Suchergebnisse

Completed the quadfecta of stresstesting: Karhu + Testmem5 1usmus + RamtestPro + Y-cruncher 8-) More specifically the tune is CL32 GDM enable 8700MT/s and 2200mhz FCLK * 100.4 BCLK for an effective ~8734MT/s + 2209mhz FCLK I'm not using any spesial tweaks to boost karhu numbers such as...

Sorry i cannot fulfill this request :cry:

Maxed out 9950X3D

Updated GPU scores Extreme: Memory ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ Godlike: Memory 1760050603 Friend did...

Pretty much nothing except the chance for a better binning since its a EPYC

I paid 702 euros + tax for it, on the 12.06.2025 from Proshop Seems it has gotten alille cheaper since then https://www.proshop.no/ProsessorCPU/AMD-EPYC-4585PX-43-GHz-processor-OEM-ProsessorCPU-16-kjerner-43-GHz-AMD-AM5-Bulk-uten-vifte/3380475 The cpu will run on the apex just like a regular...

Here you go @misterh, updated my AM5 memory numbers 4585PX (which is a clone of the 9950X3D) @ delidded + custom watercooling 8844MT/s CL32 GDM disable 2211mhz FCLK All prefechers etc enabled, no special tweaks only for this benchmark CPU memory extreme = 12902 points CPU memory godlike =...

Time to update my submission with Zen5X3D cores + 5090 (y) Settings used: Hardware used: 1080p #1 Average = 981.2 fps P1 = 374.2 fps 1080p #2 Average = 975.3 fps P1 = 377.5 fps Lastly we have the 720p run (y) Average = 1027.1 fps P1 = 383 fps

Yes it pretty much behaves like a 9950X3D with a other name Everything else seem to be 100% the same

Baby 16 core EPYC reporting in ;) 4585PX is basically the server version/clone of the desktop 9950X3D with same boost limits and FMAX etc Blender version = 4.5 CPU score = 767.63 https://opendata.blender.org/benchmarks/65e8e6f9-3a97-4aa8-857a-545615fafeca/ GPU score = 17151.17...

@misterh Dont know if these results can be added, but you have a submission from a other forum here (y)

Small update [ R9 9950X3D @ CCD0 only + 32GB 8700 Cl32 | Astral 5090 | Custom watercooling | domdtxdissar ] 1440p ultra preset, DLSS and framegen disabled, TAA 2160p ultra preset, DLSS and framegen disabled, TAA

Little birdie told me i had to run with motion blur enabled [ R9 9950X3D @ CCD0 only + 32GB 8700 Cl32 | Astral 5090 | Custom watercooling | domdtxdissar ] 1440p ultra preset, DLSS and framegen disabled, TAA 2160p ultra preset, DLSS and framegen disabled, TAA

[ R9 9950X3D @ CCD0 only + 32GB 8700 Cl32 | Astral 5090 | Custom watercooling | domdtxdissar ] Quick and easy, without cherry picking runs 9950X3D simulating a 9800X3D (8 cores on CCD1 is disabled) 2x16gb a-die @ 8700MT/s 32-47-42-80 Watercooled Astral OC 5090 1080p ultra preset, DLSS and...

Quick and easy GPU run Dont know why i'm underperforming in the first test Extreme: Godmode:

Small update after I delidded retail sample Running PBO which makes the MT runs somewhat effective in power/performance since only pulling ~285w PPT R23 MT = 50687 R23 ST = 2413 R24 MT = 2914 R24 ST = 151 points

Small update after ive delidded retail sample Running PBO which makes the MT runs somewhat effective in power/performance since only pulling ~285w PPT R23 MT = 50687

Small update after ive delidded retail sample Running PBO which makes the MT runs somewhat effective in power/performance since only pulling ~285w PPT R24 MT = 2914 R24 ST = 151 points

IMC on the CPU is the limiting factor, i'm very happy with the memory:d It was clearly the best when i was binning CL26 @ 6400MT/s with different memory sets For CL32 8000MT 1.51VDD is required CL30 8000MT/s needs 1.71 VDD The only drawback is that they dont like GDM disabled above 8200MT/s...

So i have swapped over to my sample #3 9950X3D with the strongest IMC, and for the last week i have been finetuning the settings/voltages and pushing the memory as hard as i could for a fully stable daily memory profile At the end of my journey i reached a wooping fully stable 8744MT/s CL32...