mersenneforum.org  

Go Back   mersenneforum.org > Factoring Projects > Operazione Doppi Mersennes
Reload this Page Benchmarks
Register FAQ Search Today's Posts Mark Forums Read

Reply
Page 1 of 2 1 2
 
Thread Tools
Old 2012-09-19, 08:05   #1
ET_
Banned
 
ET_'s Avatar
 
"Luigi"
Aug 2002
Team Italia

4,871 Posts
Cool Benchmarks
Hi everybody!

I am reworking productivity and testing pages on FermatSearch (http://www.fermatsearch.org/productivity.html ).

I'd need data (specifically, k/sec) for mmff working on Fermat numbers, specifically for the following ranges:

N = 28, 32, 36, 40, 44, 48, 54, 60, 80, 100, 150.

It would be fine if we could fill up the following table:

Range: k=1e15 - 1.1e15

Code:
        ----------------------------------------------------------------------------------------------------
	|   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 460 | 951 M/s| 941 M/s| 835 M/s| 483 M/s| 437 M/s| 395 M/s| 350 M/s| 303 M/s| 169 M/s| 144 M/s|  63 M/s|
GTX 470 |1560 M/s|        |        |        |        |        |        |        |        |        |        |
GTX 4802|1866 M/s|1812 M/s|1634 M/s|1010 M/s| 920 M/s| 840 M/s| 747 M/s| 651 M/s| 358 M/s| 304 M/s| 138 M/s|
GTX 560 |1243 M/s|1204 M/s|1079 M/s| 654 M/s| 593 M/s| 540 M/s| 479 M/s| 416 M/s| 235 M/s| 199 M/s|  87 M/s|
GTX 570 |        |        |        |        |        |        |        |        |        |        |        |
GTX 5801|2303 M/s|2232 M/s|2023 M/s|1230 M/s|1116 M/s|1021 M/s| 902 M/s| 784 M/s| 442 M/s| 372 M/s| 120 M/s|
GTX 6802|1689 M/s|1645 M/s|1473 M/s| 868 M/s| 789 M/s| 705 M/s| 625 M/s| 543 M/s| 303 M/s| 258 M/s|  -  M/s|
------------------------------------------------------------------------------------------------------------
1 - GPUSieveSize=16(default), GPUSieveProcessSize=8 (default).
2 - GPUSievePrimes default.

GPU SievePrimes:
Code:
        ----------------------------------------------------------------------------------------------------
	|   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 460 |  90000 |  80000 |  90000 | 200000 | 210000 | 210000 | 240000 | 270000 | 600000 | 650000 | 950000 |
GTX 470 |  70000 |        |        |        |        |        |        |        |        |        |        |
GTX 480 |        |        |        |        |        |        |        |        |        |        |        |
GTX 560 |  70000 |  70000 |  80000 | 120000 | 140000 | 150000 | 180000 | 230000 | 300000 | 400000 | 700000 |
GTX 570 |        |        |        |        |        |        |        |        |        |        |        |
GTX 580 | 100000 | 100000 |  82486 | 100000 | 150000 | 150000 | 200000 | 200000 | 300000 | 300000 | 900000 |
GTX 680 |        |        |        |        |        |        |        |        |        |        |        |
------------------------------------------------------------------------------------------------------------

Thanks, Luigi
Last fiddled with by ET_ on 2013-02-24 at 13:07
ET_ is offline   Reply With Quote
Old 2012-09-19, 14:23   #2
ATH
Einyen
 
ATH's Avatar
 
Dec 2003
Denmark

3×1,151 Posts
Default
Remember to find the optimal GPUSievePrimes for each n. Do you need data for GTX 460?
ATH is offline   Reply With Quote
Old 2012-09-19, 15:03   #3
ET_
Banned
 
ET_'s Avatar
 
"Luigi"
Aug 2002
Team Italia

4,871 Posts
Default
Quote:
Originally Posted by ATH View Post
Remember to find the optimal GPUSievePrimes for each n. Do you need data for GTX 460?
Yes, why not?... Thanks.

