mersenneforum.org  

Go Back   mersenneforum.org > Search Forums

Showing results 1 to 25 of 1000
Search took 0.14 seconds.
Search: Posts Made By: TheJudger
Forum: GpuOwl 2020-11-01, 13:37
Replies: 2,607
Views: 171,573
Posted By TheJudger
some Quick&Dirty benchmarks: A100 PCIe,...

some Quick&Dirty benchmarks:


A100 PCIe, reported clock rate and power consumption during run: 1215 MHz, 250W:# ./gpuowl.exe -iters 200000 -prp 77936867
2020-11-01 14:30:43 gpuowl...
Forum: GPU Computing 2020-11-01, 01:02
Replies: 2,832
Views: 234,612
Posted By TheJudger
CUDALucas 2.06, CUDA 11.1.1, Quick&Dirty run, did...

CUDALucas 2.06, CUDA 11.1.1, Quick&Dirty run, did a "./CUDALucas -cufftbench 2048 32768 20" before the following runs on each GPU.


A100 PCIe, actual clock rate 1200-1230 MHz and power 250 Watt...
Forum: GPU Computing 2020-10-31, 23:37
Replies: 3,434
Views: 308,526
Posted By TheJudger
Finally got in touch with a RTX 3090: mfaktc...

Finally got in touch with a RTX 3090:
mfaktc v0.22-pre8 (64bit built)
[...]
CUDA version info
binary compiled for CUDA 11.10
CUDA runtime version 11.10
CUDA driver version ...
Forum: GPU Computing 2020-07-28, 21:17
Replies: 3,434
Views: 308,526
Posted By TheJudger
I guess for many of us a RX 480 is far more...

I guess for many of us a RX 480 is far more enjoyable than a T4 for home usage (PC games :smile:)
Forum: GPU Computing 2020-07-28, 20:46
Replies: 3,434
Views: 308,526
Posted By TheJudger
Hi James, I can't remember a T4 hitting 2500...

Hi James,

I can't remember a T4 hitting 2500 GHz-d/d. And while that A100 has a TDP of 400 W it reports during mfaktc "just" 290 to 300 W. But you still might be correct that a T4 has a better...
Forum: GPU Computing 2020-07-28, 19:46
Replies: 3,434
Views: 308,526
Posted By TheJudger
Hi, seems like mfaktc runs fine with CUDA 11...

Hi,

seems like mfaktc runs fine with CUDA 11 on Ampere (no specific changes for Ampere except Makefile). :smile:

mfaktc v0.22-pre8 (64bit built)
[...]
CUDA version info
binary compiled for...
Forum: GPU Computing 2020-05-18, 19:17
Replies: 3,434
Views: 308,526
Posted By TheJudger
Yes, using atomics on CC 2.0 devices or newer....

Yes, using atomics on CC 2.0 devices or newer. Current artificial limit is 10 factors within class in a single "stage".

Oliver
Forum: JCoveiro 2020-02-03, 18:02
Replies: 12
Views: 6,185
Posted By TheJudger
Just check my recent results in the benchmark...

Just check my recent results in the benchmark thread. :smile:

Oliver
Forum: Hardware 2020-01-31, 21:21
Replies: 49
Views: 6,025
Posted By TheJudger
Hi, full benchmarks here:...

Hi,

full benchmarks here: https://mersenneforum.org/showpost.php?p=536336&postcount=788

Oliver
Forum: Hardware 2020-01-31, 21:20
Replies: 826
Views: 128,075
Posted By TheJudger
Hi, some fun with my Ryzen 9 3900X, I think most...

Hi, some fun with my Ryzen 9 3900X, I think most impressive is part 3!

BIOS defaults (142 W PPT), dual channel DDR4-2400 (dual rank):
2048K, 5760K, 6144K and 6400K flawed by some background...
Forum: Hardware 2020-01-29, 19:52
Replies: 826
Views: 128,075
Posted By TheJudger
Hi, you might want to include...

Hi,



you might want to include benchmarks for 2 workers and it is even better for certain ranges (FFT data fits twice into L3 cache(s) but not 4 times). Current LL Doublecheck fall into this...
Forum: Hardware 2020-01-28, 22:07
Replies: 49
Views: 6,025
Posted By TheJudger
Stock Ryzen 9 3900X with dual DDR4-3200 (dual...

Stock Ryzen 9 3900X with dual DDR4-3200 (dual rank).

