AMD Takes The CPU Performance Crown For The First Time in Almost a Decade.

[Glenwing]

But there IS a difference between them and that difference is that they are faster, considering how little the sandy chips were beat by the result would have been completely opposite.

 

Agreed, but they are close enough that it's not a big concern if they're not on the chart.  You can just add 10% or so in your head.  Unlike if we were missing the FX-8320 and had the Phenom II X6 1090T instead or something.

[leipero]

CPU controls things like physics and AI.  Turning down the detail settings and the resolution affects GPU stress only.  CPU load remains the same even when settings are turned down, if you know which ones to turn off and which ones not to touch.

Depending on game, in majority of games, CPU and memory subsystem is used differently depending on settings. CPU also need to do other work for GPU driver, therefore, that methodology is flawed (turning all settings on low).

 

But you are right, if they turn down only settings that are not affected by CPU that much, it should be OK, but that is not the case, since majority of tests use presets (in order to be replicable). Best way to test is at max settings (or at all settings as many did), in order to see real world scenario. That will tax all components to teh maximum, even tho games are not really good way to represent any component speed, but how well that software will work on different components (how well is optimized), at the end, that is all users care for.

[B NEGATIVE]

A heavily OC'd to the limit arch to a arch at stock with a massive clock deficit.....and you claim its a win?

 

Now lets apply a 4.5 OC to the Intel,still 500 mhz below the AMD clock,and watch it kick mud into the AMD chips face.

 

Render performance is a better indicator than just games performance. Keyshot something and take the time spent

[LAwLz]

Also, list you showed is non relevant, since it is present-past, and even so, those games are playable on all platforms, no big deal anyway.

That's not how science works. You can't say "AMD performs better than Intel!" and then only post two games, while Intel wins in 22 other games, and they are tied in 11 games. That's simply spreading misinformation and cherry picking. OP is suffering heavily from confirmation bias. A lot of the games on that list are new by the way, you can't just say "well that game is 3 months old so that doesn't count". Hell Blood Dragon is newer than FarCry 3 so by your logic, FarCry 3 shouldn't count. But where do you draw the line between what's considered "old and not relevant" and "new and relevant"?

[Lord Pantaloons]

That's not how science works. You can't say "AMD performs better than Intel!" and then only post two games, while Intel wins in 22 other games, and they are tied in 11 games. That's simply spreading misinformation and cherry picking. OP is suffering heavily from confirmation bias. A lot of the games on that list are new by the way, you can't just say "well that game is 3 months old so that doesn't count". Hell Blood Dragon is newer than FarCry 3 so by your logic, FarCry 3 shouldn't count. But where do you draw the line between what's considered "old and not relevant" and "new and relevant"?

This might be one of those "Its better because it's worse" scenarios.

[yetisunny]

the amd 8590 is only 300 euros in my country is this normal, since it's 500 dollars for you guys?

[Lord Pantaloons]

the amd 8590 is only 300 euros in my country is this normal, since it's 500 dollars for you guys?

 

 

that's about $400-ish USD, the 9590 is coming in at 800-1000 USD for me(amazon) the 9370 is only 300 USD.

[A/C]

Can anyone say how AMD handles SSD's in RAID 0 compared to intel's Z87 line? That's kinda a dealbreaker for me.

[tedjani]

AMD YEAHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH

[ProKoN]

the first graph of b4 is kinda bogus. i just averaged 58 fps @ 1680x1050 with the same settings .. ultra with no AA

 

min-31fps

 

max- 81fps

 

average 58fps

 

4670K @ 4.8GHz

 

GTX 460 SLI

 

Thats similar to a 2600K or 8350 with a titan? I doubt it.

 

I just so happen to have a 4770k to test out as well. I will most likely make a video response to this poor graph.

 

AMD takes the crown???  if your referring to the $200.00 and under cpu bin... your right!

[leipero]

That's not how science works. You can't say "AMD performs better than Intel!" and then only post two games, while Intel wins in 22 other games, and they are tied in 11 games. That's simply spreading misinformation and cherry picking. OP is suffering heavily from confirmation bias. A lot of the games on that list are new by the way, you can't just say "well that game is 3 months old so that doesn't count". Hell Blood Dragon is newer than FarCry 3 so by your logic, FarCry 3 shouldn't count. But where do you draw the line between what's considered "old and not relevant" and "new and relevant"?

