AMD FX 8370 vs Intel I7 5960x [GTX 970 SLI 4K benchmarks]

[Raichuboy17]

Show me the rendering times and then we will talk AMD... Then we will talk.

[patrickjp93]

Ok, life lesson my friend.

 

That attitude, is what will get you in troubble with your peers. I tried it, doesnt work. What work you think you've done, anyone with 3 years of work experience, would have done that or way more. I know it may seem like everyone just sits at their asses, but assuming you know so much more then people who actually are working. That is what gets you labled as "cocky new kid" pretty darn fast.

 

I am sure you know your stuff. But since you appear to be a student, a word of advice from someone who have gone through school and have been working for a few years... We got a apprentice in our company, he is REALLY good, but he has that attitude of "i know so much, and i know i'm good"... there is A LOT of small corrections and ways of doing our work that he doesnt know, or doesnt realize. Mind you, i work as an electrician, so my type of work is more on the physical side of things. But never think the work you have done compares to what those who have been working for even a few years...

 

In comparison... Before i started working full time some 4-5 years ago, i thought that the 10 miles of cables i had laid down was A LOT.... and that a 50sqmm cable was large... Lets just say, that is nothing. I have laid like 100s of miles of cables within just those 4 years and cable sizes of 240sqmm isnt uncommon. It is rather common... When you think something is alot before starting to work, it really is not...

 

Our apprentice learnt that, when we handed him 3 reels of 95 meters of cable wheighing a modest 22 pounds pr meter..... he got a bit more humble after that.

I know there are people who have plenty to teach me. That said, if you're not teaching me anything new, please get out of the way, because I have work to do. 

 

The amount of work is not as impressive as the quality. Using 10% fewer clock cycles to do 15% more work in 2000 lines fewer of code and between 10 and 30% less time is an accomplishment, especially in systems programming. I also didn;t say I knew more. I produce high quality code. It's always clear, concise, extensible, and documented. You dismissing what I've done without seeing it is your biggest mistake in this, and IBM would much agree with me, as I now have a lead developer role for a new team focused on medical applications of artificial intelligence the moment I graduate in December. It's not cockiness if you actually are a star player, and so far no one has come along and beaten me when I've set my mind to it. It hasn't happened since I started school and hasn't happened since I started computer programming. Evaluate me on my product, not my attitude, because I can learn everything you do and do it better in the end if you get in my way, and I will leave you high and dry for it. That is the kind of vicious bastard I am, and I may not be liked, but I am respected. I am a team player, but I refuse to work with egotists and morons. There may be something I can do better. Either show me what it is or go back to work on your own stuff and stay out of my hair.

 

In regards to your post, I pointed out succinctly exactly what was wrong with it on factual basis. I have posed a thesis about the quality of game programming and proven how easy it actually is to build an extensible system if you start it correctly. All it requires is planning out your communications ahead of time and some basic math. Until you're tuning to the clock cycle, you're not going to optimize any better than that template code will allow you. That's why it's the same concept used to build and validate huge HPC programs (just switch OpenMP for MPI containing OpenMP and OpenCL runtimes)  before going back and doing the threads by hand.

 

Coding and laying wires are very different, and while I will not claim to have even a clue about how to wire a house or set up power lines, I can code extensible systems and pieces of them that self-balance just as well as a 5-year industry coder who is ready to jump into supercomputing and learn from the pros. Game programmers even of five years are well beneath my skill level in everything but using Direct X/OpenGL itself. I can make the AI better than all of them because I've been involved in AI training and research since I started college. I can do the network protocols for an MMO better because I actually plan them out ahead of time to be extensible and flexible. I can do the physics engines better because I actually finished all three levels of calculus courses and linear algebra and discrete mathematics and took undergrad and graduate algorithms and know 90% of the tricks to making approximation systems for all sorts of physical calculations. I also hand-coded in assembly the SIMD versions of them to handle stuff like ray tracing using AVX 256 extensions. My 2600K can ray trace better than any GPU implementation ever will because I use the correct tool for the job (a branch predictor). I am damn good at this, and I can't believe you're being so stupid as to dismiss that just because I'm a student. Prodigies come around whoop people in the tail end now and again. If you dismiss anyone because of age or job status, you will someday miss a golden chance. It's utter foolishness. Rate me on real merits, not blind ignorance.

 

I am damn good at what I do, and I'm not taking lip from an amateur logician like you who broke 8 of the 10 fundamental rules of informal logic all in one post and wouldn't know his ass from his elbow knee deep in code at a college freshman level. I'll take lip when someone shows me how to do something better. Until then, I'm good at what I do, and I invite anyone to beat me, because being wrong or losing a fight is always more fun, because I always learn something. You could do with some humbling yourself, because you've completely misunderstood and misrepresented who I am.

