mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   Hardware (https://www.mersenneforum.org/forumdisplay.php?f=9)
-   -   Need help deciding between Athlon II X4 620 and i5 (https://www.mersenneforum.org/showthread.php?t=12668)

garo 2009-11-27 16:15

I think the reason for that is that I have distributed.net running in the background at lower priority so the benchmark ran with no idle cores.

petrw1 2009-11-27 16:16

Another observation:

Prime95 client shows that my i5-750 is a 2 core multi-threaded processor.

garo 2009-11-27 16:17

What OS are you using? You probably have the wrong drivers.

petrw1 2009-11-27 16:22

[QUOTE=garo;197169]What OS are you using? You probably have the wrong drivers.[/QUOTE]

Vista 64 Bit Business.

Control Panel shows me as having 4 CPUs; it is just Prime95 that appears confused.

garo 2009-11-27 16:24

I'm not using the 64-bit version so that may have a problem. Are you using 25.11?

petrw1 2009-11-27 16:34

[QUOTE=garo;197171]I'm not using the 64-bit version so that may have a problem. Are you using 25.11?[/QUOTE]

Hmmm I should use 25.11 but I am using 25.9b4-64Bit.

It still lets me run 4 CPUs so I am not too worried as it is not performance affecting.

CADavis 2009-11-27 22:02

That has been [URL="http://mersenneforum.org/showthread.php?t=11685"]noticed [/URL]by GW already. Easy enough to fix:

from undoc.txt:

[QUOTE]The program automatically computes the number of CPUs, hyperthreading, and speed.
This information is used to calculate how much work to get.
If the program did not correctly figure out your CPU information,
you can override the info in local.txt:
NumCPUs=n
CpuNumHyperthreads=1 or 2
CpuSpeed=s
Where n is the number of physical CPUs or cores, not logical CPUs created by
hyperthreading. Choose 1 for non-hyperthreaded and 2 for hyperthreaded. [/QUOTE]

hj47 2009-11-30 11:12

[quote=sdbardwick;195921]

Quick benchmark comparison:
[code] [B]630[/B] [B]i5-750[/B]
1024K: 20.18 15.87
1280K: 25.16 20.04
1536K: 30.28 24.18
1792K: 36.76 29.31
2048K: 41.49 33.25
2560K: 55.64 44.26
3072K: 66.97 54.09
3584K: 80.70 65.56
4096K: 91.44 74.00
[/quote]

When I see these results I begin to question whether or not the 750 is worth the extra money over the 620/630. In Australia the 620 is about US$110 while the 750 is US$215, so it's comparable to American prices (Aussie dollar FTW).

Even after reading several similar threads I'm still unsure which path to go down. I would be overclocking, but of course it would need to be prime stable for dedicated LL crunching.

Arrrgghhh so confused :cry:

em99010pepe 2009-11-30 11:20

I am with you hj47.
Here in Portugal the Phenom II X4 965(3.4 GHz) and the Core i5 750 2.66 GHz cost the same. The AMD motherboards socket AM3 are cheaper than the socket 1156 ones.
Anyway, I am very inclined for the Core i5 due to its overclock capability.

Carlos

garo 2009-11-30 12:17

Two things that made me chose the i5.
1. Overclocking potential. I have it Prime95-stable at 3800MHz with a good cooler and 0.1V extra.
2. There is minimal slowdown when running 4 LL tests in parallel. I would really like to see iteration times for the 620/630 when running four LLs.

hj47 2009-12-01 10:38

[QUOTE=garo;197365]Two things that made me chose the i5.
1. Overclocking potential.[B] I have it Prime95-stable at 3800MHz with a good cooler and 0.1V extra.[/B]
2. There is minimal slowdown when running 4 LL tests in parallel. I would really like to see iteration times for the 620/630 when running four LLs.[/QUOTE]

Wow :shock:

That definitely makes the i5 look like the better choice. What kind of % speed increase does this kind of GHz help for LL's?

I'm just concerned that at frequencies that high you may corrupt more LL tests than usual?


All times are UTC. The time now is 23:26.

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