I7 920 isn't performing to its potential

[caplus12000]

I have an i7 920 D0 stepping and ive been getting low cpu scores in various benchmarks like passmark and cinebench. i looked up other peoples scores with the same processor at stock and overclocked and theirs is much higher. I overclocked my i7 920 to 4GHz and in geekbench, its floating point score was only 11,000. Other's cpus were nearing the 20k mark at 4GHz. why is this?

 

[clutchc]

Is your system bogged down with resource wasting crap and registry errors? Might try running a clean up program like CCleaner.
http://www.piriform.com/ccleaner

Pretty amazing getting a 2.66GHZ/2.93GHZ processor up to 4GHZ.

 

[caplus12000]

No, I make sure only the benchmark is running. I anyway have 6 gigs of DDR3 1600MHz ram and it barely goes over 2gigs of usage. I've tried CCleaner but no change. I moved from a Q9550 @ 4ghz and that scored about the same as my i7 at 4GHz!!! i thought there was a difference, at least 20% clock for clock???

 

[PassMark]

Check for over heating.

Check the CPU load during the benchmark to make sure all CPUs are fully loaded.

Check that hyperthreading is turned on. It might not be if this was an upgrade.

 

[caplus12000]

Cpu maxes out about 65 degrees after 20mins of prime 95 ( no errors ). all the cores are fully maxed when benchmarking the cpu. Hyperthreading is turned on.

 

[clutchc]

If your low comparative scores are scores showing up in one of Futuremark's ratings, I would take the comparison with the proverbial grain of salt. I have had the same thing happen; both too low and too high. I compare between rigs of my own, but I don't base too much credence on comparison with others' ratings.

 

[mcnumpty23]

have you set power mode to high performance in control panel?

 

[skit75]

I find real world testing to be much more satisfying. I've had my i7 920 @ 3.8GHz w/ a Noctua HSF for a couple years now and I still consistently load gaming maps and other CPU intensive tasks way ahead of my peers. I finish first in my number crunching.... i'd say more than 95% of the time.
The question I ask myself lately is now that I finish first in loading and still have to wait for other gaming competitors to load, was the extra money I spent on my rig worth it?
Who am I kidding, of course it was =)

 

[clutchc]

I find real world testing to be much more satisfying. I've had my i7 920 @ 3.8GHz w/ a Noctua HSF for a couple years now and I still consistently load gaming maps and other CPU intensive tasks way ahead of my peers. I finish first in my number crunching.... i'd say more than 95% of the time.
The question I ask myself lately is now that I finish first in loading and still have to wait for other gaming competitors to load, was the extra money I spent on my rig worth it?
Who am I kidding, of course it was =)

One reason you may be loading maps faster than your peers, may be because you have a faster LAN adapter and/or internet connection. That's usually why some are ready before others.

 

[caplus12000]

In terms of real world performance, it is definitely fast and its a noticeable difference from my Q9550 with DDR2. BF3 is a lot smoother now and I can now multitask better. I have set the power management setting to high performance in control panel but still low scores. What if I just have a bad D0 chip??? my mobo by the way is an X58-UD5 rev 1.0 with the latest bios. I bought these second hand for 300 including 6gigs of DDR3 1600MHz of triple channel ram.

 

[skit75]

One reason you may be loading maps faster than your peers, may be because you have a faster LAN adapter and/or internet connection. That's usually why some are ready before others.

The loading I was referring to is much more dependent on CPU cycles than on my connection speed which actually is not the greatest to speak of. I was able to test this because I just moved also... going from 12-15mb connection to a more relaxed approx. 8 mb connection =/

More than half of my peers have faster connections than I do, now. The example I am using, by the way, is League of Legends and I am not saying you are wrong, but CPU cycles have the most effect in this example by my testing. My setup is not much different than the original posters. I am using a 1st gen Core i7 2.6GHz CPU @ 3.8GHz with 6GB of Corsair Dominator Tri-channel series DDR3 and a single GTX 560 ti on an Asus P6T. I remember testing it using assorted benchmarking wares, post build, and not being much impressed by what I was seeing either, results wise.

Back on topic.... As with any i7 OC, did you make sure to turn off the "speed boost" stuff in the BIOS? Some BIOS call it by different names, but you don't want your CPU flexing more cycles out of an already OC'd setting. You may just be getting throttled for it being enabled.