I guess you didn't read my post. Science? Game testing =/= science, in fact, far away from it.

[LAwLz]

I guess you didn't read my post. Science? Game testing =/= science, in fact, far away from it.

Science (noun):

1) The observation, identification, description, experimental investigation, and theoretical explanation of phenomena.

2) Methodological activity, discipline, or study.

 

Comparing different processors is science, and OP is terrible at it. Also, you are extremely biased and fanboy-ish. Please stop that and let the numbers speak for themselves, okay? Even if you consider all of them "playable" it still doesn't make up for the fact that OP is spreading misinformation and lying.

[leipero]

Science (noun):

1) The observation, identification, description, experimental investigation, and theoretical explanation of phenomena.

2) Methodological activity, discipline, or study.

 

Comparing different processors is science, and OP is terrible at it. Also, you are extremely biased and fanboy-ish. Please stop that and let the numbers speak for themselves, okay? Even if you consider all of them "playable" it still doesn't make up for the fact that OP is spreading misinformation and lying.

1. OP actually did that, you missed the point completely.

 

Comparing different processors is not what "game tests" do, as i said already, you didn't read my post, or you didn't understand it. I really can't repeat myself, read again.

[LAwLz]

1. OP actually did that, you missed the point completely.

 

Comparing different processors is not what "game tests" do, as i said already, you didn't read my post, or you didn't understand it. I really can't repeat myself, read again.

Yes exactly. Now you're arguing against yourself. You said that it wasn't science (trying to argue semantics instead of the core issue), and now you're saying it is. So good that you agree with me and realize you were wrong.

 

About your second part of the post, it doesn't really make any sense. Which post are you referring to exactly? The one where you say my list isn't relevant because it contains some slightly older games as well (even though they are about the same age as the games OP posted)? Yeah, that post makes no sense whatsoever. It just shows how biased you are when you start going "well these 2 games where AMD wins are more important than the 22 other games where Intel wins".

So what is your point? My point was that OP is a lying bastard, and that people on this forum seems to just thumb up posts which agrees with their opinion, even if the post is a big pile of lies and bullshit.

[leipero]

Yes exactly. Now you're arguing against yourself. You said that it wasn't science (trying to argue semantics instead of the core issue), and now you're saying it is. So good that you agree with me and realize you were wrong.

 

About your second part of the post, it doesn't really make any sense. Which post are you referring to exactly? The one where you say my list isn't relevant because it contains some slightly older games as well (even though they are about the same age as the games OP posted)? Yeah, that post makes no sense whatsoever. It just shows how biased you are when you start going "well these 2 games where AMD wins are more important than the 22 other games where Intel wins".

So what is your point? My point was that OP is a lying bastard, and that people on this forum seems to just thumb up posts which agrees with their opinion, even if the post is a big pile of lies and bullshit.

I've just applied your logic on OP, since you are the one who said he is biased, and science do not work that way. However, i didn't contradict my post on last page (i guess you still didn't read it).

 

For the second thing, don't put the words I didn't said in my mouths. I'm sorry if you missed the point, but that wouldn't happen if you did read my first post here. Now stay on topic.

 

Also, using your "scientific" methodology, and just reading title of this topic and context of it, you can clearly see who is biased here.

[eshbop]

I don't agree with that methodology. Maxing out games makes more stress on CPU, memory subsystem, and all components in general. Also, you missed the point of OP, it shows that when coded properly, games/programs do run same or better on AMD platform, and with new systems (maybe even with upgraded DX) it will be potentially easy for anyone to code and use CPU's to it's potential.

 

Also, list you showed is non relevant, since it is present-past, and even so, those games are playable on all platforms, no big deal anyway.

With the methodology LAwLz suggested you are making the CPU a bottleneck rather than the GPU. If a GPU is bottlenecking frame rates then you can't gauge CPU performance because the CPU is providing more than enough processing power and still isn't fully loaded. Testing 'normally' with i5 2500k plus a TITAN VS i5 4670k plus a TItan VS 8350 plus a Titan are all essentially tests of the GPU. When you make it so the GPU isn't 100% loaded (ie isn't the bottleneck) then the CPU is the limiting factor, allowing you to compare just CPU performance. And if you want to be pedantic about it all the CPUs should be running the same frequencies too.

