Post your Cpumark99 scores

[OBLAMA2009]

418 e7500/g31, its just a business machine

 

[n7]

i7 860 @ 3.9 GHz - 601

Q9650 @ 4 GHz - 577

E5300 @ 4 GHz - 568

I suspect this benchmark favors RAM speed a lot.

 

[Rubycon]

i7 860 @ 3.9 GHz - 601

I suspect this benchmark favors RAM speed a lot.

Yes it probably does. I use a fair amount of apps that do respond well to ram speed/timings/latency so I put lots of emphasis on them for better throughput. It's one area you can push to the limits without worry of degradation or breaking something outright. Well as long as you don't get carried away. (the taller modules tend to break easier and kill the slots on the mainboard if a large tool like a crescent wrench or crowbar happen to fall on them! )

 

[BD231]

a lowly 527 for a 965BE @ 3.8/2.8

Your CPU north bridge is running at 2.8ghz?

What motherboard are you using?

 

[konakona]

biostar TA790GX A3+

cpu-nb is running at a bit high 1.44v, but hasn't given me any problems yet.

 

[Miramonti]

I've used many 1366 CPUs including X model Xeons. (!)
Overclocking is a lottery, plain and simple.
Case in point - I have two W3520s that overclocked and beat out three i7 975s by quite a bit! I had W3540s that were DOGS!

The guy posting his air o/c (more than likely) is "benchmark stable". I just don't see that speed and voltage working on commercially available air cooling. Unless he lives in the arctic region with the box outside. Or maybe he's a butcher and the workplace allows his pc in the walk-in?!

That voltage and frequency when fully stressed with a program such as Linpack is going to generate crazy heat. High end air won't work. Water cooling will if a large radiator is used. Even then I would expect mid 70s temperatures.

I'm BCLK limited with this chip because of the memory configuration - I use 12GB at 7-7-7-18-1T 1693 / 1066 strap. This is on a 4X Classified X58 (762) Mobo. Cooling is achieved with a Venomous-X with a pair of Gentle Typhoon 1800RPM fans in push pull. Idle temps around 35C, load temps in the lower 60s - peak load temps lower 80s.

My definition of stability and endurance are different than most people's. If the box cannot pass the harshest loading test while getting thrown on the floor while at sea along with two gorillas playing volleyball with the case - then it's of no use to me! Ok that may be slightly exaggerated. I will say those Gentle Typhoon fans are quite durable and really can take a beating without dying!

That burst my bubble a bit but thanks for the assessment.

I agree with the lottery part...I don't think I drew the best of the crop with my e8600 but am making the best of it. Your threshold for stability requirements are commendable, altho my needs for stability are vastly different than yours...all I need is for it to make it thru the work day until its rebooted, and then start all over again fresh for tomorrow. I rarely ever have a crash, but rebooting definitely helps.

I knew from the outset there might be challenges and I didn't win the lottery with this one...my cpu idles at 60c and full throttle reads 90c. I suspected it was probably a cpu diode(?) issue with the temperature reading (after seating and reseating heatsink, lapping, and my case can be a turbine if turned up 100% etc.) Since I already lapped it as soon as I saw high temps, I eventually said screw it, I'll just buy another cpu if I'm wrong, I need the o/c'd speed. That was 18 months ago, affirming my conclusion imo.

But overall I'm closer to an upgrade than I thought and hope to 'play the lottery' soon. This time I won't lap the cpu so quickly so I can return it in case its an obvious loser.

Thanks for your feedback and hardware configuration info as well.

 

[Miramonti]

i7 860 @ 3.9 GHz - 601

I suspect this benchmark favors RAM speed a lot.

Iirc, I do recall better numbers with lower latency settings.

