Recent Posts

1

[Max] Bug Reporting / Re: Very slow geometry phase 10x slower than old pc

« Last post by Juraj on Today at 15:57:43 »

Yeah I can understand, but now that devs have latest high-end dual-Epyc, let's see what happens :- ). There are lot of after-market Epycs on market so that would be quite good for studios if their performance becomes bulletproof in Corona.

2

[Max] Bug Reporting / Re: Very slow geometry phase 10x slower than old pc

« Last post by webuilddreams on Today at 15:30:21 »

That is not what I meant, there is nothing faulty with CPU. But that doesn't mean all software can utilize it in same way. How Embree can utilize the hardware depends on many aspects of software stack, from motherboard firmware to OS scheduler. If those things are written for retail SKU, tiny discrepancy can cause misutilization.

Years ago, one of such solutions was upgrading Visual Studio redistributable used for compiling. Just saying it isn't as easy as updating libraries often, though I do hope it is for this particular issue.

Hi Juraj, thanks for your input, we or I am a bit frustrated , spend a small fortune and now this shit, hope the devs solves this soon.

3

[Max] I need help! / Re: Slow DR Parsing etc

« Last post by dj_buckley on Today at 15:25:18 »

Finally had to chance to upload a scene Ticket #234970.  A couple of things worth noting

It's already been part optimized so not as heavy as normal but it does still exhibit everything discussed in this thread to the point of failure on one of nodes despite it rendering fine locally etc etc

It's not as optimized as it could be, I'm well aware it could be optimized further but that's besides the point.  RAM usage and subsequent failure on the nodes due to commit ram size when rendering via DR is the main point.

If you have any questions about the scene.  Why it's setup a certain way etc etc just email me.  No screenshots here please as it's confidential.

Oh and apologies for the size of the upload it's roughly 10GB worth of assets but that's typical of day to day scenes and the stuff I work on.

4

[Max] Bug Reporting / Re: Very slow geometry phase 10x slower than old pc

« Last post by Juraj on Today at 14:54:11 »

That is not what I meant, there is nothing faulty with CPU. But that doesn't mean all software can utilize it in same way. How Embree can utilize the hardware depends on many aspects of software stack, from motherboard firmware to OS scheduler. If those things are written for retail SKU, tiny discrepancy can cause misutilization.

Years ago, one of such solutions was upgrading Visual Studio redistributable used for compiling. Just saying it isn't as easy as updating libraries often, though I do hope it is for this particular issue.

5

Hardware / Re: Threadripper & Ryzen only builds (3rd Gen starts on page 50)

« Last post by Juraj on Today at 14:46:58 »

Perhaps not much else to do with left-over retail stock since most systems are built as all-in-one solutions. There might also be quite less interest in them compared to sourcing any available GPU stock.

6

[Max] Bug Reporting / Re: Very slow geometry phase 10x slower than old pc

« Last post by webuilddreams on Today at 13:24:26 »

I hope solution for your problem is found, but I will like to stress that if in some case you have non-retail stepping (ES/QS/etc..), the issue can be with on-chip memory controller which creates certain spiral of firmware/OS issues that can manifest under certain workloads (whether it's scheduler issues, CPU groups, NUMA nodes, etc..).

So if Corona team received fully retail unit (which in 2024 means all-in-one system for most Epyc stuff), they might not be able to replicate your issue.

It's bit risky proposition to build custom multi-CPU servers today, too many variables with zero support. Super hard to investigate and find what can be culprit.

Not saying that's the case, just be prepared.

Nice to know that Corona devs got the machine though.

Hi we have a single retail Epyc 9754 (so no Numa nodes), we ruled out that the processor (and the whole system) is the culprit:

the slow geometry phase is caused by a particular step 'Embree scene calculation'.

The epyc is in the calculation of the 'Embree scene building' step very slow, 10x slower than for example a ryzen 5900x (12 cores)!

So to rule out that the Epyc cpu is slower when it comes to Embree calculation we want to test this slow Embree scene calculation and we tested it with Embree benchmarked software : phoronix-test-suite benchmark embree, the result was :

AMD Epyc 9754                              : 127 frames/sec
 
Compared to
 
AMD Ryzen Threadripper Pro 7995WX  : 116 frames/sec
 

Conclusion the Epyc is a bit faster in Embree calculations compared to a 7995WX (96 cores) so our system is not the issue, could it be the Embree version inside corona ?( Corona using old embree version?)Intel Embree is currently at version 4.3 do you know what version currently is used in Corona 11.1?

7

Gallery / Re: vis.juju renderings

« Last post by vis.juju on Today at 13:21:12 »

Outstanding renderings! I love the shaders more than the reference. Nice skills, very inspiring :))

Wow, thanks a lot! This was quite challenging

8

Hardware / Re: Threadripper & Ryzen only builds (3rd Gen starts on page 50)

« Last post by Nejc Kilar on Today at 13:14:00 »

...

I was waiting for the "I need moar sleep" comment when you broke then news :) We just got a dog recently and while it is clearly not the same as having a child boy did I not sleep for a week or two. Still recovering lol.

Yeah the CPUs I meant are the Epyc ones and lately we are really putting in a lot of focus on making sure the performance with these system is as it should be.

Still, I don't quite understand why the top of the line chips seem to be discounted by roughly 50% at most stores I'm looking at. Great time to buy though lol.

As for your 7980x, dang, that looks sweet! I'm currently contemplating getting one so all your feedback here is super duper useful. Thank you!

9

Gallery / Re: The Cat

« Last post by romullus on Today at 12:32:50 »

Unfortunately i can't help you, i don't use cryptomatte, but i think this should not happen. Maybe try to post a new topic in i need help board, or leave a message in this topic and someone more knowledgeable should help you to solve this issue.

10

[Max] Feature Requests / Re: Better implementation of corona maps in Nitrous viewport

« Last post by NazarVitkovskyi on Today at 12:28:59 »

@Jpjapers by saying "more complex material graphs in viewport", what exactly do you mean? What would you like to be added?