Discussion in 'General Hardware' started by lemonadesoda, Mar 18, 2009.
coming right up, 2600k system at 4.2ghz.
Geekbench score- 12797
I do got to say I like how it does both single core and multi core benching and comparison. And I can confirm it uses all 8 processing threads of my 2600k(watched task manager while benching)
Name ....... | CPU . | Speed .. | MemoryType | GBScore | Integer | FP . | Memory | StreamScore | GeekScore/Ghz | comment
BarbaricSoul | 2600k | 4.20 Ghz | DDR3 1600 | 12797 .... | 13242 |1682 | 7976 .... | 9384 ....... | 3046.9 ........... | 2600k system in system specs
Was that Geekbench version 2.4.x or 2.1.x? http://www.sendspace.com/file/zputtn Please use v2.1.x for this thread. The results are (unfortunately) not compatible.
@all. Please post your result in the correct submission format, so I can copy+paste into the table. Thanks! Example here
I got this.
MetalRacer|i7 3930k|5.2GHz|DDR3 2133|20767|23498|26279|9649|14153|3993|
MetalRacer|i7 3770k|4.8GHz|DDR3 2133|15330|16261|18787|9850|10934|3193|
Aquinus | i7 3820 | 4.62 Ghz | DDR3-2400 (10-12-11-33-1T) | 13131 | 13827 | 15091 | 8923 | 12254 | 2842 | 128.6Mhz bclk
Melvis | Phenom II X4 965 | 3.4 Ghz | DDR3 1333 | 6317 | 7353 | 8291 | 2960 | 2501 | 1858 | Windows 7 64
Here's what I'm getting
chevy350 |2600k | 4.4GHz | DDR3-1866 | 13228 | 13994 | 16661 | 7937 | 9120 | 4017 |
Add it if you want, here is my >$300 laptop with it's 1.3ghz E-300 APU
Melvis | AMD A8-5600k | 3.8 Ghz | DDR3 1333 | 6063 | 6427 | 7646 | 3884 | 3612 | 1596 | Customers Comp
CPU-Z said it only hit 3.8GHz, its meant to do 3.9
I make that 4204 Geekscore / Ghz
Could you kindly rerun that under Geekbech 4.1.x (see OP). Thanks.
Look into reducing TRFC a tad. Might help your ram latency. Help squeak even a little more performance out of your build.
Yes I can re-run with the earlier version, just paid for the unlock a year or so ago and like to keep it updated. As far as the score/ghz I tried following the op but it wasn't very clear as to which score to use since your example shows a score that wasn't listed in your example post lol.....will get the results updated with the 4.1.x later tonight when I get home from work
The Geekscore / Clock speed = Take the main overall result / your actual OC clockspeed
Anyone else browsing this thread... please add your result, esp if you run a Xeon or AMD dual / quad CPU system.
Arctucas | i7 950 | 4273 MHz | DDR3 1600 | 10986 | 12428 | 13436 | 6387 | 6571 | 2571 | http://browser.primatelabs.com/geekbench2/1369754
Well I would re-run and re-post but get this from op link
Download not available
The file of the above link no longer exists. This could be for several reasons:
The file was deleted by the user who uploaded it.
The file contained illegal contents and was deleted from our servers by our Anti-Abuse team.
The link is incorrect.
The server is busy and can not process the request.
File not found. (e029a7af)
Good thing to know. I usually leave TRFC on auto, I figured the motherboard/memory has a better idea for how long and often it should refresh. Do you know much about secondary memory timings? I usually leave them alone. It was my impression that improvements are minimal on most secondary timings unless they're way off.
I really need to do some thorough testing with my rig to find out exactly what limits on what components are there because there is a lot to configure on SB-E. I can get a decent overclock out of it but I feel like it wants to go faster but can't. Maybe it's just me wanting it to go faster and it doesn't. (Not to say that it is slow, it certainly is not.)
Here is my copy of v2.1.4. Not sure how long the upload will stay there. Perhaps this could be transferred to TPU/downloads... I'll PM Mr W1zzard
My previous post was with 2.1.10.
I ran 2.1.4 and got:
Arctucas | i7 950 | 4273 MHz | DDR3 1600 | 11207 | 12518 | 13688 | 6496 | 7366 | 2622 | http://browser.primatelabs.com/geekbench2/1373607
A small increase.
I think we are suppose to be running the 32bit test.
I'll upload the re-run. But in general don't worry about which 2.1.x version you use... they are pretty close results and well within a reasonable tolerance of +/-3%. The problem is when someone uses 2.4.x, or 64bit, where the results can be quite different. To put it all into perspective, the idea of the table isn't about small %, but seeing the order of magnitude differences between old and new, and atom vs. xeon. etc.
Yup I just noticed that as well, grabbed the 2.1.4 that was upped so I'll get it re-run....and in 32-bit as well lol. I paid the small fee to run the 64-bit and was just out of habit to run it on 64 instead of 32
Edited results in post #111 big change over the 64-bit results lol
Separate names with a comma.