31
Hardware / Re: Threadripper Builds
« on: 2019-06-12, 18:20:15 »
Wasnt that new 16c with 80% of 2990wx performance seriously overclocked with like 1.45Vcore ? I dont think thats sometihng you would like to run 24/7.
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
That said, could be anything from heatsink seating to fan profiles, to software used to monitor temperature readings (super inaccurate) to what package is being run. Cinebench and Blender are less intense than Corona.
I'm happy to report that it's all going well so far after 1 day. The RAM is clocking in at 3000, and I'm seeing average temperatures of around 68 degrees under full load.
I agree with you there fully Nox3d. There is no reason to avoid genres, themes, looks, even cliches just because they were done before or popularized by someone else. There is always space for another Farnsworth House :- ).
First, it's how we all learn, we emulate what we find beautiful and speaks to us. As long as it's not plagiarization, it's new for person who made it. And just because it's common to you, it's not to someone else. Given there's a billions of us, reaching for originality above all is futile attempt, sometimes bordering on pretentious. You can all proclaim photography to be dead with approach like this, which is hardly the case.
I am also tired of backhanded "compliments" that speak more about commenter than the author of artwork, like the endless "another XYZ? (more boring archviz?uuh...). For those who don't find the artwork interesting, find another one to comment that's more to your liking. I don't go around the street telling people I don't like their commuting family cars, why don't they buy more unique sport cars ? Well that's lot of stupid analogies from me but you get the point. Ignore if you don't enjoy the genre or feel the need to nitpick irrelevant crap.
You could try manually inputting all the timings and clock for memory without having XMP set it for you (it sets all sort of other stuff as well, potentially overriding your overclock/power settings).
AVX code is still affecting AMD architecture in Corona, so there will never be absolute parity with something like Cinebench or Vray, I've asked devs multiple times about this and there is no way they can get rid of it due to their compiler.
3.3-3.4 +/- is the correct all-core speed roughly for 250W ceiling/default settings.
I had all sorts of odd shanenigans where the Spectre hot-fixes bundled with regular Windows updates (every 3 months +/-) would cap the 2990WX clock at 3.0 for 250W all-core due to affecting power settings. But outside of that, no issue with memory clock/timings. This sounds like XMP-profile/AVX-turbo issue.
How did someone get 37 secs bench time. I have a 2990wx or to 4.0 with 32 go ram and I had a score of 1:12. What am I doing wrong?
https://www.hardocp.com/news/2019/01/03/amd_ryzen_threadripper_2990wx_performance_regressions_linked_to_windows_bug
https://bitsum.com/portfolio/coreprio/
For slovak / czech speakers https://www.svethardware.cz/chaby-vykon-threadripperu-2990wx-byl-identifikovan-jako-dusledek-chyby-v-kernelu-windows/48337
Seems to slow down Corona benchmark by 2 seconds for me. Doesn't appear to have any impact on Vray benchmark. May still be a good idea to run, either way, to ensure that dynamic local mode is running as a permanent service. For 7-zip there is apparently some benefit.
Anyone else loses a second or two with corona bench?
I 'm using Noctua NH u14s with double mounted fan to cool my system.
I' m curious about other idle and "loaded" temps with solutions like wraithripper, enermax, or custom.
my Tdies are : idle 28-29 C and loaded 48-49 C . Room temp is 20 C.
loaded rendering 67 c