mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Hardware > GPU Computing > GpuOwl

Reply
 
Thread Tools
Old 2021-05-02, 21:40   #2707
Xyzzy
 
Xyzzy's Avatar
 
Aug 2002

2·3·5·277 Posts
Default

In general it seems like:

AMD advantage: Better FP64 ratio & open Linux drivers.

Nvidia advantage: CUDA & better (non-free) drivers.

Xyzzy is offline   Reply With Quote
Old 2021-05-11, 00:48   #2708
Xyzzy
 
Xyzzy's Avatar
 
Aug 2002

2·3·5·277 Posts
Default

How do you configure gpuowl to use a little bit less than maximum?

It would be nice to leave gpuowl running when we use a web browser.

We do have -yield enabled which doesn't seem to do much?

Xyzzy is offline   Reply With Quote
Old 2021-05-12, 10:30   #2709
drkirkby
 
"David Kirkby"
Jan 2021
Althorne, Essex, UK

3×149 Posts
Default

Here's a benchmark result that may be helpful for an Nvida P2200 graphics card in a Dell 7920 fitted with a pair of Xeon 8167Ms and 384 GB RAM. The GPU is considerably slower than the CPUs - see a more detailed discussion at



https://www.mersenneforum.org/showthread.php?p=578245


Here's the benchmark, running the exponent you want.



Code:
drkirkby@jackdaw:~/gpuowl/gpuowl-master$ ./gpuowl -maxAlloc 5G -prp 77936867
2021-05-12 11:00:26 GpuOwl VERSION 
2021-05-12 11:00:26 GpuOwl VERSION 
2021-05-12 11:00:26 Note: not found 'config.txt'
2021-05-12 11:00:26 config: -maxAlloc 5G -prp 77936867 
2021-05-12 11:00:26 device 0, unique id ''
2021-05-12 11:00:27 Quadro P2200-0 77936867 FFT: 4M 1K:8:256 (18.58 bpw)
2021-05-12 11:00:27 Quadro P2200-0 77936867 OpenCL args "-DEXP=77936867u -DWIDTH=1024u -DSMALL_HEIGHT=256u -DMIDDLE=8u -DMM_CHAIN=1u -DMM2_CHAIN=2u -DMAX_ACCURACY=1 -DWEIGHT_STEP=0.33644726404543274 -DIWEIGHT_STEP=-0.25174750481886216 -DIWEIGHTS={0,-0.44011820345520131,-0.37306474779553728,-0.29798072935699788,-0.21390437908665341,-0.11975874301407295,-0.014337887291734644,-0.44814572555075455,} -DFWEIGHTS={0,0.78609128957452257,0.5950610473469905,0.42446232150303748,0.2721098723818392,0.1360521812214803,0.014546452690911484,0.81207258201996746,}  -cl-std=CL2.0 -cl-finite-math-only "
2021-05-12 11:00:27 Quadro P2200-0 77936867 

