mersenneforum.org  

Go Back   mersenneforum.org > Search Forums

Showing results 1 to 25 of 212
Search took 0.04 seconds.
Search: Posts Made By: xx005fs
Forum: GPU Computing 2020-12-25, 23:47
Replies: 3,457
Views: 332,774
Posted By xx005fs
Because TF uses integer operations, which is much...

Because TF uses integer operations, which is much faster on turing than on pascal.
Forum: Hardware 2020-11-20, 19:43
Replies: 124
Views: 7,563
Posted By xx005fs
Sadly they used aggregated result, if only they...

Sadly they used aggregated result, if only they listed the specific results for each scientific compute category (like GEMM, FFT, and N-Body) because it is well known that Ampere has good FP32...
Forum: GpuOwl 2020-11-02, 02:52
Replies: 2,686
Views: 200,222
Posted By xx005fs
Actually if those numbers are accurate, it is...

Actually if those numbers are accurate, it is highly probable that the 6900xt will be more than 2x the performance of 5700xt, since assuming the same FP32 to FP64 ratio with RDNA2 the FP32...
Forum: GpuOwl 2020-11-01, 17:41
Replies: 2,686
Views: 200,222
Posted By xx005fs
Thank you for the benchmark numbers. Very...

Thank you for the benchmark numbers. Very impressive performance from the A100, almost scales 1:1 with volta when comparing their memory bandwidth. I can't imagine the performance if the memory is...
Forum: GpuOwl 2020-10-23, 02:41
Replies: 2,686
Views: 200,222
Posted By xx005fs
Stock: 1600MHz Core, 850MHz HBM2 memory, 250W ...

Stock: 1600MHz Core, 850MHz HBM2 memory, 250W
gpuowl-win -prp 77936867 -maxAlloc 8192 -nospin
2020-10-22 19:30:10 gpuowl v7.0-66-gebe49cc
2020-10-22 19:30:10 Note: not found 'config.txt'...
Forum: Software 2020-09-21, 21:09
Replies: 35
Views: 3,857
Posted By xx005fs
There are already AIDA64 benchmarks on the 3080...

There are already AIDA64 benchmarks on the 3080 out in the wild such as https://www.overclockersclub.com/reviews/nvidia_geforce_rtx3080_founders_edition/4.htm
The single precision is 31523GFLOPs and...
Forum: Hardware 2020-09-03, 00:31
Replies: 173
Views: 20,316
Posted By xx005fs
It does make me wonder if AMD does actually have...

It does make me wonder if AMD does actually have something up their sleeve because it will be totally unreasonable for Nvidia to price the newer generation so low. I think they made the 3070 500$...
Forum: Hardware 2020-09-02, 16:14
Replies: 173
Views: 20,316
Posted By xx005fs
The 3080 looks promising for trial factoring, but...

The 3080 looks promising for trial factoring, but 700$ is still really expensive for a purely 80-class GPU even compared to last couple generations. Sadly though it doesn't seem to have a decent FP64...
Forum: GPU Computing 2020-08-10, 19:15
Replies: 3,457
Views: 332,774
Posted By xx005fs
That is blisteringly fast! Wonder if there will...

That is blisteringly fast! Wonder if there will be GPUOWL benchmark coming up for this beast. The 1.5+TB/s memory bandwidth and its near 10TFLOPs of FP64 capabilities will be very interesting.
Forum: GpuOwl 2020-07-01, 22:02
Replies: 4
Views: 889
Posted By xx005fs
Tested mfaktc, 0 difference between enabled and...

Tested mfaktc, 0 difference between enabled and disabled and CPU usage is equally low at near 0%.
Forum: GpuOwl 2020-07-01, 22:00
Replies: 4
Views: 889
Posted By xx005fs
Windows build 2004 and Nvidia Driver version...

Windows build 2004 and Nvidia Driver version 451.48
Forum: GPU Computing 2020-06-29, 18:38
Replies: 3,457
Views: 332,774
Posted By xx005fs
CUDA 11?

Any builds that support the newest CUDA 11? Hopefully there's something about CUDA 11 that could potentially bring some sort of speed up?
Forum: GpuOwl 2020-06-29, 04:02
Replies: 4
Views: 889
Posted By xx005fs
GpuOwl Nvidia Hardware Accelerated Scheduling Test (Windows)

With the newest Nvidia driver release on Windows along with the newest Windows update, there's apparently a new feature that is called hardware accelerated scheduling under the...
Forum: GPU Computing 2020-05-14, 19:55
Replies: 278
Views: 37,761
Posted By xx005fs
That's certainly possible, but I don't think many...

