[VR-Zone] Intel Core i7 3770K@4.6GHz vs AMD FX-8150@4.6GHz

[Homeles]

And you , you do ??...





Assuming that the code is equally optimuzed , wich is not the case
at all , otherwise in FP calculus the BD should do way better given
its instrcution set.

You can't even write proper English... should probably learn how to do that before arguing on a forum that uses the English language.

The floating point unit is what's flexing its muscle in a hash workload, so I'm not even sure what you're trying to argue there.

And yes, I do. Cache sizes are different between IVB and BD. Cache latencies are different. Cache miss penalty is different. Cache associativity is different. Memory controllers are different. Load/store units are different. Integer execution units are different. Floating point units are different. Schedulers are different. Decoders are different. IVB has a µop cache, Bulldozer does not. Branch prediction units are different. Physical register files are different. Latencies between all these units are different. Fab processes are different. Flip flops are different

Need I say more?

Thinking that a CPU is going to perform the same relative to a completely different CPU requires a juvenile perspective of processors. There are far too many variables, and each architecture is going to have its own strengths and weaknesses.

Yes, Bulldozer supports some fancy instructions that Ivy Bridge doesn't. Even if Bulldozer is using XOP or FMA4, that doesn't necessarily mean that it will outperform Ivy Bridge.

Blaming Bulldozer's shortcomings solely on "poor optimization" is just asinine. Is it part of the problem? Sure. The reason that Bulldozer failed lies far deeper than code and compiling though.

 

[superccs]

You clearly don't understand the fundamentals of a CPU. Different architectures are better at different things -- or in the case of Bulldozer, worse at all the things. With the exception of hashing that is, which is something that is unarguably better left for GPGPU.

Yep no clue. I work with carbon not silicon.

Doesn't AMD always deliver disappointments before real performers though? (ie, Phenom -> Phenom II)

 

[Vesku]

Not sure what the point of this is, Bulldozer was not great for the MSRP at launch versus SB. The Phenom II x6 was a better choice while it was available. To add to the WTF, why didn't they compare the 3570K at 4.6GHz. Better to compare in same price tier than same thread processing count. Not that Bulldozer would look any better but that the information would actually have some purpose to it.

Better to save energy for comparing the Piledriver refresh to original BD and current IB.

 

[2is]

And you , you do ??...





Assuming that the code is equally optimuzed , wich is not the case
at all , otherwise in FP calculus the BD should do way better given
its instrcution set.

Lets go ahead and assume for a minute you're right. That BD has all this untapped potential... Who cares? Hardware is NOTHING without software. If very little to nothing is designed to take advantage of BD's greatness, then it's shit.

 

[AnonymouseUser]

The difference in power consumption between the 2 is astronomical. http://www.bit-tech.net/hardware/2012/04/23/intel-core-i7-3770k-review/8

244 vs 586 watts. The difference is enough to power an entire midrange gaming PC.

FWIW, the high-end i7's don't fair much better at load, and are actually worse at idle, all while using a mATX board vs the AMDs on an ATX.

Note: the AMD chips were tested in an ATX motherboard, while the Intel LGA1155 chips were tested in a micro-ATX board. This difference can account for up to 20W, as we found in our Energy Efficient Hardware feature.

 

[Rifter]

AMD is out of the high end CPU game, this is old news now.

Besides AMD cant even compete with their own best Phenom II's(the 6 core thubans)what makes anyone think they will stand a chance against intels newest chips?

 

[Homeles]

Yep no clue. I work with carbon not silicon.

Doesn't AMD always deliver disappointments before real performers though? (ie, Phenom -> Phenom II)

Eh, maybe. Phenom simply was too much for the 65nm process. Quad core on a single die wasn't optimal, and Phenom flopped hard as a result. Phenom would have been perfectly fine had it been a native dual core design. Of course it wasn't, and AMD wasn't able to get what Phenom should have been until they added some L3 and shrank the transistors, giving us Phenom II. I find it hard to believe that there was such a huge difference made by a simple shrink -- the improvement was larger than Penyrn, Intel's best "tick" to date.

