooling this processor. So funny. I realized the problem when I have tested one of my scenes on another PC -which should be slower than mine, but the results were the opposite. Just then I've installed a software called HWMonitor and saw the temperature readings above the TT threshold for this particular CPU when rendering. (85 or 95 something) This software also reports the frequency of the processor, and reports all values for three states: Maximum frequency reached, minimum frequency and current frequency. My max frequency was 3400mHz but throttled down to 2600mHz while rendering -even less while rendering. I bought a tower fan and teperatures do not exceed 50C (celcius) even while rendering and my frequency was 3400 in full performance.
Power Configurations affect speed drastically. You might be using Power Saving algorithm.
This is the Windows Power Schemes that I mention. Balanced is ok in terms of performance, as far as I've tested, but if you have enabled Power Saving Preset then expect a seriously slow machine. I recommend you to switch to Performance and see what happens. Download and install HWMonitor, which is freeware. By that, you'll be able to tell whether the PC is operating slower than it should.
There might be other things to consider. These are some of the probable culprits I could come up with with a quick thinking. There are also BIOS settings that can slow you down. Just turn off Turbo Boost and your computer becomes a 10 year old model equivalent. There are also preset speed settings in some mobo bioses. I remember my PC being slow when Quiet Preset was chosen. There was also an Energy Saving Preset and Hi Performance one. Try it with a preset with fires, thunderbolts in its logo :D
-
Hi! Thanks for your answer. Currently i'm having this result. What does this means ? https://prnt.sc/1YDEAC54amNW
-
Hi All,
Ryzen 7950X3D
32GB DDR5 6000MHz CL30
Render Time: 0:00:36, Rays/sec: 13 443 100
Where I can find a database with submitted results?
Cheers!
-
Where I can find a database with submitted results?
Cheers!
On the Benchmark page, swap to User Results (small banner up top) to get to https://corona-renderer.com/benchmark/results
-
HI Everyone,
I'm currently commissioning a new machine into our studio and am finding the performance is lacking in corona 9 and on the benchmark.
I've also run the machine on Cinebench and Vray Benchmark and they are giving results that I would expect from this CPU. We also have a 3975WX 32-Cores machine which I have run a comparison against in the table attached.
Any help would be greatly appreciated as this is a bit weird that it is performing on vray and cinebench but not corona.
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
AMD Ryzen Threadripper PRO 5995WX 64-Cores
Real CPU Frequency [GHz]: 3.9
Render Time: 0:00:36, Rays/sec: 13,413,500
Build details-
CPU AMD Ryzen Threadripper Pro 5995WX CPU 64 Core / 128 Thread -
Max Boost 4.5Ghz
CPU Cooler Phanteks Glacier One 240T30 DRGB 240mm Liquid Cooler
Motherboard ASUS Pro WS WRX80E SAGE SE Wi-Fi AMD WRX80 Workstation Motherboard
Graphics Card N/A
RAM Supermicro Samsung 128GB DDR4-3200 2Rx4 LP ECC Registered
DIMM (2X 64GB)
SSD WD BLACK SN850X 1TB SSD M.2 NVME R/W 7300/6300MBS 5YRS
WTY
Case Corsair 7000D Airflow Black ATX MidTower Gaming Case Tempered
Glass
Power Supply Corsair HX Series HX1000 1000W Power Supply 80 Plus
Platinum - High Performance - Full Modular
OS Windows 11 Professional OEM
Thanks in advance for an help.
-
Do you have only two physical DIMMs populated? If so, that means you are running dual-channel, which could influence performance as the platform is 8-channel and it's suggested to at least run quad-channel.
Those 64GB Registered ECC LP (Low-Power = 1.2V) are also quite slow, they are 3200 MT/s but have very high timings (CL22 for DDR4 is not great at all) which slower the system memory latency and that can have strong influence on rendering speed.
It's best to avoid bigger than 32GB DIMMs, as those densities are reserved for server platforms where capacity is more important than latency.