Luigi
ET_ is offline   Reply With Quote
Old 2012-09-19, 18:27   #4
ATH
Einyen
 
ATH's Avatar
 
Dec 2003
Denmark

3·1,151 Posts
Default
Code:
        ----------------------------------------------------------------------------------------------------
	|   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 460 | 938 M/s| 926 M/s| 822 M/s| 740 M/s| 676 M/s| 395 M/s| 350 M/s| 311 M/s| 170 M/s| 142 M/s| 62 M/s |

GPUSievePrimes:
Code:
        ----------------------------------------------------------------------------------------------------
	|   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 460 |  80000 |  90000 | 100000 | 120000 | 160000 | 210000 | 240000 | 270000 | 450000 | 650000 | 900000 |

Note that currently mmff only works for k>=224 and k*2n+1 <= 2188, so for n=150 you can only do:
FermatFactor=150,174,188
which only takes 16min on a GTX 460.

The highest working n=159 only works for:
FermatFactor=159,183,188
which takes 40sec.

The highest working n for which there is still work to do < 2188 according to: http://www.fermatsearch.org/stat/done.php
is n=146:
FermatFactor=146,3e12,188
Last fiddled with by ATH on 2012-09-19 at 18:39
ATH is offline   Reply With Quote
Old 2012-09-19, 19:43   #5
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

2×3×7×199 Posts
Default
Thanks for reporting the GPUSievePrimes too. I'm fairly certain that the optimum setting won't change much for each Nvidia model.

Note that timings will change as k increases. As the bit size of the factor increases, wider multiplications are required.

Let's specify a specific k range to benchmark - something representative of current reservations (like 1e15 to 1.1e15).
Prime95 is offline   Reply With Quote
Old 2012-09-20, 03:25   #6
ATH
Einyen
 
ATH's Avatar
 
Dec 2003
Denmark

1101011111012 Posts
Default
Quote:
Originally Posted by Prime95 View Post
Thanks for reporting the GPUSievePrimes too. I'm fairly certain that the optimum setting won't change much for each Nvidia model.

Note that timings will change as k increases. As the bit size of the factor increases, wider multiplications are required.

Let's specify a specific k range to benchmark - something representative of current reservations (like 1e15 to 1.1e15).
Ok, I'll need to rerun test tomorrow then at new k if we agree on 1e15-1.1e15.

Btw GPUSievePrimes have a range that gives the same speed and the range is longer for higher n. For n=150 basically 600k to 1075k give about the same speed.
ATH is offline   Reply With Quote
Old 2012-09-20, 08:28   #7
ET_
Banned
 
ET_'s Avatar
 
"Luigi"
Aug 2002
Team Italia

114078 Posts
Default
Quote:
Originally Posted by ATH View Post
Ok, I'll need to rerun test tomorrow then at new k if we agree on 1e15-1.1e15.
[...]
I agree.

The first post has been updated.

Thanks ATH.

Luigi
ET_ is offline   Reply With Quote
Old 2012-09-21, 06:37   #8
ATH
Einyen
 
ATH's Avatar
 
Dec 2003
Denmark

345310 Posts
Default
k=1e15-1.1e15:
(except n=150 as I said, I can only do 2187 to 2188 ~ k=1.37*1011-2.74*1011)

Code:
        ----------------------------------------------------------------------------------------------------
	|   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 460 | 951 M/s| 941 M/s| 835 M/s| 483 M/s| 437 M/s| 395 M/s| 350 M/s| 303 M/s| 169 M/s| 144 M/s| 63 M/s |


GPUSievePrimes:
        ----------------------------------------------------------------------------------------------------
	|   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 460 |  90000 |  80000 |  90000 | 200000 | 210000 | 210000 | 240000 | 270000 | 600000 | 650000 | 950000 |
Last fiddled with by ATH on 2012-09-21 at 06:38
ATH is offline   Reply With Quote
Old 2012-09-24, 10:11   #9
Jatheski
 
Jatheski's Avatar
 
Apr 2012
993438: i1090

2·73 Posts
Default
Quote:
Originally Posted by ATH View Post
k=1e15-1.1e15:
(except n=150 as I said, I can only do 2187 to 2188 ~ k=1.37*1011-2.74*1011)
Benchmark for a GTX560

Code:
        ----------------------------------------------------------------------------------------------------
        |   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 560 |1243 M/s|1204 M/s|1079 M/s| 654 M/s| 593 M/s| 540 M/s| 479 M/s| 416 M/s| 235 M/s| 199 M/s|  87M/s |


