86 Processors compared over at Tomshardware

[pauldun170]

http://www.tomshardware.com/reviews/cpu-performance-comparison,3370.html

 

[zebrax2]

I'm slightly disappointed with that review when i saw that it only covers CPUs up to 3 generations old.

 

[exar333]

I'm slightly disappointed with that review when i saw that it only covers CPUs up to 3 generations old.

They reviewed 86 CPUs. They have to draw the line somewhere...

 

[zebrax2]

They reviewed 86 CPUs. They have to draw the line somewhere...

I just wished they tested older processors rather than testing multitudes of processor whose only difference really is a speed bump of 100mhz

 

[Soulkeeper]

This makes me glad I chose llano

 

[Gigantopithecus]

Useful, but disappointing they left out SB i3s as well as IB i3s and PDCs.

 

[Avalon]

Too many models where the only difference was clock speed, IMO. Made me want to stop reading after the first chart.

 

[Ketchup]

Are they just trying to make themselves look impressive? As posted above, a bunch of really close CPU's, and they don't even go back that far. Not wasting my time reading all that. Heck, Anandtech goes back that far on just about EVERY CPU review, don't they?

 

[Idontcare]

This makes me glad I chose llano

Amazing when you look at some of those charts in that it shows just how much of a step backwards AMD took with the bulldozer microarchitecture.

If only they had kept with the optimization and shrinks of the thuban lineage

As an FX-8350 owner, it is not surprising all to see piledriver dominate the following chart though.

Who would have guessed that making the 32nm equivalent of Prescott would be history repeating itself? Somehow Hector didn't see that coming, wonders never cease...

 

[mrmt]

Who would have guessed that making the 32nm equivalent of Prescott would be history repeating itself? Somehow Hector didn't see that coming, wonders never cease...

Hector didn't. Dirk did, and still pressed on.

 

[Tempered81]

I just wished they tested older processors rather than testing multitudes of processor whose only difference really is a speed bump of 100mhz

http://www.hardware.fr/articles/778-10/crysis.html

 

[AtenRa]

22nm 3770K uses more power than 32nm 2700K ?? really ???

They could measure power consumption in a highly MT application like HandBrake, Super Pi is such a fail for power consumption measurements for today's microprocessors.

 

[Homeles]

Hector didn't. Dirk did, and still pressed on.

Bulldozer was announced during Hector's leadership.

http://www.bit-tech.net/news/hardware/2007/07/28/amd_goes_modular/1

Dirk probably pushed on in light of the bigger picture. Brilliant concept, terrible execution.

It's unfortunate that the first iteration was such a disaster, and the second iteration is still bad, but I imagine things being much better for AMD once they make it to Steamroller. The changes that are being put in address every critical flaw of the architecture.

The question is whether or not that design will ever see the light of day.

 

[fixbsod]

bah 1333 RAM in SB?

 

[Idontcare]

22nm 3770K uses more power than 32nm 2700K ?? really ???

They could measure power consumption in a highly MT application like HandBrake, Super Pi is such a fail for power consumption measurements for today's microprocessors.

That does seem suspect. Even if they didn't want to use handbrake, how about a free MT bench like Cinebench?

Super PI is a rather deprecated application.

 

[ShintaiDK]

Remember different boards etc. Such roundups usually consist of different hardware and setups due to reused numbers.

 

[Arkaign]

I'm confused by the Mafia II numbers. It says they used a 7870 or something like that, but all the CPUs are running 20fps or less? At lowest settings it looks like? What the hell?

 

[inf64]

Yeah Mafia's fps seems very low considering low resolution they used. 8350 did rather well I must say,except for the power draw it's a great step up from 1100T(since 8150 was not). They need to shrink this thing to 28 or 20nm or release the Kraken(SR) soon. VIshera on 28nm would be so much better product just thanks to lower power draw.

 

[exar333]

Yeah Mafia's fps seems very low considering low resolution they used. 8350 did rather well I must say,except for the power draw it's a great step up from 1100T(since 8150 was not). They need to shrink this thing to 28 or 20nm or release the Kraken(SR) soon. VIshera on 28nm would be so much better product just thanks to lower power draw.

28nm will not be any better for AMD most likely. They will be losing SOI and only making a half-node drop. Likely a draw for them...

 

[Rifter]

28nm will not be any better for AMD most likely. They will be losing SOI and only making a half-node drop. Likely a draw for them...

I was going to mention this as well, if AMD is going to seriously compete against intel again shrinks arnt going to cut it they need a new design from the ground up. One that cuts power useage by over 50% from their current lineup.

 

[Kristijonas]

Some results are suspicious and choice of CPUs to test is dumb. Otherwise everything is more or less as expected.

 

[beginner99]

Some results are suspicious and choice of CPUs to test is dumb. Otherwise everything is more or less as expected.

Yeah like that my 3 year old lynnfield CPU beats an fx-8350 in performance / watt. That just shows how bad BD uArch is.

 

[coffeejunkee]

22nm 3770K uses more power than 32nm 2700K ?? really ???

They could measure power consumption in a highly MT application like HandBrake, Super Pi is such a fail for power consumption measurements for today's microprocessors.

Hmm? Says in the image they are using Prime95...but still strange to see 3770K use more than 2700K. I remember from the initial reviews all Ivy's using less than their Sandy counterparts.

Remember different boards etc. Such roundups usually consist of different hardware and setups due to reused numbers.

I assume TH is smart enough to use the same motherboard for same socket cpu's, but unfortunately it's not mentioned which boards were used.

 

[SPBHM]

I'm confused by the Mafia II numbers. It says they used a 7870 or something like that, but all the CPUs are running 20fps or less? At lowest settings it looks like? What the hell?

maybe they left the physx settings (made for GPU acceleration) at the maximum settings but running on the CPU, or they found a really bad spot on the game...

anyway, interesting test, but they should have included at least the i3 2100, and the i3 3220.

also, since we are talking about 86 CPUs, at least two Core 2 (on duo and one quad) and one k8 (6000+) would be nice.

 

[moonbogg]

People with Nehalems are still getting by just fine, lol.