Conroe E6300ES review =)

[Gorod]

Thought it might be interesting to compare low-end Conroe to prev. generation CPU's

Opteron 170
Cooler Master Hyper 6
DFI LP UT nF4 Xpert
2x512Mb , Mushkin Redlines XP3500 , Winbond UTT
X1900XTX 750/900
250Gb WD2500JS-00MHB1 , S-ATA 16Mb
Win XP Pro SP2 , Cat 6.5

Vs.

T2600ES
Tt Big Typhoon
Aopen i975Xa-YDG unmoded
2x512Mb , Corsair Value Select VS1GBKIT667D2
X1900XTX 750/900
250Gb WD2500JS-00MHB1 , S-ATA 16Mb
Win XP Pro SP2 , Cat 6.5

Vs.

E6300ES
Tt Big Typhoon
Intel D975XBX "Bad Axe" Rev.304
2x512Mb , Corsair Value Select VS1GBKIT667D2
X1900XTX 750/900
250Gb WD2500JS-00MHB1 , S-ATA 16Mb
Win XP Pro SP2 , Cat 6.5

• Opteron 170 @ 3.0GHz , 300x10.0 , HTT x3 , 250MHz @ 2.0-2-2-5-1T
• Opteron 170 @ 2.8GHz , 280x10.0 , HTT x4 , 255MHz @ 2.0-2-2-5-1T
• T2600ES @ 3.0GHz , 231x13.0 , 385MHz @ 4.0-4-4-12
• T2600ES @ 2.8GHz , 215x13.0 , 359MHz @ 4.0-4-4-12
• T2600ES @ 2.6GHz , 200x13.0 , 333MHz @ 4.0-4-4-12
• E6300ES @ 2.67GHz , 381x7.0 , 381MHz @ 4.0-4-5-12
• E6300ES @ 3.0GHz , 429x7.0 , 429MHz @ 4.0-4-5-12
  

Bad Boys :







Games :











DVD > DiVX Encoding with Auto Gordian Knot 2.27 :

Movie : Night_Watch ( 1h54m )
Movie size on HDD : 4.06Gb
DiVX preset file size : 700Mb





Synthetic Benchmarks :


























































Links to custom test apps :

TesteR106
Stream
FactorLite
Nuclearus

[[XC]melymel]

Nice work I bet doing all those benches took a while so thanks What happened to the 170 @ 3.0ghz after the first 5 benchmarks it isn't shown? Anyway the opty kept up better than I thought it would but then this is the bottom end for conroe .

[DTU_XaVier]

Nice one, and so comprihencive that even some larger hardware-sites could learn something from it
Though, I have to point out (as it was to me a few days ago), that technically, every chip below E6600 is currently not Conroe, but Allensdale, with a halved L2 Cache

Best Regards

[[XC]melymel]

Though, I have to point out (as it was to me a few days ago), that technically, every chip below E6600 is currently not Conroe, but Allensdale, with a halved L2 Cache

Best Regards

No point in being pedantic as it's the same awesome arc as the conroe and they're all c2d.

[Gorod]

melymel2789

I bet doing all those benches took a while

Heh ...

What happened to the 170 @ 3.0ghz after the first 5 benchmarks it isn't shown?

It's in "progress" I'l try to add 3.0GHz for Opty in a few days when i get time for that . I'l try to bench all that on my Venice 3000+ @ 2.8GHz and CedarMill661 @ ~ 4.8-5.0GHz also . I'l try to get more games too

[Gorod]

DTU_XaVier

Though, I have to point out (as it was to me a few days ago), that technically, every chip below E6600 is currently not Conroe, but Allensdale, with a halved L2 Cache

My bad ... it's 7.30 in the morning here , i think i stayed up too late today lol
Best regards guys

[DTU_XaVier]

No point in being pedantic as it's the same awesome arc as the conroe and they're all c2d.

Awesome indeed
Why can't it be august?

Gorod, what's the max. stable clock on that E6300??