Bulldozer looks to be a similar issue -- AMD trying to accomplish too much before the fab tech is ready to handle it. It's a prime example of why you shouldn't introduce a new microarchitecture on a new process. The concept behind Bulldozer is brilliant. The execution...

For a simple "tweak," Piledriver is looking really good, but there's no way in hell it'll be as good as the boost that Phenom II gave over Phenom. We already know the ballpark figures for it, and you can't dig yourself out of a hole that large on the same process. Steamroller based FX won't appear until 2014 -- if at all. Still, Piledriver will likely put AMD in a similar position to where they were with Phenom II: a budget alternative to Intel.

 

[Conscript]

We don't have a clue how Haswell would perform but do you think that it would perform worse than its predecessor considering that this is Intel we're talking about?

So cute how innocent the young'ins can be...

 

[BD231]

Piledriver will barely improve upon BD, you can count on it. AMD CPU's will not be in contention with Intel for quite a few years to come, they just don't have the tech or the engineers to pull it off right now.

 

[Homeles]

So cute how innocent the young'ins can be...

Considering Hawell's just an evolution, the only way it'd be worse than Ivy Bridge is if there was some sort of errata issue.

Piledriver will barely improve upon BD, you can count on it. AMD CPU's will not be in contention with Intel for quite a few years to come, they just don't have the tech or the engineers to pull it off right now.

Brilliant thing to say, considering there's already benchmarks and reviews of the architecture out there.

 

[AtenRa]

Im not trying to say that Core i7 3770K is not faster but this is not a review, it is a parody.

No NB overclocking on the FX system
No information on IQ settings on Games
Not to mention the 3770K is ~75% more expensive.

Also to note, the performance difference is only at 20-25% on average in real life applications like x264, CB 11.5 MT and more.

 

[dma0991]

So cute how innocent the young'ins can be...

I see your point, Intel has its share of failures in the past but that doesn't mean that they are taking it easy. Even by an off chance that Haswell doesn't even compete head to head with SB/IB, the extra TDP headroom from IB(77W) to Haswell(95W) should offer some higher clockspeeds to compensate for whatever Haswell is not good at. With AMD as bad as it is right now in terms of competitiveness, Intel could release a lemon and people would still buy it.

 

[dma0991]

No NB overclocking on the FX system

I doubt overclocking the NB would be a game changer in this test.

No information on IQ settings on Games

We could assume that the IQ settings are the same for both setups. Not sure how stating the IQ settings will add more value to the test.

Not to mention the 3770K is ~75% more expensive.

This is supposed to be a test of what's the best that each company has to offer. I suppose that it isn't fair since the best from Intel would be the Core i7 3960X but that would make it a lot worse in terms of price difference compared to the Core i7 3770K.

 

[Mallibu]

Im not trying to say that Core i7 3770K is not faster but this is not a review, it is a parody.

No NB overclocking on the FX system
No information on IQ settings on Games
Not to mention the 3770K is ~75% more expensive.

Also to note, the performance difference is only at 20-25% on average in real life applications like x264, CB 11.5 MT and more.

Since you like percentages so much, here's one for you:
FX 8150 is 250% the price of Pentium G850, while being 15% slower in games.
http://www.anandtech.com/bench/Product/434?vs=404

 

[Abwx]

You can't even write proper English... should probably learn how to do that before arguing on a forum that uses the English language.

Yet , it was enough to be understood , so you re just contradicting
yourself with this kind of trashy remarks.....

The floating point unit is what's flexing its muscle in a hash workload, so I'm not even sure what you're trying to argue there.

Cinema4D rendering is massively floating point , yet BD does
not do well in this bench as well as in anything that is FP dependant.

