e8600 Slower than e8500???

[El<(')>Maxi]

a Bios update has nothing to do with cpu bench tests.

Sorry but that is 100% wrong. People have been chasing BIOS files in search of performance (and getting it in many cases) for a very long time. One current example has guys who bench Super Pi running non-supported BIOS's to boost performace and it's very effective. Look at the DFI NF4 heydays, Foxconn BlackOps and every other high-end boards out there, performance is usually improved with each BIOS release. There are also a myriad of situations where a BIOS update is issued by system builders to improve system performance, in fact I was looking at updating a server today for this very reason. The updated BIOS included updated Intel(R) Memory Reference Code to revision 1.3.2. Too many examples on this forum alone which prove the point.

 

[wingman99]

Sorry but that is 100% wrong. People have been chasing BIOS files in search of performance (and getting it in many cases) for a very long time. One current example has guys who bench Super Pi running non-supported BIOS's to boost performace and it's very effective. Look at the DFI NF4 heydays, Foxconn BlackOps and every other high-end boards out there, performance is usually improved with each BIOS release. There are also a myriad of situations where a BIOS update is issued by system builders to improve system performance, in fact I was looking at updating a server today for this very reason. The updated BIOS included updated Intel(R) Memory Reference Code to revision 1.3.2. Too many examples on this forum alone which prove the point.

Your are 100% Wrong, i said BIOS UPDATE, is from the manufacture, find me a bios up date and score where it made a difference in benching on a home PC.

It should be easy for you to do, just look for performance increase fix on the updated bios listing changes LOL.

 

[gigabit]

Your are 100% Wrong, i said BIOS UPDATE, is from the manufacture, find me a bios up date where it made a difference in benching score on a home PC.

The bios can make a big difference in benching.The last bios(P06) for the 780i gave me much weaker performance from my ram in Everest just to name one benchmark.

 

[wingman99]

The bios can make a big difference in benching.The last bios(P06) for the 780i gave me much weaker performance from my ram in Everest just to name one benchmark.

This post is about CPU performance not ram or north bridge or south bridge or motherboard performance.

 

[Gautam]

Well so far I can tell you that in the Canyon Flight test in 3DMark06, they are absolutely dead even. I had numbers I knew for that particular test which is why I quoted it. I'm not sure if I'll have time to do comprehensive testing tonight, but based on that, I'm expecting both to be completely equal.

And wingman, for the most part you're correct but there are some strange BIOS issues involving 45nm chips that cause them to be all over the place in superpi depending on which you have.

 

[theELVISCERATOR]

Your results were done back to back on the same install? I guess I didn't read carefully enough but it sounded like you pulled the E8500, reinstalled Windows, then put the E8600 in.

Nonetheless you piqued my interest. I'll try both chips in a P5E64 Evolution tonight...should be somewhat comparable.

Using identical drivers, I could not match my 2nd vantage run..ran at 4194 at 4500 mhz! My cpu scores are faster about the right amount..

the in game FPS is just lower and I am about 350 pts down from my best with my 3110, and get this with higher card clocks also!

I had flashed to the 17 bios though.

I have flashed back to 16 but havent had the time to bench.


I did test at 500x9.0 for 4500 and no gain over 450x10...strange... mem settings about the same..


I wonder if this chip is limited by p35 performance?

 

[wingman99]

And wingman, for the most part you're correct but there are some strange BIOS issues involving 45nm chips that cause them to be all over the place in superpi depending on which you have.

Well i agree there is latency timing they can play with in Bios adjustment and also different boards with a different setup, but it's vary small.

Could you test with PC WiZard 2008 bench like this below E8400 stock 3.0Ghz with both CPU's or just post one

 

[gigabit]

This post is about CPU performance not ram or north bridge or south bridge or motherboard performance.

I know what the thread is about i can read
I was making a point.But what ever.I dont argue with people that are always right

 

[wingman99]

I wonder if this chip is limited by p35 performance?

Your defiantly reaching, that is like saying there is ghosts in my chip set.

 

[Evilsizer]

Using identical drivers, I could not match my 2nd vantage run..ran at 4194 at 4500 mhz! My cpu scores are faster about the right amount..

the in game FPS is just lower and I am about 350 pts down from my best with my 3110, and get this with higher card clocks also!

I had flashed to the 17 bios though.

I have flashed back to 16 but havent had the time to bench.


I did test at 500x9.0 for 4500 and no gain over 450x10...strange... mem settings about the same..