Best Regards

[Gorod]

Awesome indeed
Why can't it be august?

Gorod, what's the max. stable clock on that E6300??

Best Regards

I realy dont know , need a better mobo to find out With MCH v-moded "BAD AXE" i been able to game/prime all day long @ 3.0GHz , 430MHz FSB with Default Vcore for CPU

Max benchable was 3233MHz (462FSB with crapy 1333MHz FSB strap ) (default Vcore too ) , i guess i will never find out what that CPU is capable off unless there are any mobos with 500+ fsb coming out any time soon

[DTU_XaVier]

I realy dont know , need a better mobo to find out With MCH v-moded "BAD AXE" i been able to game/prime all day long @ 3.0GHz , 430MHz FSB with Default Vcore for CPU

Max benchable is 3233MHz (462FSB with crapy 1333MHz FSB strap ) (default Vcore too ) , i gues i will never find out what that CPU is capable off unless there are any mobos with 500+ fsb coming out any time soon

75% oc at default voltage...
Though, the low multi is a drag

Best Regards

[Shadowmage]

Anyway the opty kept up better than I thought it would but then this is the bottom end for conroe .

Don't forget this is at 3GHz, and cache size doesn't really matter that much (minimal increases).

[Cooper]

Great review. So much work done

I also added few links at the bottom of 1st post to some custom applications Gorod used in this comparison. Perhaps some of you might be willing to bench something new
I guess Gorod also might add few more links which I haven`t found myself.

PS Just click on the link how it`s shown on picture below

[^don.k's^]

Doesn't 6300/6400 ES have 4mb of L2 ? At least the first samples had 4mb L2...

[DTU_XaVier]

Doesn't 6300/6400 ES have 4mb of L2 ? At least the first samples had 4mb L2...

No, they do not... They have only half the L2 cache of E6600 and above, and are codenamed as "Allensdale", though, other than that, they're identical to 4mb chips

Best Regards

[ineedaname]

Like what melymel2789 said the opty did quite well compared to the lower end conroe's. Conroe's are sweet but i already have an opty so it makes me feel not as bad lol.

[hicookie]

Awesome professional review!!!

Nice chip,default vcore oc @ 75% lol

[vengance_01]

Great review, it shows the Opteron can hold it owns clock for clock, but because conroe scales so well, the opterons will have a hard time keeping up

[bachus_anonym]

Damn, Gorod! Your latest comparisons are awesome! Thanks for sharing, man

[K404]

Thanks for putting the work in

BUT

As you said, Gorod...the results for the Opto170 at 3GHz REALLY need to be added, else the review doesnt really hold up. The E63 and Opto170 arent compared at the same speeds for a lot of tests, and the E63 is the one to have the clock speed advantage as well.

Once those results are added...I reckon that`ll be a damn good review

[Mako88]

Once those results are added...I reckon that`ll be a damn good review

It's a damn good review already.

And those saying the Opteron somehow holds a candle to those results are dilusional at best. Clock for clock even the 2MB Conroe seems a generation ahead as it should. Now we wait for K8L to answer (which according to DigiTimes breaking news today isn't due until 2008 now...) and so on and so forth.

Love all this competition, in the end the consumer wins.

[K404]

I really dont want to knock the reviewer, but who games at 640.480? Yes, Conroe has a good lead at lower res, but at 1280.1024 and above...the lead will be much smaller.

[Mako88]

I really dont want to knock the reviewer, but who games at 640.480? Yes, Conroe has a good lead at lower res, but at 1280.1024 and above...the lead will be much smaller.

As discussed many times on this site and others, you test games at the lowest resolution possible when probing a CPU's limits because it isolates the CPU as the limiting factor as opposed to the video card GPU limiting speeds.

It's not about "real world" gaming frame rates in tests like these, the games are actually more like a synthetic benchmark for these purposes.

As far as the results being "much closer" at higher resolutions, that's correct, but not due to any sort of misplaced fanboy hopes regarding the Opteron holding a candle to Conroe. They're closer as the resolution climbs because the video cards have to endure more load. Eventually, say at 2560x1600, there is actually zero difference in the frame rate scores between any two CPUs...because the video card is maxed out and incapable of scaling any higher.

