Iâm a little bit confused with your benchmark results.
I see that you have âall GPUsâ for your benchmark results but the way I understand things now, 5.3+, does not support multi-GPUs on standard licenses or are these results from the so-called TVAi Pro version? Thank you for your time.
What are your thermals under that air-cooled at full load?
Also, Iâm sure you may know, though, the issue you could be experience is oxidation.
Did you upgrade your BIOS to microcode 0x129 yet?
Just a heads-up, if you are having issues, Intel has deployed a massive RMA for defected 13- 14-gen CPUs and Intel has extended users warranties an extra 2 or 3-years (I forget which) to DIY or pre-build systemâs owners.
This is with the new 0x129 microcode BIOS. My Noctua D15S cooler has never been able to handle the 13900K at full tilt and the new BIOS didnât change that. However, Asus has improved memory compatibility compared to my ~year-old BIOS and Iâm now able to run the Asus âtweakedâ setting for this kit. (This is an extra one-click option, with some tuned sub-timings, that Asus offers for some memory kits) This is a 2-stick G-Skill 48GB kit running at 7200MT 36-46-46-115
After trying a few different settings, I ended up using the Intel âExtremeâ Default settings, but with a 1.4V VID limit to keep temps under control. Iâm also running a 90C temp limit, but the benchmark doesnât get anywhere near it.
Yes, thatâs true, as I have some unexpectedly poor results.
I want to do cross-platform testing and follow Topazâs recommendations with GPU generations.
Iâm also testing all parts separately to determine any bottlenecks, but for now the 9950X is performing very well outside of TVAI, eating 210W, peaking 6 GHz single core, 5.4GHz on all cores, but now have to dissipate all that heat. The AM5âs IHS is so tiny⊠what a stupid choice
It is significantly ahead my previous 7950X, but also heats much more.
Iâll be back (soon?) with something like 2x RTX 4090 or 5x RTX 3070, depending on the time I spend coding
This is X3D, that is known to butkick because of its embeded cache.
I do want to massively o/c it so I canât go to a X3D else I will fry the cache instanly.
Also device is positionned to 0 which means âautoâ, I do also have better result using âautoâ than âall GPUâ, that is not normal : The load on GPU should be full and CPU only sending jobs to them. On âAutoâ itâs working much !
EDIT : Sorry I missread, please put your bench under quotes, itâs unreadable and forget all those things out of 1920x1080, we canât compare else
This is indedd very low with a dual RTX 4090, absolutely not normal.
Give mobo, RAM details please
EDIT 2 : Changed âproâ seating to workstation, Ada slightly o/ced
I think CEP should be enabled as well as TVB clipping. The 25% fan speed limit on the CPU cooler sounds pretty low, but I guess that depends on how quiet you want things. The 307 current limit probably wonât effect the benchmark, but could hurt performance when you run multiple renders at once. (25% fan speed will hurt in that scenario as well)
Other than that, the biggest limiting factor I see is your RAM. 6000 C32 vs 7200 C36 (and Asus âtweakedâ timings)
Also, are the âN/Aâ results tests that did not complete?
If we are both running the same benchmark resolution, we should either both get âN/Aâ or both get results. I cannot think of a reason why a 13900K / 4080S rig would get results but a 13900K / 4090 would get N/A. (Other than system setup)
I think the TVB frequency clipping is a safety thing that lowers frequency (and thus voltage) at higher temperatures to avoid degradation. Itâs called something different on my Asus board, but the result is lower boost clocks and lower voltage as temperatures increase.
The higher memory speed does kick the IMC voltage up to 1.4V but Intel seems to think the upper voltage limit for the cores needs to be 1.55V or less so Iâm okay with 1.4V on the IMC. I know AMD mentioned their X3D cache layer is more sensitive than the CCD silicon, but the 32mb of on-die cache didnât require any special voltage considerations.
I donât know that 1.4V is safe long term, but Intel didnât mention IMC voltage in their investigation of CPU failures and didnât change it in the 0x129 microcode thatâsupposed to save these processors from self distruction, so Iâm going with it.
I have no choice, my expensive license flows very quickly
i just tried for the first time Rhea on a bullshit video 720x400 to 4K, I am completely stuned bu the result, on some tests pictures. Going to encode full now.
I say on picture because, several videos are VERY resilient with a disappointment impossible to patch before coding : some with clear spot punishing the stream at random location, ruining the stream
very hard to reproduice, will post it in buggs section
X670E Aorus Extreme has arrived ! Week-end of tests, itâs so heavy, omg, a panzer
Also received my Nvme Gen4 x4 to PCIE Gen 4 x16, will see how much bandwidth it disaggregates. If few I order tons of it 10$, very worth it to get PCIE x16
Thereâs also a fu**** confirmed bug on X670E Crosshair and Ryzen 99xx, thermal managment workaround and CPU current delivery. The CPU bumps instantly to a skyrocket 100% current load and FANs scream to death trying to cool this ⊠500$ , incredible
Impossible to reproduce, it is random, when Asus chips decides NCT6796D
I discovered the âtweakedâ RAM timings were not fully stable, and when I tried using generic XMP I or XMP II the performance loss was bad enough that I decided to try and find a way to get some of it back. After some experimentation, it seems that the timings were more important than the raw MT/s.
By selecting the âtweakedâ XMP profile and then manually reducing the speed from 7200 down to 7000, I was able to get performance on-par with the 7200 speed. My IMC VDD is down to 1.350 too.
I discovered greta instability over all releases since 5.0.4. Latest release just did something completely newe : when adding a 4th GPU it becomes completely nuts
One of GPU crash and become invisible, and before that instead of using âAll GPUâ it uses only 2 and worst of them âŠ
I tried susing same, unsynced gen, exact same (same board w/ same firmare), multiple platform, always the same problem : everything turned wrong with multi GPU after 5.0.4 and I think we are not mcuh to see that.