2021-05-12 11:00:27 Quadro P2200-0 77936867 OpenCL compilation in 0.01 s
2021-05-12 11:00:27 Quadro P2200-0 77936867 trig table : 65 points, cos 73.98 bits, sin 73.34 bits
2021-05-12 11:00:27 Quadro P2200-0 77936867 trig table : 257 points, cos 73.98 bits, sin 73.11 bits
2021-05-12 11:00:27 Quadro P2200-0 77936867 trig table : 262145 points, cos 72.64 bits, sin 72.56 bits
2021-05-12 11:00:28 Quadro P2200-0 77936867 maxAlloc: 5.0 GB
2021-05-12 11:00:28 Quadro P2200-0 77936867 P1(0) 0 bits
2021-05-12 11:00:28 Quadro P2200-0 77936867 PRP starting from beginning
2021-05-12 11:00:31 Quadro P2200-0 77936867 OK         0 on-load: blockSize 400, 0000000000000003
2021-05-12 11:00:31 Quadro P2200-0 77936867 validating proof residues for power 8
2021-05-12 11:00:31 Quadro P2200-0 77936867 Proof using power 8
2021-05-12 11:00:40 Quadro P2200-0 77936867 OK       800   0.00% 1579c241dc63eca6 7271 us/it + check 3.12s + save 0.19s; ETA 6d 13:24
2021-05-12 11:01:47 Quadro P2200-0 77936867        10000   0.01% fc4f135f7cf4ad29 7241 us/it
2021-05-12 11:02:59 Quadro P2200-0 77936867        20000   0.03% 3cd1bd9d5e09cbc5 7209 us/it
2021-05-12 11:04:10 Quadro P2200-0 77936867        30000   0.04% c4e0ff35e3290d98 7131 us/it
2021-05-12 11:05:22 Quadro P2200-0 77936867        40000   0.05% dffe1b1b0d748128 7210 us/it
2021-05-12 11:06:34 Quadro P2200-0 77936867        50000   0.06% 52e286945371ed29 7129 us/it
2021-05-12 11:07:44 Quadro P2200-0 77936867        60000   0.08% 0945da4dc08bdd95 7075 us/it
2021-05-12 11:08:55 Quadro P2200-0 77936867        70000   0.09% 7131fa4eb77f4bb2 7100 us/it
2021-05-12 11:10:08 Quadro P2200-0 77936867        80000   0.10% 8d76071d27ee4221 7237 us/it
2021-05-12 11:11:20 Quadro P2200-0 77936867        90000   0.12% 0bacff453b2f470e 7236 us/it
2021-05-12 11:12:33 Quadro P2200-0 77936867       100000   0.13% 6d7296b9e2830f50 7246 us/it
2021-05-12 11:13:46 Quadro P2200-0 77936867       110000   0.14% 8cbfd4435622bda7 7354 us/it
2021-05-12 11:14:59 Quadro P2200-0 77936867       120000   0.15% 79ae5dad855057ad 7271 us/it
2021-05-12 11:16:11 Quadro P2200-0 77936867       130000   0.17% 50c97bcbf876231f 7205 us/it
2021-05-12 11:17:22 Quadro P2200-0 77936867       140000   0.18% e1db15f897271496 7138 us/it
2021-05-12 11:18:34 Quadro P2200-0 77936867       150000   0.19% 127631386c6a9b17 7156 us/it
2021-05-12 11:19:45 Quadro P2200-0 77936867       160000   0.21% 25b7b6206fc6f085 7096 us/it
2021-05-12 11:20:56 Quadro P2200-0 77936867       170000   0.22% 416816b0d9f4bba8 7101 us/it
2021-05-12 11:22:07 Quadro P2200-0 77936867       180000   0.23% 6bee5d054f770861 7100 us/it
2021-05-12 11:23:18 Quadro P2200-0 77936867       190000   0.24% f37f068f014b18a0 7101 us/it
2021-05-12 11:24:32 Quadro P2200-0 77936867 OK    200000   0.26% f0b04b45b0855bd2 7099 us/it + check 2.95s + save 0.19s; ETA 6d 09:18
2021-05-12 11:25:43 Quadro P2200-0 77936867       210000   0.27% 43eb2fc2424d8aac 7063 us/it
2021-05-12 11:26:53 Quadro P2200-0 77936867       220000   0.28% a1081c6dc6a7689f 7064 us/it
2021-05-12 11:28:05 Quadro P2200-0 77936867       230000   0.30% 2387818d3d3d0d01 7202 us/it
drkirkby is offline   Reply With Quote
Old 2021-05-12, 13:37   #2710
moebius
 
moebius's Avatar
 
Jul 2009
Germany

2·313 Posts
Default

Thx, sheet is updated! Always bring up the values....

Last fiddled with by moebius on 2021-05-12 at 13:41
moebius is offline   Reply With Quote
Old 2021-05-15, 16:00   #2711
De Wandelaar
 
De Wandelaar's Avatar
 
"Yves"
Jul 2017
Belgium

73 Posts
Default

Benchmark RTX 5000 under windows.
Iteration between 2806 en 2841 microsec.
See attach :
Attached Thumbnails
Click image for larger version

Name:	GPUowl  RTX 5000.jpg
Views:	94
Size:	117.2 KB
ID:	24893  

Last fiddled with by De Wandelaar on 2021-05-15 at 16:01 Reason: typo
De Wandelaar is offline   Reply With Quote
Old 2021-05-16, 01:04   #2712
moebius
 
moebius's Avatar
 
Jul 2009
Germany

2×313 Posts
Smile

I took the trouble to get a benchmark value for the AMD RX 6800 XT graphics card for the users of this forum. I thank the users of guru3d.com
for their fast support. A little bit better than the Tesla P100!
Attached Thumbnails
Click image for larger version