 

In saying that, since no one runs games like this it is kind of pointless.

[LAwLz]

I've just applied your logic on OP, since you are the one who said he is biased, and science do not work that way. However, i didn't contradict my post on last page (i guess you still didn't read it).

 

For the second thing, don't put the words I didn't said in my mouths. I'm sorry if you missed the point, but that wouldn't happen if you did read my first post here. Now stay on topic.

OK since you refuse to link me to the post you keep referring to I looked it up myself. It is this one, correct?

 

I don't agree with that methodology. Maxing out games makes more stress on CPU, memory subsystem, and all components in general. Also, you missed the point of OP, it shows that when coded properly, games/programs do run same or better on AMD platform, and with new systems (maybe even with upgraded DX) it will be potentially easy for anyone to code and use CPU's to it's potential.

 

Also, list you showed is non relevant, since it is present-past, and even so, those games are playable on all platforms, no big deal anyway.

No, you should use low settings when testing a CPU. Again, it eliminates GPU bottlenecks as much as possible, which is a big factor when using games as CPU benchmarks (and why you often see so many gaming benchmarks showing next to no difference between CPUs).

Of course there are some exceptions to the rule but I am talking about general modern games which are mostly bottlenecked by your GPU. If you run these games on high settings then it becomes a GU test rather than a CPU test.

 

I disagree that my list is not relevant. Please explain why my list is objectively not relevant. The whole "some games are not brand new" argument which I assume you were implying when you said "it's present-past" is wrong because a lot of the games are newer than some of the games OP posted, yet you think OP's games are relevant. Your "those games are playable on all platforms" argument doesn't really make any sense either, because you use that when Intel wins, but don't use that logic when AMD wins. Again, your bias is shining through.

 

 

 

Yes, you did contradict your post (about this not being science). First you said:

 

I guess you didn't read my post. Science? Game testing =/= science, in fact, far away from it.

Then when I posted the definition of science (which fits what OP was doing, and what I was doing in my first post in this thread) you replied with:

OP actually did that, you missed the point completely.

So yeah, first you say OP wasn't doing science and then you said he was. Anyway we are arguing about semantics here which is pointless.

[leipero]

In saying that, since no one runs games like this it is kind of pointless.

Most important part, however, it should be noted that majority of games use system differently, therefore, that methodology is flawed in two ways, one you already said, and another considering how games stress components differently on different settings (not only GPU as many like to suggest). If you carefully read OP and title of the topic, you will see that OP said most things from neutral point of view.

 

Here is not OP or any post now in question, but some other things.

[leipero]

I disagree that my list is not relevant. Please explain why my list is objectively not relevant. The whole "some games are not brand new" argument which I assume you were implying when you said "it's present-past" is wrong because a lot of the games are newer than some of the games OP posted, yet you think OP's games are relevant. Your "those games are playable on all platforms" argument doesn't really make any sense either, because you use that when Intel wins, but don't use that logic when AMD wins. Again, your bias is shining through.

 

 

 

Yes, you did contradict your post (about this not being science). First you said:

 

Then when I posted the definition of science (which fits what OP was doing, and what I was doing in my first post in this thread) you replied with:

So yeah, first you say OP wasn't doing science and then you said he was. Anyway we are arguing about semantics here which is pointless.

I argue that your list is irrelevant for this topic (considering what OP said, and the name of the topic), maybe i didn't expressed myself well. Just by reading OP and title of the topic, you would see that list you gave here is completely irrelevant, unnecessary and it shows something completely different apart from me or OP being biased etc.

 

And i need to ask you one more time, do not put something i didn't said in my mouth. It would be very nice from your side not to do that, thank you.

 

I already told you that all i did is applied YOUR logic on OP, from YOUR perspective, that isn't my perspective, it is your own, and it contradict your point of view.