For accurate benchmarking, it's best to set Windows mode to "Ultimate performance" in Power Settings, this will shorten the time the system will boost to higher frequency.
-
Thanks Juraj,
I’ll swap out the ram tomorrow for a test.
What would you suggest for ram to support a cpu such as this?
-
Short answer: In your case, I would probably stay with ECC to be safe and just use 256GB DDR4-3200 REG ECC (8x32GB) from either Samsung or Micron making.
Long answer blabbering:
That is controversial topic on "prosumer" platforms like Threadripper Pro :- ). "Go ECC or not go ECC" (There are also two types of ECC, unbuffered, and buffered/registered).
Threadripper Pro and Asus WX80 Sage to my knowledge supports all types of memory, unlike Xeon-W for example. And I currently don't own any Threadripper Pro, just Threadripper and Xeons, so I can't talk from experience, but few Corona team guys can, like Nejc.
1) ECC, particularly Registered ECC, is important for memory integrity, esp. at higher capacities (like 2TB). But for offline rendering, you don't need either integrity nor such capability. Contrary to the belief of ECC evangelists, your PC will almost never blue-screen or otherwise negatively affect your workload performance.
2) Motherboards like Asus WX do allow over-clocking of ECC memory. There is certain happy middle-ground between full JEDEC standard, and consumer XMP Profile. Something like 3200 CL18 would be still very conservative settings (compared to consumer 3600/CL16-CL18) but still much faster than JEDEC 3200/CL22+.
3) I would avoid high-capacity (64+) DIMMs, they have lot of ranks, often come in low-power stock mode, and are meant for different usage than workstations. 8x32GB is the ideal configuration for Threadripper Pro if you only need 256GB Ram (that is more than rendering needs). Plenty of capacity, all memory channels used (8), and 32GB DIMMs are easy to drive to 1.35V which enables faster settings.
-
Thanks Juraj,
That has sorted the bottleneck problem. Results below.
I'll look into your suggestions for ram to replace the two big 64gb pieces.
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
AMD Ryzen Threadripper PRO 5995WX 64-Cores
Real CPU Frequency [GHz]: 3.2
Render Time: 0:00:17, Rays/sec: 27,524,400
-
Corona 1.3 Benchmark Finished
BTR Scene 16 passes
AMD Ryzen 9 7950X 16-Core Processor
Real CPU Frequency [GHz]: 5
Render Time: 0:00:35, Rays/sec: 13,849,600
Top-down air cooler and 175W PPT. About as tuned as it gets without giving it more juice.
-
I think the Threadripper 5995WX (64 cores) and Ryzen 7950X (16cores) showcase interesting situation today. No matter the budget, you can get absolutely kickass performance, long gone are days when you had to have Threadripper or Xeon. The Ryzen gives you half the performance for fraction of cost.. Mainstream class of CPUs are simply very good.
-
Hello how are you I am new to this, I have a question I have an Intel Core i9 13900KF, 64gb ram, RTX 4090, cooling is H170I ELITE CAPELLIX, when I render a scene my CPU temperature reaches 92° I did the Corona benchmark test and The temperature only reached 72°, why did that happen? My scenes are poorly optimized and that's why it gives me 92° temperature? Thank you very much the truth is that I am very worried
-
The old Corona Benchmark is not intended as a temperature/stress test for a machine, and the rendering code has changed a lot since 1.3 - for example, Denoising was introduced in Corona 1.4, and Denoising is known for taking full advantage of what a CPU can do (and thus, because the CPU is working at its maximum for an extended period, will see higher temperatures). To be honest, poor optimization would likely cause lower temperatures as the renderer will not be able to perform as efficiently, if anything :)
So long as 92 is within the thermal tolerances for your CPU, then there is no problem - and if that is outside the tolerances, it would be wise to check the cooling (is it mounted properly, is the thermal paste properly applied, is it recommended as sufficient cooling for that CPU, etc.)
-
Un-stickied and posted a new thread here: https://forum.corona-renderer.com/index.php?topic=40702.0