 

[Jovec]

a lowly 527 for a 965BE @ 3.8/2.8

530 - 955Be at 3840, 2640 CPU-NB, 1920 HT, and 800MHz (3.33x240) DDR2

Going to piggy back on this thread, but what benchmarks would you recommend that might stress the HT link? I'd like to test between 1920 and 2160. Also, with 8GB DDR2 1066 I know my ram isn't the fastest, but any links/guides for ram settings? All I know is at 960 (240x4) I won't pass p95 or linpack but I will at 240x3.33. Possibly more NB voltage? Or adjust timings?

 

[ljtatej]

Got 554 on my rig below..sadly this processor would see a lot more with a better board.

 

[konakona]

tightening the ram timings from 9-9-9-24-36 to 6-6-6-18-26 still retained the same score @ 527.
time to play around with ref clock and multi to give it a slight boost

 

[loafbred]

Q9550 @ 3.6: 520.

 

[zoic]

547 with i7 920 3,5GHz HT Off (with wine on ubuntu 9.10).

I'll test on Win7 when I bother to reboot.

548 in Win7.

 

[loafbred]

Q6600 @ stock 2.4Ghz:
No affinity: 245
Affinity to two processors: 261
Affinity to one processor: 339
Affinity to one processor and priority set to real-time: 340

Moral of the story, set your affinities for your single-threaded applications.

What O/S are you using? With Vista 64, my q9550 @3.6 scores 521 with affinity set to core 3, but any other combination of cores, and any other single core, scores consistent 520.

 

[Cogman]

What O/S are you using? With Vista 64, my q9550 @3.6 scores 521 with affinity set to core 3, but any other combination of cores, and any other single core, scores consistent 520.

Weird, Vista 64bit Ultimate.

Actually, I think I know where the difference is coming in. I have speedstep (or whatever intels power saving thing is) turned on. When no priority or affinity is set, I betcha the CPU is ramping down on me, distorting my score.

 

[n7]

tightening the ram timings from 9-9-9-24-36 to 6-6-6-18-26 still retained the same score @ 527.
time to play around with ref clock and multi to give it a slight boost

I did some testing w/ RAM @ DDR3-1560 & 7-8-7, 7-7-7, 7-8-8, & 7-10-10, & let's just say this benchmark is extremely unreliable IMHO.

I was getting better scores @ 7-10-10 than 7-8-7, which is silliness.

 

[loafbred]

Weird, Vista 64bit Ultimate.

Actually, I think I know where the difference is coming in. I have speedstep (or whatever intels power saving thing is) turned on. When no priority or affinity is set, I betcha the CPU is ramping down on me, distorting my score.

It must not be caused by that, because mine is turned on too. At idle it runs at multiplier 6, under load at 8.5.

 

[Cogman]

It must not be caused by that, because mine is turned on too. At idle it runs at multiplier 6, under load at 8.5.

I'm using a pretty aggressive power profile in vista, you have to change that specifically.

 

[Miramonti]

Bump for some updates.

I'd love to see some Sandy Bridge scores!

 

[Puppies04]

613

I5 2500k @ 4.0ghz (dropped it from 4.5 for some benching)

 

[Chiropteran]

490 first run. stock speed Phenom II 1090t
Other runs: 484, 480, 472, 464, 452 (literally every time I run it it is slower than the previous run)

453, 430 Same system, run from inside a Windows XP VM

 

[Kenmitch]

2700K

Running at 4.0ghz gets me 617
Running at 4.2ghz gets me 648
Running at 4.8ghz gets me 737
Running at 5.0ghz gets me 770
Running at 5.1ghz gets me 785
Running at 5.2ghz gets me 801

Guess it's a very light hitting app for todays CPU's as my vcore at 4.0ghz was 1.104v's under load

 

[Joseph F]

508
Core i5 2500k 3.3GHz, turbo off.

 

[ljtatej]

682 with current rig @4.4Ghz. Can't remember what I was running on the first go round.

 

[nemesismk2]

I get 418 with my amd athlon ii x3 455 running stock at 3.3ghz

 

[Steelski]

Phenom I at 2.6Ghz
score : 323