Name:	RX6800XT.JPG
Views:	136
Size:	87.3 KB
ID:	24899  

Last fiddled with by moebius on 2021-05-16 at 01:16
moebius is offline   Reply With Quote
Old 2021-05-16, 15:38   #2713
moebius
 
moebius's Avatar
 
Jul 2009
Germany

2×313 Posts
Default

Vega56, flashed with an XFX V64 BIOS, 1672P7, 1100HBM
Attached Thumbnails
Click image for larger version

Name:	Vega56.png
Views:	75
Size:	35.1 KB
ID:	24900  
moebius is offline   Reply With Quote
Old 2021-05-16, 16:00   #2714
moebius
 
moebius's Avatar
 
Jul 2009
Germany

2·313 Posts
Default

"The values in the post above are from a MSI Radeon 6800 XT Gaming X Trio"

Last fiddled with by moebius on 2021-05-16 at 16:01
moebius is offline   Reply With Quote
Old 2021-05-17, 21:02   #2715
moebius
 
moebius's Avatar
 
Jul 2009
Germany

10011100102 Posts
Default

I have changed the gpuowl benchmark list accordingly to the estimated values ​​of the Quadro GV100 and the Quadro GP100 based on the values ​​available for the GV100 and GP100 chipset. I still dream of getting values ​​for the graphics card accelerator MI Instinct 100, which in my opinion should perform better than the Nvidia A100.I simply have no comparison values ​​for the CDNA architecture.
gpuowl ComparisonList https://drive.google.com/file/d/10fC...enkBdAaRP/view
moebius is offline   Reply With Quote
Old 2021-06-06, 09:26   #2716
vasyannyasha
 
vasyannyasha's Avatar
 
"Vasiliy"
Apr 2017
Ukraine

3F16 Posts
Default

RTX 3060 Mobile 4250-4300 us/it

Code:
2021-06-06 11:55:43 GpuOwl VERSION v7.2-69-g23c14a1
2021-06-06 11:55:43 Note: not found 'config.txt'
2021-06-06 11:55:43 device 0, unique id ''
2021-06-06 11:55:44 NVIDIA GeForce RTX 3060 Laptop GPU-0 Bye
2021-06-06 11:56:26 GpuOwl VERSION v7.2-69-g23c14a1
2021-06-06 11:56:26 Note: not found 'config.txt'
2021-06-06 11:56:26 config: -prp 77936867 -maxAlloc 4.5G 
2021-06-06 11:56:26 device 0, unique id ''
2021-06-06 11:56:27 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 FFT: 4M 1K:8:256 (18.58 bpw)
2021-06-06 11:56:28 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 OpenCL args "-DEXP=77936867u -DWIDTH=1024u -DSMALL_HEIGHT=256u -DMIDDLE=8u -DMM_CHAIN=1u -DMM2_CHAIN=2u -DMAX_ACCURACY=1 -DWEIGHT_STEP=0.33644726404543274 -DIWEIGHT_STEP=-0.25174750481886216 -DIWEIGHTS={0,-0.44011820345520131,-0.37306474779553728,-0.29798072935699788,-0.21390437908665341,-0.11975874301407295,-0.014337887291734644,-0.44814572555075455,} -DFWEIGHTS={0,0.78609128957452257,0.5950610473469905,0.42446232150303748,0.2721098723818392,0.1360521812214803,0.014546452690911484,0.81207258201996746,}  -cl-std=CL2.0 -cl-finite-math-only "
2021-06-06 11:56:28 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 

