Van Smith's COSBI benchmark in action - UPDATE: Download COSBI!

[Whitedog]

Is simple how he fudged the scores on the Dell system
His code searches for system make

Sub ScrewDell()
...If system is Dell Then
......results / 2
...Else
......results *2
...End If
End Sub


LOL😀

 

[formulav8]

I think benchmarks should have NO optimizations for MMX, 3dNow, SSE, or SSE2 instructions. It should measure the true performance of the system out of the box. That would make for a fair comparison. But thats just my opinion.



Jason

 

[acejj26]

any possibility that the passively cooled P4 is severely clock throttled??

 

[dexvx]

Clock Throttled? These arent laptops.

Passively Cooled? I can assure you that no Dell P4 system is passively cooled. There is always an exhaust fan somewhere near the CPU. If Precision workstations were clock throttled, then there would be a lot of unhappy companys out there waiting for their data to process.

On a side note...

anyone have any K6 2/3's or lower clocked Athlon/Durons (well under the 1Ghz mark) that might want to give us their scores and see how they compare against the 2Ghz Northwoods?

Enough talk about fairness in benchmarks... I wanna see some fully unabridged source code!

 

[aswedc]

COSBI threatens quite a number of people. Intel has invested a great deal in controlling benchmarks, and COSBI will be completely out of their grasp. The people behind Intel-friendly commercial benchmarks are also not terribly happy to see COSBI either. The Intel supporters (operatives?) have tried to already paint conspiratorial colors on COSBI even before the project is released. A few detractors display almost a pathological antipathy to anything we do. -Van

The Athlon XP 2000+ must really be twice as fast as the P4 2.53ghz! Quake 3 is really a conspiracy with Intel!

I think it'd probably be better if we just ignored Van instead of trying to figure out his benchmark results...

 

[acejj26]

the P4 is passively cooled on dell machines....there is only a large heatsink on the processor...i doubt it is throttled, but it remains a possibility

maybe van should post the source code for his benchmark so people can verify if it is legit or not

 

[SexyK]

Originally posted by: acejj26
the P4 is passively cooled on dell machines....there is only a large heatsink on the processor...i doubt it is throttled, but it remains a possibility

maybe van should post the source code for his benchmark so people can verify if it is legit or not

I've read reports of P4's running at 70+c and not throttling. In fact, the only time i've ever seen a chip throttle is in Tom's video where he completely removes the heatsink. If a major OEM like dell can't produce a system with temps below 70c, something is majorly wrong. I'd be willing to bet a heafty sum that no one on these forums could force a Dell system to throttle straight out of the box, no matter what they ran on it.

Kramer

 

[Pabster]

aswedc wrote:

"I think it'd probably be better if we just ignored Van instead of trying to figure out his benchmark results..."

Ditto.

 

[BDawg]

All this talk about COSBI and not one Jello Pudding Pop joke. 😀

 

[dexvx]

No P4's are NOT passively cooled. I have been through dozens of Dell P4's. They have a HUGE heatsink on the processor and no fan ON the heatsink, but there are fan exhausts located in VERY close proximity to the heatsink. If you disable that fan, you can bet that your computer will auto-shut down due to the heat. I define passively cooled as a proc with a heatsink thats sitting in the middle of nowhere (like my Dell pentium system).

 

[glugglug]

ROFL.

Perhaps the benchmark is CPU cycles taken, hence higher MHz rating = worse results since the rest of the system generally isn't proportionally scaled?

Not gonna run this thing until I see source so I can compile it myself. Otherwise I think you're asking for a trojan/virus.