Prime95 64-bit version 29.8, RdtscTiming=1
Timings for 2880K FFT length (12 cores, 1 worker): 1.33 ms. Throughput: 750.80 iter/sec.
Timings...
Forum: GPU Computing 2019-09-11, 19:36
Replies: 3,434
Views: 308,526
Posted By TheJudger
That perfectly explains why it crashes at 2048......

That perfectly explains why it crashes at 2048... :sad:

Oliver
Forum: GPU Computing 2019-09-11, 18:16
Replies: 3,434
Views: 308,526
Posted By TheJudger
Actually I didn't spent much time on thinking...

Actually I didn't spent much time on thinking about this. I'm not sure wheter TF to 255 hits the wrap around or not.
I don't have any evidence that 2047 doesn't work, I'm just not a fan of "changed...
Forum: GPU Computing 2019-09-11, 17:43
Replies: 3,434
Views: 308,526
Posted By TheJudger
Hi, unless you're really sure about the...

Hi,

unless you're really sure about the increased sieve size limit I suggest to stay with 1024...
"Doesn't crash" and "passes the builtin selftest" doesn't prove that 2047 is OK. 2048 crashes...
Forum: GPU Computing 2019-07-21, 20:17
Replies: 3,434
Views: 308,526
Posted By TheJudger
Hi there, I was recently able to "reproduce"...

Hi there,

I was recently able to "reproduce" the issue where mfaktc reports 38814612911305349835664385407 as a (false) factor of M<insert prime number here>. While the origin of the factor is well...
Forum: GPU Computing 2019-06-07, 22:23
Replies: 3,434
Views: 308,526
Posted By TheJudger
Yes, all you need to do is to modify the Makefile...

Yes, all you need to do is to modify the Makefile as you already did!

Oliver
Forum: Hardware 2019-05-29, 19:17
Replies: 54
Views: 2,709
Posted By TheJudger
AFAIK most (all expect prefetch?) of AMDs 3Dnow...

AFAIK most (all expect prefetch?) of AMDs 3Dnow instructions are no longer available on recent AMD CPUs. Not sure whether you call this x86 or not.

Oliver
Forum: GPU Computing 2019-03-15, 19:21
Replies: 3,434
Views: 308,526
Posted By TheJudger
No problem. And yes, those run to run variations...

No problem. And yes, those run to run variations are annoying. On a stock Geforce you have powertarget, temperature target, actual temperature and so on. Even when you try to lock a specific...
Forum: GPU Computing 2019-03-14, 22:56
Replies: 3,434
Views: 308,526
Posted By TheJudger
Are you sure about this? I'm not! Hint: check...

Are you sure about this? I'm not! Hint: check kernel_possible() in the same file.

Last time I did some benchmarks barrett 87 and 88 was faster than 77 (Pascal series).

Oliver
Forum: Data 2019-02-23, 18:04
Replies: 211
Views: 19,386
Posted By TheJudger
Correct, that kernel can handle numbers up to...

Correct, that kernel can handle numbers up to 276. And the lower limit for all current barrett based kernels in mfaktc is 264.

Oliver
Forum: GPU to 72 2019-02-18, 18:08
Replies: 5,049
Views: 247,434
Posted By TheJudger
Once the runtime per class is lets say something...

Once the runtime per class is lets say something like 0.5s it is more efficient to do it step by step (read: stages or just on bitlevel at a time).
Forum: Mlucas 2019-02-16, 22:42
Replies: 11
Views: 2,058
Posted By TheJudger
Hi just add a counter to your signal...

Hi



just add a counter to your signal handler. Hint: on Windows you have to re-register your signal handler once it was triggered.

Oliver
Forum: PrimeNet 2019-02-10, 01:36
Replies: 2,037
Views: 160,399
Posted By TheJudger
xxx005fs: doing some work on GPUs an submitting...

xxx005fs: doing some work on GPUs an submitting manually? If so than this is normal!

Oliver
Forum: GPU Computing 2019-02-09, 20:04
Replies: 3,434
Views: 308,526
Posted By TheJudger
GPU sieve

GPU sieve
Showing results 1 to 25 of 1000

 
All times are UTC. The time now is 06:31.

Thu Nov 26 06:31:02 UTC 2020 up 77 days, 3:42, 3 users, load averages: 1.77, 1.66, 1.53

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, 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.