GPUSievePrimes:
        ----------------------------------------------------------------------------------------------------
        |   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 560 |  70000 |  70000 |  80000 | 120000 | 140000 | 150000 | 180000 | 230000 | 300000 | 400000 | 700000 |
Jatheski is offline   Reply With Quote
Old 2012-09-24, 10:17   #10
ET_
Banned
 
ET_'s Avatar
 
"Luigi"
Aug 2002
Team Italia

4,871 Posts
Default
Quote:
Originally Posted by Jatheski View Post
Benchmark for a GTX560

Code:
        ----------------------------------------------------------------------------------------------------
        |   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 560 |1243 M/s|1204 M/s|1079 M/s| 654 M/s| 593 M/s| 540 M/s| 479 M/s| 416 M/s| 235 M/s| 199 M/s|  87M/s |


GPUSievePrimes:
        ----------------------------------------------------------------------------------------------------
        |   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 560 |  70000 |  70000 |  80000 | 120000 | 140000 | 150000 | 180000 | 230000 | 300000 | 400000 | 700000 |
Thanks Fabrice!

Luigi
ET_ is offline   Reply With Quote
Old 2012-09-27, 17:31   #11
frmky
 
frmky's Avatar
 
Jul 2003
So Cal

7·383 Posts
Default
Here's my GTX 480. I didn't optimize GPUSievePrimes, but instead let it just use the mmff-0.25 defaults.

Code:
        ----------------------------------------------------------------------------------------------------
	|   28   |   32   |   36   |   40   |   44   |   48   |   54   |   60   |   80   |  100   |  150   |
------------------------------------------------------------------------------------------------------------
GTX 480 |1866 M/s|1812 M/s|1634 M/s|1010 M/s| 920 M/s| 840 M/s| 747 M/s| 651 M/s| 358 M/s| 304 M/s| 138 M/s|
frmky is offline   Reply With Quote
Reply
Page 1 of 2 1 2

« Previous Thread | Next Thread »
Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Benchmarks MurrayInfoSys Information & Answers 3 2011-04-14 17:10
LLR benchmarks Oddball No Prime Left Behind 11 2010-08-06 21:39
benchmarks Unregistered Information & Answers 15 2009-08-18 16:44
Benchmarks for i7 965 lavalamp Hardware 21 2009-01-06 04:32
Benchmarks Vandy Hardware 6 2002-10-28 13:45

All times are UTC. The time now is 04:50.


Sun Sep 24 04:50:17 UTC 2023 up 11 days, 2:32, 0 users, load averages: 1.31, 1.21, 1.24

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2023, 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.

≠ ± ∓ ÷ × · − √ ‰ ⊗ ⊕ ⊖ ⊘ ⊙ ≤ ≥ ≦ ≧ ≨ ≩ ≺ ≻ ≼ ≽ ⊏ ⊐ ⊑ ⊒ ² ³ °
∠ ∟ ° ≅ ~ ‖ ⟂ ⫛
≡ ≜ ≈ ∝ ∞ ≪ ≫ ⌊⌋ ⌈⌉ ∘ ∏ ∐ ∑ ∧ ∨ ∩ ∪ ⨀ ⊕ ⊗ 𝖕 𝖖 𝖗 ⊲ ⊳
∅ ∖ ∁ ↦ ↣ ∩ ∪ ⊆ ⊂ ⊄ ⊊ ⊇ ⊃ ⊅ ⊋ ⊖ ∈ ∉ ∋ ∌ ℕ ℤ ℚ ℝ ℂ ℵ ℶ ℷ ℸ 𝓟
¬ ∨ ∧ ⊕ → ← ⇒ ⇐ ⇔ ∀ ∃ ∄ ∴ ∵ ⊤ ⊥ ⊢ ⊨ ⫤ ⊣ … ⋯ ⋮ ⋰ ⋱
∫ ∬ ∭ ∮ ∯ ∰ ∇ ∆ δ ∂ ℱ ℒ ℓ
𝛢𝛼 𝛣𝛽 𝛤𝛾 𝛥𝛿 𝛦𝜀𝜖 𝛧𝜁 𝛨𝜂 𝛩𝜃𝜗 𝛪𝜄 𝛫𝜅 𝛬𝜆 𝛭𝜇 𝛮𝜈 𝛯𝜉 𝛰𝜊 𝛱𝜋 𝛲𝜌 𝛴𝜎𝜍 𝛵𝜏 𝛶𝜐 𝛷𝜙𝜑 𝛸𝜒 𝛹𝜓 𝛺𝜔