I wonder if this chip is limited by p35 performance?

stick with bios 16, bios 17 is more for 45nm quad cores then 45nm duallies. 500x9 wouldnt be any gain over 450x10, because the NB timings are looser when you run 500mhz fsb vs even 499mhz. im sure Guatam could atest to that, the number differences your seeing are truely odd. P35 is not the problem here, must be something else going on... sorry i have skimed over some of the thread but have you by anychance installed difference video card drivers at all?

gigabit,
dont worry about wingman...

 

[saxile]

Hey Stilletto, this might be a stupid question, but then again it could be part of the problem. Have you tried benchmarking the L1/L2 caches to see if there is any differences between your two chips? Could be the e8600 has slightly slower cache than the e8500, sounds wierd, but not much in the world of computers always makes sense.

 

[wingman99]

I know what the thread is about i can read
I was making a point.But what ever.I dont argue with people that are always right

I'm not always right, however you have to weed out things that don't effect cpu performance with the same chip architecture and same Bios. When you drop in a new CPU the Bios has specs for that chip or it wont POST simple.

Yes i know what your Point is but those Bios bench marks don't change the bench marks from cpu performance any significant amount with 45nm 775 platform has a abundance of memory bandwidth and FSB speed more than any Intel cpu could ever use.

 

[IInvader]

I've noticed the L2 cache on the E8600's seems to be registering a higher latency by many posters, compared to the L2 cache on E8400 and E8500 C0 processors, at similar FSB and MHz.

I know the L2 cache is much slower on the Wolfdales as compared to the Conroes and Allendale Core2Duo, but because the cache is bigger on the Wolfdales, there is little performance hit. But the E8600 has the same size L2 cache. On my E8400 C0, I'm getting a 4.0ns L2 cache latency at 500 FSB x 8 = 4,000 MHz, 2G Ram 1:1 500 MHz 5-5-5-15, Performance Level 9.

 

[Evilsizer]

I've noticed the L2 cache on the E8600's seems to be registering a higher latency by many posters, compared to the L2 cache on E8400 and E8500 C0 processors, at similar FSB and MHz.

I know the L2 cache is much slower on the Wolfdales as compared to the Conroes and Allendale Core2Duo, but because the cache is bigger on the Wolfdales, there is little performance hit. But the E8600 has the same size L2 cache. On my E8400 C0, I'm getting a 4.0ns L2 cache latency at 500 FSB x 8 = 4,000 MHz, 2G Ram 1:1 500 MHz 5-5-5-15, Performance Level 9.

im wondering about this, seems really odd that new E0's L2 would have a higher latency. the L2 is still running at CPU speed so the only thing i can think of going on is if they use "timings" for L2 cache. i dont know if that is possible, i cant see any other way to "slow down" L2.

 

[White_Pawn]

the moment you guys started this thread. I automatically thought that perhaps the are higher cycles in the new E0 steppings. That might be the reason why they can overclock higher. I'm guessing it has something to do with the L2 cycles.

I wouldn't be surprised either, as if you guys read the nehalem articles. They have increased the cycle latency in the cache for nehalem. They might be playing around with it here, in the E0 chips.

 

[White_Pawn]

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3382&p=9

Nehalem, like AMD’s Phenom, features a 3-level cache hierarchy. There’s a 64KB L1 cache (32KB I + 32KB D), a 256KB L2 cache (per core, unshared) and up to an 8MB L3 cache (shared among all cores).

The L1 cache is the same size as what we have in Penryn, but it’s actually slower (4 cycles vs. 3 cycles). Intel slowed down the L1 cache as it was gating clock speed, especially as the chip grew in size and complexity. Intel estimated a 2 - 3% performance hit due to the higher latency L1 cache in Nehalem.

 

[Evilsizer]

http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=3382&p=9

right i remember reading that, i wonder then if intel implemented the same then on E0's.

 

[wingman99]

im wondering about this, seems really odd that new E0's L2 would have a higher latency. the L2 is still running at CPU speed so the only thing i can think of going on is if they use "timings" for L2 cache. i dont know if that is possible, i cant see any other way to "slow down" L2.

Unless there using new sleep transistors for more power saving less leakage and there is more latency off and on in the Cache when they power down, that's what I've been thinking.

 

[Stilletto]

So far, the cache speed/latency according to my testing is looking to be slower indeed.....

 

[White_Pawn]

Unless there using new sleep transistors for more power saving less leakage and there is more latency of and on in the Cache when they power down, that's what I've been thinking.

it isn't a new architecture like nehalem. I'm pretty sure E0 is just a test to see how far the cpus can overclock when they loosen the cycles for L1 cache. The 2~3% decrease in performance clock per clock sounds about right. What is MOST important, is to find out if the cpu is fasterin the end, E8600 highest OC vs. older stepping highest OC.