That's certainly possible, but I don't think many workstation users will use Radeon Pros over Quadros. However, I am sure that the release of new gaming cards will make the Radeon VII gamers dump...
Forum: GPU Computing 2020-05-14, 18:55
Replies: 278
Views: 37,761
Posted By xx005fs
Actually not bad when considering brand new, but...

Actually not bad when considering brand new, but the pro cards can't be overclocked so it will for sure be memory bound. However, RVII looks to be a much better deal when buying used.
Forum: Hardware 2020-05-07, 17:00
Replies: 844
Views: 140,417
Posted By xx005fs
It seems that Hyperthreading is only significant...

It seems that Hyperthreading is only significant for smaller FFT size since Ryzen 3000 series has a rather large cache. At the same time on smaller FFTs power consumption also goes through the roof...
Forum: GpuOwl 2020-05-07, 16:56
Replies: 2,686
Views: 200,222
Posted By xx005fs
PM1 Result not understood

It seems that for factored PM1 results out of GPUOWL, primenet won't be able to understand it.

{"status":"F", "exponent":"98141611", "worktype":"PM1", "B1":"750000", "B2":"15000000",...
Forum: Hardware 2020-05-06, 21:40
Replies: 844
Views: 140,417
Posted By xx005fs
More Ryzen 3900x data

CPU clockspeed locked at 4.2GHz, running DDR4 3733 with Dual channel and Dual Rank memory, FCLK at 1866MHz.
Prime95 64-bit version 29.8, RdtscTiming=1
Timings for 2048K FFT length (12 cores, 1...
Forum: Cloud Computing 2020-02-11, 03:42
Replies: 1,123
Views: 75,794
Posted By xx005fs
Interesting. Based on my experience the T4 is...

Interesting. Based on my experience the T4 is significantly slower though, like 6000+ vs K80's 3000 or so. Will experiment that tomorrow cuz i've reset so many times that google isn't giving me back...
Forum: Cloud Computing 2020-02-11, 02:22
Replies: 1,123
Views: 75,794
Posted By xx005fs
After the colab premium launch, it has been a...

After the colab premium launch, it has been a complete disaster for me since the only GPU I am receiving is the Tesla T4s, and I can't even get a half decent GPU like the K80 to run my PRP tasks....
Forum: Cloud Computing 2020-02-09, 20:35
Replies: 1,123
Views: 75,794
Posted By xx005fs
Is there anyway to run 1 session but execute 2...

Is there anyway to run 1 session but execute 2 programs at once, say executing primenet.py and GPUOWL to manage assignments or MISFIT and mfactx? or is there a need to open up a second session to run...
Forum: GPU Computing 2020-02-05, 05:56
Replies: 353
Views: 43,459
Posted By xx005fs
I am honestly really disappointed on how AMD is...

I am honestly really disappointed on how AMD is handling OpenCL right now, since they have basically neglected support of it on Windows machines, now ROCm 3.0 breaks OpenCL. I think what they need to...
Forum: JCoveiro 2020-02-03, 00:24
Replies: 12
Views: 9,307
Posted By xx005fs
Yes leave your memory overclocked but undervolt...

Yes leave your memory overclocked but undervolt the core. I don't have a Ryzen 3000 series CPU on hand so I can't proof that undervolting doesn't hurt that much but it's not going to be huge....
Forum: JCoveiro 2020-02-02, 19:12
Replies: 12
Views: 9,307
Posted By xx005fs
For GIMPS I seriously recommend undervolting your...

For GIMPS I seriously recommend undervolting your CPU. I dunno how it works on Ryzen 3000 series but I think setting a manual power limit to say 90W instead of the stock 140W wouldn't drop clockspeed...
Forum: GpuOwl 2020-02-02, 08:20
Replies: 2,686
Views: 200,222
Posted By xx005fs
Interesting. I saw only around 5% improvement...

Interesting. I saw only around 5% improvement going from CUDALucas to gpuowl on my 1070. Did RTX series get higher than 1/32 DP ratio?
Showing results 1 to 25 of 212

 
All times are UTC. The time now is 10:31.

Fri Jan 22 10:31:48 UTC 2021 up 50 days, 6:43, 0 users, load averages: 2.32, 2.06, 2.16

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.