2021-06-06 11:56:28 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 OpenCL compilation in 0.02 s
2021-06-06 11:56:28 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 trig table : 65 points, cos 73.98 bits, sin 73.34 bits
2021-06-06 11:56:28 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 trig table : 257 points, cos 73.98 bits, sin 73.11 bits
2021-06-06 11:56:28 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 trig table : 262145 points, cos 72.64 bits, sin 72.56 bits
2021-06-06 11:56:28 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 maxAlloc: 4.5 GB
2021-06-06 11:56:28 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 P1(0) 0 bits
2021-06-06 11:56:28 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 PRP starting from beginning
2021-06-06 11:56:30 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 OK         0 on-load: blockSize 400, 0000000000000003
2021-06-06 11:56:30 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 validating proof residues for power 8
2021-06-06 11:56:30 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 Proof using power 8
2021-06-06 11:56:35 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 OK       800   0.00% 1579c241dc63eca6 4157 us/it + check 1.70s + save 0.14s; ETA 3d 17:59
2021-06-06 11:57:14 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867        10000   0.01% fc4f135f7cf4ad29 4228 us/it
2021-06-06 11:57:57 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867        20000   0.03% 3cd1bd9d5e09cbc5 4244 us/it
2021-06-06 11:58:40 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867        30000   0.04% c4e0ff35e3290d98 4300 us/it
2021-06-06 11:59:23 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867        40000   0.05% dffe1b1b0d748128 4305 us/it
2021-06-06 12:00:05 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867        50000   0.06% 52e286945371ed29 4267 us/it
2021-06-06 12:00:48 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867        60000   0.08% 0945da4dc08bdd95 4242 us/it
2021-06-06 12:01:30 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867        70000   0.09% 7131fa4eb77f4bb2 4259 us/it
2021-06-06 12:02:13 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867        80000   0.10% 8d76071d27ee4221 4249 us/it
2021-06-06 12:02:55 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867        90000   0.12% 0bacff453b2f470e 4257 us/it
2021-06-06 12:03:38 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867       100000   0.13% 6d7296b9e2830f50 4268 us/it
2021-06-06 12:04:21 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867       110000   0.14% 8cbfd4435622bda7 4248 us/it
2021-06-06 12:05:03 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867       120000   0.15% 79ae5dad855057ad 4244 us/it
2021-06-06 12:05:45 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867       130000   0.17% 50c97bcbf876231f 4248 us/it
2021-06-06 12:06:28 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867       140000   0.18% e1db15f897271496 4261 us/it
2021-06-06 12:07:11 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867       150000   0.19% 127631386c6a9b17 4267 us/it
2021-06-06 12:07:53 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867       160000   0.21% 25b7b6206fc6f085 4246 us/it
2021-06-06 12:08:36 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867       170000   0.22% 416816b0d9f4bba8 4264 us/it
2021-06-06 12:09:18 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867       180000   0.23% 6bee5d054f770861 4250 us/it
2021-06-06 12:10:01 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867       190000   0.24% f37f068f014b18a0 4303 us/it
2021-06-06 12:10:46 NVIDIA GeForce RTX 3060 Laptop GPU-0 77936867 OK    200000   0.26% f0b04b45b0855bd2 4309 us/it + check 1.77s + save 0.14s; ETA 3d 21:02
Attached Thumbnails
Click image for larger version

Name:	vid.jpg
Views:	50
Size:	56.9 KB
ID:	25034   Click image for larger version

Name:	sens.jpg
Views:	44
Size:	67.1 KB
ID:	25035  
vasyannyasha is offline   Reply With Quote
Old 2021-07-06, 01:12   #2717
storm5510
Random Account
 
storm5510's Avatar
 
Aug 2009

3·661 Posts
Default

Question: Why is it necessary for the the "-cpu" configuration item to be repeated in every line? It seems to serve no other purpose than to take up a lot of space.
Attached Thumbnails
Click image for larger version

Name:	owl.JPG
Views:	55
Size:	40.3 KB
ID:	25231  
storm5510 is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
mfakto: an OpenCL program for Mersenne prefactoring Bdot GPU Computing 1680 2021-09-13 17:01
GPUOWL AMD Windows OpenCL issues xx005fs GpuOwl 0 2019-07-26 21:37
Testing an expression for primality 1260 Software 17 2015-08-28 01:35
Testing Mersenne cofactors for primality? CRGreathouse Computer Science & Computational Number Theory 18 2013-06-08 19:12
Primality-testing program with multiple types of moduli (PFGW-related) Unregistered Information & Answers 4 2006-10-04 22:38

All times are UTC. The time now is 18:58.


Thu Oct 21 18:58:16 UTC 2021 up 90 days, 13:27, 1 user, load averages: 2.25, 2.06, 1.65

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.