Were the AMD fx processors still competitive against haswell in price and performance?
I'm not talking about just gaming.
It's hard to beat an core processor for $149.
They definitely seem to be holding their own against Intel.
>>52027905
So are these benches legit? Can AMD outperform an I7 in some tasks?
Is there an Intel processor within a similar price range that can compete?
How close would the 4460 get?
>>52027893
In most very heavily threaded workloads the 4 module/8 thread Vishera chips perform pretty well against a mainstream i7.
The only caveat is that they lose out in FPU heavy workloads, as well as serial workloads.
>>52027893
factor in the electricity costs of that 220W chip.
ask your mommy how much it costs.
the 8370/8350 is a way better deal and it has a tdp of 95/85watt instead of 200watt.
but how are the single core benchmarks?
Btw. i recently pulled out my 1090T 6 core and my asus crosshair iv mobo, still works godly.
coupled it with a 290x and i played battlefront with real life mod with Ultra settings and was getting close to the same performance as my 4670k @ 4.4ghz
>>52029037
8320 is a 200w chip?
Are you okay anon?
>>52029037
funny thing is, not enough in the processor's life time if it was running at 220W the whole fucking time to make up the difference in amd vs intel, power consumption is a non issue in real countries.
>>52029069
single core is where the cpu dies, intel is at least twice as powerful there.
also, battlefront is on an engine that properly uses 6 cores (i think also 8) cores, but its the only engine that does that.
>there is a 1 fps difference between an fx 8350 and an i7 3960x
reminder
>>52029712
>comparing on 2 different OSes
>cherrypicking highest and lowest scores
>macbook = ultrabook?
Yeah that's some pretty shitty bait you got there.
>guaranteed replies
And I fell for it.
>>52029753
>comparing on 2 different OSes
OS X usually gets lower scores over Winblows
> cherrypicking highest and lowest scores
both highest scores
> macbook = ultrabook
yeah, it's thin as fuck and uses a 45W CPU (that AMD cpu is over 100W)
not even b8 m8.
>>52029778
Yeah, no.
Fuck off.
>>52029778
That image needs to die, and so do you.
If you know what you are doing generally vishera is highly competitive against haswell in many areas, not really until skylake does intel become flat superior 24/7.
>>52029653
>also, battlefront is on an engine that properly uses 6 cores (i think also 8) cores, but its the only engine that does that.
Its frostbite which powers just about every major EA game now and will do so in the future.
Seems the latest CoD game can also into threads.
>>52027893
If you're gaming you're better off with fewer Intel cores for most games, but in newer games Vishera will hold it's own against haswell.
If you're rendering or anything that makes use of all 8 threads then AMD is what you want. At that price nothing can beat it.
It's priced against an i5 but outperforms an i7.
Fuck hyperthreading.
>>52029957
>tomb raider
As much as I agree with your assesment don't use tomb raider in any cpu related fashion. The game is infamously cpu agnostic which (while a good thing) doesn't really paint a picture of relative performance of the cpus.
>>52027893
The cheapest skylake chips will soon be overclockable so even ZEN is DOA.
>>52029992
You're clueless.
The game isn't "CPU agnostic", the results are showing a huge GPU bottleneck.
>>52029992
That's why usually they recommend a stronger gpu with a weaker cpu for budget builds, isn't it?
Games tend to be more gpu limited for the most part, with a few exceptions like civ.
>>52029678
Now go play cities skylines and get back to us.
>>52030004
So you are saying a 7970 - the most powerful gpu at the time of the benchmark - is massively crippling every cpu on that chart? Plus you WANT a gpu bottleneck so you can simply add more gpu power to get better performance.
>>52030018
In a nutshell, yes.
>>52030058
>- is massively crippling every cpu on that chart?
Yes, you tech illiterate retard. If there was a sufficiently powerful GPU in the system then you would expose CPU performance despite the resolution. It would show the same CPU scaling you'd see at lowest res and settings.