[patrickjp93]

You are limiting it to background services..

It is still take up execution ports, it could also have influence in regards to the different data structures (IE limiting the amount of entries the thread can have in the different queues/windows, register files, cache, and so on). It certainly is not as big as an issue as it used to be, as Intel have improved hyper-threading over generation, but some might experience worse performance when having hyper-threading enabled (The older the "version" of hyper-threading, the more apparent the problem is).

Haswells hyper-threading should not really have such a big penalty on this.

If you're running a ton of background executables, that's your fault. You can also designate them all to a single core in Windows anyway, as in shove them all on core 7/8 so 3/4 (0-indexed or not?) is still dominant and has a less heavy thread of your main program loaded on it as compared to having them interspersed or even running on core 0/1. This isn't that hard. Cache being affected is unavoidable anyway, a bit of a moot point.

[Revinval]

If you save $600 on your cpu that $600 more sweet sweet gpu power.

[Sintezza]

Would like to see these tests done again, with a double check on the 5960X setup.

 

But as far as 4k gaming goes, you are basicly gpu limmited in most scenario´s.

which basicly means, that you should see similar scores arround both cpu´s, or atleast tiny diffrences within a margin of error. (depending from game to game)

the cpu doesnt realy make any significant diffrence to that. (imo)

 

The bottom line of this article is basicly, that a 5960X shouldnt be conciderd as a gaming cpu in the first place.

But thats of course allready a known fact.

 

what do you think?

[Biggerisbetter]

This tread needs to be closed. Not one person who oppose the findings has come up with enough credible evidence to disprove the benchmark findings. All I see are a bunch of opinions and no hard DATA. 

[Imabigmac]

This tread needs to be closed. Not one person who oppose the findings has come up with enough credible evidence to disprove the benchmark findings. All I see are a bunch of opinions and no hard DATA. 

It should have been closed a long, long time ago.

 

-snip-

 

I just, I just don't know anymore. 

[tono]

It should have been closed a long, lon

 

 

I just, I just don't know anymore. 

g time ago.

Its a pretty controversial thread, but yeah I agree it needs to be closed or at least have a time limited lock. So people can at least cool down and stay civil. 

[Majestic]

@Majestic what do you think?

 

Basically the same, min. fps don't make sense. The tests at which CPU is significantly different don't seem to favor the 5960X or overclocking, which is odd. Some of the additional tests make no sense, see patrick's post. And the caveat needs to be placed in the test, or different games have to be tested, that CPU intensive games (gameworks titles, multiplayer games, RTS/MMO) aren't represented by these tests.

 

Trucker2 needs to alleviate his cognitive dissonance.

[Dabombinable]

Basically the same, min. fps don't make sense. The tests at which CPU is significantly different don't seem to favor the 5960X or overclocking, which is odd. Some of the additional tests make no sense, see patrick's post. And the caveat needs to be placed in the test, or different games have to be tested, that CPU intensive games (gameworks titles, multiplayer games, RTS/MMO) aren't represented by these tests.

 

Trucker2 needs to alleviate his cognitive dissonance.

That didn't happen in the last AMD thread he was involved with, I doubt it will happen in this one.

[rentaspoon]

Surprised the ban hammer isn't out yet

Would like to see these tests done again, with a double check on the 5960X setup.

I personaly dont believe it either that a FX8370 scores higher then a 5960X on minimums at the same clock speed.

Maybe the 5960X wasnt running stable? idk.

 

But as far as 4k gaming goes, you are basicly gpu limmited in most scenario´s.

which basicly means, that you should see similar scores arround both cpu´s, or atleast tiny diffrences within a margin of error. (depending from game to game)

Its just a simple fact that haswell cpu´s have a better core performance, so its not very likely that it scores that much lower on minimums.

Unless there was something going on with that particular test system...?

 

But still the fact stands that at 4k you are gpu limmited mainaly,

and the cpu doesnt realy make any significant diffrence to that. (imo)

 

The bottom line of this article is basicly, that a 5960X shouldnt be conciderd as a gaming cpu in the first place.

But thats of course allready a known fact.

 

@Majestic what do you think?

While I agree with 99% of what you say, the truth is I think we are going to see a lot of benchmarks this way, 4k is the new 1080p and I wouldn't be surprised if we start to see 4k hardware and monitors pushed for next year.

 

The way I see it is if your still running old hardware, there is no reason to change.. yet

[marldorthegreat]

Not surprising given the GPU dominant nature of 4K. The i7 will wipe the floor with the FX during anything else.