And yes, I do. Cache sizes are different between IVB and BD. Cache latencies are different. Cache miss penalty is different. Cache associativity is different. Memory controllers are different. Load/store units are different. Integer execution units are different. Floating point units are different. Schedulers are different. Decoders are different. IVB has a µop cache, Bulldozer does not. Branch prediction units are different. Physical register files are different. Latencies between all these units are different. Fab processes are different. Flip flops are different

Need I say more?

NO , it s enough hollow wordings with no usefull point....

Thinking that a CPU is going to perform the same relative to a completely different CPU requires a juvenile perspective of processors. There are far too many variables, and each architecture is going to have its own strengths and weaknesses.

Hence the necessity of specific optimizations for each uarch.....

Yes, Bulldozer supports some fancy instructions that Ivy Bridge doesn't. Even if Bulldozer is using XOP or FMA4, that doesn't necessarily mean that it will outperform Ivy Bridge.

See my previous quote above.....

Blaming Bulldozer's shortcomings solely on "poor optimization" is just asinine. Is it part of the problem? Sure. The reason that Bulldozer failed lies far deeper than code and compiling though.

Optimizations are still lacking for this CPU , otherwise , its
main problem are internal intermediate latencies that had
to be hugely increased to get te silicon working properly.

These shortcomings wont be completely adressed before
steamroller cpus.

 

[AtenRa]

I doubt overclocking the NB would be a game changer in this test.

There are applications that will get a performance increase with OC the NB.

We could assume that the IQ settings are the same for both setups. Not sure how stating the IQ settings will add more value to the test.

Im sure you would find the same review a lot more professional and without raising any doubts if they would state more information about each bench.

This is supposed to be a test of what's the best that each company has to offer. I suppose that it isn't fair since the best from Intel would be the Core i7 3960X but that would make it a lot worse in terms of price difference compared to the Core i7 3770K.

So, next one in line would be a $349 Core i7 3770K against a $119 A8-3870K. Those two are the best that each company has to offer with integrated iGPUs 😛

 

[sm625]

But BD can import pictures faster. That is the secret sauce in their design.

 

[Edrick]

So, next one in line would be a $349 Core i7 3770K against a $119 A8-3870K. Those two are the best that each company has to offer with integrated iGPUs

You know it is funny that before DB launched, people were comparing it to the Intel "E" class of CPUs (980X, 3960k, etc) due to its core count. Then when it got crushed by those CPUs, all the AMD fans started posting about market segments and such saying that BD should be compared to the 2500k and 2600k series of CPUs, in other words mainstream. Now that BD is getting crushed by those mainsteam CPUs, now we have AMD fans saying that is not a fair comparison either.

So bottom line is the AMD fans will ONLY compare BD to something that BD can beat, regardless of core count, market segment, price, etc. Good to know.

 

[Phynaz]

You can't even write proper English...

Blaming Bulldozer's shortcomings solely on "poor optimization" is just asinine.

Let me acquaint you with the rules around here:

1) No trolling, flaming or personally attacking members. Deftly attacking ideas and backing up arguments with facts is acceptable and encouraged. Attacking other members personally and purposefully causing trouble with no motive other than to upset the crowd is not allowed.

 

[Homeles]

Let me acquaint you with the rules around here:

Is it not important to know English around here? Is it not a good suggestion to learn how to write before trying to make an argument?

The second one is not a personal attack -- it's attacking an argument. Learn the difference.

 

[StrangerGuy]

Is it not important to know English around here? Is it not a good suggestion to learn how to write before trying to make an argument?

The second one is not a personal attack -- it's attacking an argument. Learn the difference.

Yours is a much as a personal attack as calling a black guy black.

 

[Phynaz]

Is it not important to know English around here?

Checks forum rules....Nope, not in the rules here.

 

[StrangerGuy]

Checks forum rules....Nope, not in the rules here.

Hey look do us a favor and take your juvenile musings elsewhere.

 

[Homeles]

Checks forum rules....Nope, not in the rules here.

Hint: I'm not talking about the rules.

 

[Phynaz]

Edit: Nevermind.