[LAwLz]

Most important part, however, it should be noted that majority of games use system differently, therefore, that methodology is flawed in two ways, one you already said, and another considering how games stress components differently on different settings (not only GPU as many like to suggest). 

Turning up the graphics will mostly affect the GPU. Yes higher graphics settings will put slightly more stress on the CPU since it has to do more draw calls for example, but even then you will still isolate CPU performance a lot better when you run games at low settings. That's why you see a much bigger difference between CPUs at low settings, compared to the difference on high settings where CPUs usually performs very similarly.

 

If you carefully read OP and title of the topic, you will see that OP said most things from neutral point of view.

Ahahaha yeah right. Like how it says that AMD now performs better than Intel CPUs, and then cherry picked the only 2 benchmarks AMD wins in and compeltely ignored, didn't even mention (didn't even post a link to the source so everyone could check out the other tests) the other 33 tests where AMD didn't win.

Please note how I posted a ton of benchmarks and made a very easy list to give people a quick overview of the benchmarks Intel wins in, AMD wins in and which games they are tied in, as well as linked to the original article so people can double check everything I said. That's how you are unbiased. Cherry picking benchmarks is not unbiased/neutral.

OP clearly had an agenda when he made that post, and being honest or being neutral was not on that agenda.

[LAwLz]

I argue that your list is irrelevant for this topic (considering what OP said, and the name of the topic), maybe i didn't expressed myself well. Just by reading OP and title of the topic, you would see that list you gave here is completely irrelevant, unnecessary and it shows something completely different apart from me or OP being biased etc.

Oh right, so my list is irrelevant because it disproves OP? So if I make a thread which says "Intel is better than AMD", then only post 2 benchmarks, both of which I have carefully cherry picked in order to avoid posting anything where AMD beats Intel, then anything that shows AMD being better is "irrelevant"? The benchmarks I posted from that shady Russian site (same as OP used) are very relevant to this thread, because they exposes how biased OP is. You can't say it is irrelevant just because it proves that OP is flat out wrong.

[eshbop]

Intel has better single threaded performance, this is still what's most important for current games. Games that aren't making use of more than the 4cores of an i5 aren't going to be much better on a high end AMD chip (because performance is GPU bound anyway).

 

If current games (and some are starting too) made efficient use of 8 cores then this comparison wouldn't be so simple . But chances are, by the time most games are heavily multi-threaded, Intel will also have CPUs that can take advantage of this.

[LAwLz]

Intel has better single threaded performance, this is still what is most important for current games. Games that aren't making use of more than the 4cores of an i5 aren't going to be much better (if they are) on a high end AMD chip.

 

If current games made efficient use of 8 cores then this comparison wouldn't be so simple. But chances are, buy the time most games are heavily multi-threaded Intel will also have chips that can take advantage of this.

Yeah even if the new consoles has 8 cores it will still take quite a while for that to be passed onto the PC and be properly utilized (remember, just being able to use 8 cores doesn't mean they are effectively used). If you take into consideration that most games are heavily GPU bottlenecked as well, then it makes even less of a difference. When games start using (and demanding) 8 cores, the CPU market will probably look very different compared to how it looks today.

[eshbop]

Yeah even if the new consoles has 8 cores it will still take quite a while for that to be passed onto the PC and be properly utilized (remember, just being able to use 8 cores doesn't mean they are effectively used). If you take into consideration that most games are heavily GPU bottlenecked as well, then it makes even less of a difference. When games start using (and demanding) 8 cores, the CPU market will probably look very different compared to how it looks today.

Plus not all the cores in the next gen consoles will contribute to the processing power being thrown at the game (running OSs in the background, etc).

 

This makes it interesting to see what AMD's Mantle brings with it's ability to automatically offload processing power to each core despite how multi-thread aware (or unaware) a game's code may be.

[Lord Pantaloons]

Yeah even if the new consoles has 8 cores it will still take quite a while for that to be passed onto the PC and be properly utilized (remember, just being able to use 8 cores doesn't mean they are effectively used). If you take into consideration that most games are heavily GPU bottlenecked as well, then it makes even less of a difference. When games start using (and demanding) 8 cores, the CPU market will probably look very different compared to how it looks today.

just think, 6 core i5's, 8 core i7's, 12 core E-i7's