[Shadowmage]

In most situations it looks like a 3.2GHz Core Duo Sonoma would do just as well as a 3GHz Allendale.

[K404]

Please dont mention fanboy here, looks like you`re trying to get a rise out of me.

If a game is being tested at 640.480 to show one CPUs better performance against another, at the expense of real-world use...whats the point? Yes...one CPU will be shown to have better performance...but who will game at 640.480 to feel the benefit? I wont drop my resolution to feel better about my purchase

If, at higher resolutions, the FPS difference is only 3 or 4.... is it worth the upgrade? People will be using their computers for high-res gaming, video encoding etc as well as, if not instead of benchmarking, they`ll want to see a benefit for real-world apps. Thats what i`m getting at.

I dont care if CPU X gets whupped or CPU Y costs £400 more. I want a fair comparison and real-world performance...nothing more, nothing less.

Kenny

As discussed many times on this site and others, you test games at the lowest resolution possible when probing a CPU's limits because it isolates the CPU as the limiting factor as opposed to the video card GPU limiting speeds.

It's not about "real world" gaming frame rates in tests like these, the games are actually more like a synthetic benchmark for these purposes.

As far as the results being "much closer" at higher resolutions, that's correct, but not due to any sort of misplaced fanboy hopes regarding the Opteron holding a candle to Conroe. They're closer as the resolution climbs because the video cards have to endure more load. Eventually, say at 2560x1600, there is actually zero difference in the frame rate scores between any two CPUs...because the video card is maxed out and incapable of scaling any higher.

[TarTheDark]

With the speed of GPU advancement and single-PCB multi-GPU cards in SLI or CF you can SOON expect to see a big difference at higher resolutions as well.

So yes, the game benches are pretty relevant. Don't want your CPU to hold you back in the end, do you?

[Mako88]

If a game is being tested at 640.480 to show one CPUs better performance against another, at the expense of real-world use...whats the point?

What's the point of Sandra? What's the point of SuperPI?

They're synthetic tests designed to illustrate differences in CPU/memory performance without interferance from other peripherals in the system. As I said earlier, that's the same reason you test games at the lowest resolution possible when looking at CPUs...to isolate them from the GPU. Real world gaming benchmarks? No, of course not. These tests serve a different purpose, which is clear.

Yes...one CPU will be shown to have better performance...but who will game at 640.480 to feel the benefit? I wont drop my resolution to feel better about my purchase

Same as my response abive. Again, in this particular review the games are being used in a way that best isolates the processor in the system. Think of them as you would CPUMark, or SuperPI. That's the point.

If, at higher resolutions, the FPS difference is only 3 or 4.... is it worth the upgrade? People will be using their computers for high-res gaming, video encoding etc as well as, if not instead of benchmarking, they`ll want to see a benefit for real-world apps. Thats what i`m getting at.

Right, and this review wasn't meant to illustrate whether it's "worth" the upgrade only to gamers. Again, the video card GPU will always be the limiting factor in game frame rate performance above resolutions of 1280x1024x4AA. Always. So to judge a CPU from that aspect is foolish, as the fps difference is so small it's irrelevant. Luckily from the real-world apps shown above, the Conroe destroys the Opteron, clock for clock. So this review game you the exact information you're claiming you sought. Hooray!

I dont care if CPU X gets whupped or CPU Y costs £400 more. I want a fair comparison and real-world performance...nothing more, nothing less. .

And that's exactly what you got. Was it tailored specifically for the gamer? Nope. As we've said, games only vary by 3-4fps at higher resolutions because they're limited by the GPU unless you're part of the 1% running an SLI/Crossfire solution. This review isn't intended for that audience, you'll have to look elsewhere for that.


Glad we got that sorted out, this review turned out to be perfect for its intended audience. Nice work Gorod, couldn't have done it better myself.