Leaderboard

Some notes: - 2650Mhz is the boost setting applied from Performance Tuning panel, not the real clock during load (GPU is power-limited) - Tesselation modified, Not 3DMark Hall of Fame valid (valid score with similar setup around 45K - https://www.3dmark.com/fs/24052851 - It's rather difficult managing the cold bug on the 2-CCD CPU, this one lost the GPU (d6 error) and generally having errors at -100C, changing FCLK to lower doesn't help, not sure exactly what's holding it back -It's been a LONG time since we saw an AMD CPU and GPU on any of the 3DMark, this is refreshing

There are other comps? 🤔

Guys there is a few bugs on the latest Beta bios 3.62 released 11/11/2020 ( Enhance AMD Ryzen 5000 CPU performance ) Bug # 1 - Cannot use more voltage for v.mem higher than 1.82v , system will go dark screen mode and freeze until clear cmos ( i need at least 1.985v. for extreme overclocking in subsero temps.) Bug # 2 - Bios won't allow IF ratio properly if memory is overclock higher than 3800Mhz it will revert back to stock settings , IF/Memory ratio is completely broken and memory won't do more than 3800Mhz due to this bug. Bug # 3 - A lot less cpu performance than other boards on th

Here give it a try , but neither i or splave are responsible if something goes wrong ( use at your own risk , it is a Beta bios ) http://www.mediafire.com/file/nfvjsn0vvg5o9aw/tc358.zip/file

This score is insane ! Absolutely massive score ... maaaassssssive 👍. Well done Alva ! 🔥

Holy shit! Now wait for 6900xt! Might it be even stronger?

Good Job mate, nice Card

Shared on FB, looking good

Yes, it was weird to me too, even with Tess ON, the default score is close to 50K Graphics on the 6800 XT. TS and PR is pretty OK, but not this strong. On raster-based games it was around 3080-level, not 3090. (On DXR games beating 2080 Ti FE by small percentage,but still behind 3080 by a decent margin)

Simply amazing!

Wow... good job

Awesome!

THX , FCLK = 1866 = CB -178 🙂

Congrats on the first RX 6800 XT results at HWBot!

Optimal list is something like 1090t, qx9650, x5470, and ??? Og phenom could be used but they kinda suck so may lose to qx9300 or socket F

Personally I found it scored similar on my E8400 and 2600x

Thats indeed what is in the pipeline, points for everyone, no more cutoffs for scores outside the 70% margin of the nr 1 score. But again that's 2021 biss, now we need to make the site more secure, stable and with an up to date benchmark list before we go into 2021.

My 5 cent. First day after release in house results were always strong and yes it can destroy the spirit of an average Joe even to try to compete. But it is not the issue or to blame the 1st places. I see no competition in the runner up category where it was enoromus presence in the past. Nowdays it can easily happen that between Allen's WR and a watercooled daily rig's result its like 3 to 5 entry uploaded and even the half of it is subzero. That is even more killing the will to enter. I think one root cause is the pointing system. Let's get the GPUPI for CPU - 1B as example, 6x global

Its often confusing which category to choose for the IGP part of a APU, so here is a incomplete list to help find you the correct one. AMD ZEN: AMD choose to change their name scheme in 2020. Before we had Radeon Vega 3 to 11 Graphics, now it's just "Radeon Graphics" independent of existing shaders. Also there is actually "RX Vega" for everything from Vega 10 but for some reason and for some cards they omitted that. In late 2019 they introduced with the Athlon 3000G a new entry level CPU featuring a native 2 core die with a 3 CU graphics part in the desktop market. But for some

https://kingpincoolingeuropesupplie.com/ seal2fast@kingpincooling.com Check it out

There is LOD for DX9 and DX11, but if you enabling LOD on DX11 you can't have Tesselation advantages.

I think that there need to be a time to mature/quarantine new hardware and new benchmarks before they are used in competitions/point giving challenges. like 2-3 months even if scores are posted on the bot. For hardware. Based on that some new ryzens have even been noticed to break even CPUz with wrong reporting. Thus any new hardware need to be going through informal tests by users and experts on a number of benchers before they are confirmed. To for example update benchmate or other envelope. For benchmarks. Based on that its not fair to have updates in the middle of