If it is gpu bound, why the the i7 taking the hit and not the 8370? By your logic the i7 should be performing better.

[Dabombinable]

If it is gpu bound, why the the i7 taking the hit and not the 8370? By your logic the i7 should be performing better.

The GPU is the limit, which is why the FPS are so similar.

[Norman_the_Owl]

What could people possibly be arguing about?

 

It's 4k, of course their will be less difference. GPU is the bottleneck, m80s

[Overkilled]

So thats what AMD fanboys came to? Faked benchmarks

[Dabombinable]

So thats what AMD fanboys came to? Faked benchmarks

Not fanboys as such, more like some one who didn't do their benchmarks thoroughly.

[Majestic]

What could people possibly be arguing about?

 

It's 4k, of course their will be less difference. GPU is the bottleneck, m80s

 

Maybe you should read the actual argument then.

[Norman_the_Owl]

Maybe you should read the actual argument then.

Do you really think i'm gonna read 12 fucking pages of stupidity?

[Majestic]

Do you really think i'm gonna read 12 fucking pages of stupidity

 

Then don't enter the discussion. Mr project-a-lot.

 

http://linustechtips.com/main/topic/429688-amd-fx-8370-vs-intel-i7-5960x-gtx-970-sli-4k-benchmarks/page-4#entry5763917

http://linustechtips.com/main/topic/429688-amd-fx-8370-vs-intel-i7-5960x-gtx-970-sli-4k-benchmarks/page-6#entry5764806

[patrickjp93]

This tread needs to be closed. Not one person who oppose the findings has come up with enough credible evidence to disprove the benchmark findings. All I see are a bunch of opinions and no hard DATA.

My argument is strong enough to stand in forensic science. You don't need data when you can disprove the initial data with raw facts and strong logic. That is the essence of debate and informal logic (formal being mathematical, some of which I provided anyway in proof form).

[patrickjp93]

What could people possibly be arguing about?

 

It's 4k, of course their will be less difference. GPU is the bottleneck, m80s

We're arguing it's not possible the 5960X is getting lower minimum frame times on the same GPU setup all else being equal, and it's not. There's no room for that amount of error between DDR4's higher latency and the 5960X's lower clocks. The 5960X's IPC is far too high for this to be possible, yet the OP claims otherwise, and I claim he rigged the data, fabricated it, had an unstable X99 system, or had the biggest statistical outlier possible on both tests. Not to mention, his mathematical reasoning doesn't even stand the smell test, much less actual scrutiny by inductive proof.

[Dabombinable]

We're arguing it's not possible the 5960X is getting lower minimum frame times on the same GPU setup all else being equal, and it's not. There's no room for that amount of error between DDR4's higher latency and the 5960X's lower clocks. The 5960X's IPC is far too high for this to be possible, yet the OP claims otherwise, and I claim he rigged the data, fabricated it, had an unstable X99 system, or had the biggest statistical outlier possible on both tests. Not to mention, his mathematical reasoning doesn't even stand the smell test, much less actual scrutiny by inductive proof.

I'll put it this way, if Haswell's voltage is even a tad too low, performance starts dropping leading up to the full instability.

[Majestic]

Don't worry guys, with DX12 CPU won't matter anymore.

 

 

[Yoinkerman]

It's because of all the Intel fanboys.

"It reeks of bullshit because it is sponsored" Yet Linus has many sponsored by Intel videos but he has no hint of bias. (I am not accusing him of anything just using your logic)

Let me clear some things up to everyone here, AMD posted this to their Facebook. They sent us hardware and that is it. We have never received any pay from AMD or advertising and in no way shape or form did that "Taint" our results.

Our test benches is in the article itself if you take the time to read it. To many people this may not make sense that we aren't making it a CPU bound environment, but if you are a gamer you never want to be in a CPU bound environment. Please tell me any gamer who wants 400FPS that isn't a competitive CS:GO player or the like.

Anandtech and many other outlets put CPUs in synthetic situations that frankly put AMD in a much worse light for gamers than they deserve. The amount of people who assume you need an i5 to play dota because its Intel and "AMD is so much worse" shows how uneducated the masses are. This test is to educate people that you don't need to throw money at your CPU till your games run better.

If anyone here has some constructive criticism, Donny and myself will be looking through the posts

What we'd like to see is suggestions for games etc. Thanks and we're happy you've taken the time to read it.

BTW Sintzza, aren't you from the Teksyndcate forums as well?

Praise the dark Lord someone with a brain

[Majestic]

Praise the dark Lord someone with a brain

 

He starts his scentence with "it's because of all the Intel fanboys". Nothing that comes next is going to be particularly intelligent.