This does not mean a game or engine is "CPU agnostic." That statement was 100% ignorant nonsense from a tech illiterate dipshit who should stay in /v/ with all the other children.
>>52030092
oh shit nigger you are right, a 690 is a bottleneck.
>>52030113
Yes, it is. The game is not locked to 108fps, you tech illiterate retard.
It is very heavily GPU bottlenecked.
>>52030201
Given these guys had to go to 1366x768 on low settings to show a difference in cpu for all intents and purposes yes, the game is cpu agnostic. If you need to run at over 300fps to show a cpu bottleneck you realistically don't have one.
http://www.gamingbench.com/benchmarks/tomb-raider-performance/
>>52030303
The game uses heavily taxing compute functions.
It is heavily GPU bottlenecked.
You're a tech illiterate retard so fucking clueless you don't even know how to interpret simple bench results. No amount of your childish bullshit will change the facts.
>>52030343
I understand the results just fine - I am also aware of real world usage. Until you start going to potato resolution with monstrously powerful gpus (be it single card or multiple) cpu is basically irrelevant to the game. The how or why this scenario comes into affect doesn't matter.
>>52030526
Why are AMD processors better than Intel at 4k?
Tweaktown did a review which showed the same iirc.
>>52030853
They aren't - 4k is just so gpu bound the differences in cpu are mostly irrelevant.
>>52030938
http://www.tweaktown.com/tweakipedia/56/amd-fx-8350-powering-gtx-780-sli-vs-gtx-980-sli-at-4k/index.html
It does look a tiny bit faster.
>>52030996
Mosto f the differences are within margin of error. What that test does tell you is 1) fx chips can feed multiple gpus just fine 2) 4k gives no fucks about cpu and 3) /g/ is wrong when it thinks you need a high end cpu for 4k gaming.
See also: tweaktown using a 9590 for 4k testing for a while and all the experts on /g/ declaring the tests null and void because as we all know fx chips bottleneck all gpus.
>>52029884
from that benchmark i would say that it doesn't use 8 cores and maxes out performance gains on 4, with ghz mattering more, though i would need to see core use benchmarks to confirm that, the small jump from the 4770 to the 3970 is kind of proof of that.
and like i said, one engine, sure ea uses it as a backbone for most games now, but you have to like ea games to really get much out of it, and i have to say, the list that uses it is fairly small, by the time some of those games come out, dx12 or vulcan would introduce better cpu performance then that engine alone.
>>52029678
take a look below that, the i7 920 is right there too. the 8350 is 50% stronger than it in single core and almost 100% stronger multicore, so i want to question where the bottleneck is there.
>>52030938
i would argue minimum is more important than average when it comes to 4k performance right now. i would rather have the game hit higher lows than have higher average.
>>52031027
the only thing you need for 4k is a cpu that can go to 60 fps unless the game does something weird like enabling denser effects the higher the resolution is.
>>52030003
the cheapest skylakes are dual core, the cheapest zen on arrival is true 8 core 16 threads
if amds cpus at the very least have more performance than their current newest mobile ipc, intel wont have an answer for amd that costs below 400$, and no answer in the current product line that is out now, if amd is right with the performance they get, intel wont have an answer preformance wise under 1000$, and amd is aiming for mainstream, their first run zen will be the same as a server chip, so no gpu space, its all dedicated cpu.
good for intel fanboys because they will be forced to stop jewing you.
good for amd because they can stay afloat,
good for everyone because intel will have to get off their dicks and compete again.
>>52031027
I never understood where the "AND processors are shit" meme came from.
Is it because backwards developers like bethesda release poorly threaded dx9 games that don't run as well on AMD?
>i would argue minimum is more important than average when it comes to 4k performance right now. i would rather have the game hit higher lows than have higher average.
Very true, but these 4k charts at least are showing minimums on the AMD chips to be on par with the Intel chips, so clearly in this scenario the AMD chips aren't performing any worse.
>>52031143
That and trying to use instructions badly on fx chips. Correct me if i'm wrong but x87 runs faster on fx chips than sse.
