Jump to content

AMD or Intel processor for BF4 build (ultra)

Okay there is something called Mantle. It is a low level API that can utilize all of the cores you throw at it. Also there is optimizations made specifically for AMD CPU's/GPU's. Typically the i5's, and the i7's are faster but with Mantle the 8350 will pull ahead. Not too mention Hyperthreading only produces a tangible benefit to your FPS, so going with an i5 seems more logical doesn't it? Not really since they only have four cores, and AMD has 8 cores in the 8350. Overclock it to 5 Ghz and it will beat an i5-3570K(Running at stock that is). What matters is the draw calls made, and when you have 8 cores instead of 4 cores working at executing draw calls as fast as possible, you gain an advantage. Even if the game does not support 8 cores, if the game uses Mantle, Mantle will utilize all 8 cores to assist the GPU with rendering graphics. Combine this with an AMD based GCN GPU and you have a real killer. The new cards are coming out soon. Not too mention it makes sense that graphical applications such as games have most of the burden placed upon the GPU not the CPU. So the 8350 will preform just as fast, unless you have an Intel Extreme edition 6 core.

Link to comment
Share on other sites

Link to post
Share on other sites

Here are the BETA benchmarks, the AMD CPUs outperform their intel counterparts in every price segment.

I see no reason to go for Intel especially considering the significantly higher price you'd have to pay.

http--www.gamegpu.ru-images-stories-Test

https://linustechtips.com/main/topic/61740-bf4-beta-benchmarks-released-amd-takes-cpu-performance-crown/

Activision announced that its future PC games including the upcoming call of duty will also support mantle.

Upcoming Ubisoft games including Watchdogs will also be optimized for the 8 core AMD CPUs.

 

The one thing that has mainly confused me with benchmarks such as this, the 9590 just being a cherry picked 8350 Oc'd to 4.7 with the turbo of 5 being pitted against a non-overclocked 3970x, of course the 9590 is still sorta classed as it's own CPU but I would've much preferred if these benchmarks were done with most if not everything overclocked, (with the exception of the i3's, etc). As many people would say if you own a 3970x and you aren't overclocking it, then son, you're doing it wrong, you get the point anyway. All in all, another set of these benches with all the CPU's that are able to overclock, overclocked, then we MAY have a different story.

Cpu: i5-2500k @4.8Ghz, MB: Asus Maximus V Formula, CPU cooler: Be quiet! Dark rock pro 2, GPU: Evga Gtx660 FTW@1.24ghz. Ram: Corsair Vengeance 8GB 1866Mhz, PSU: Be quiet! 730w Semi modular, SSD: Corsair force 3 240Gb, HDD: WD Green 1TB, Case: Nzxt H2 with 4 Corsair SP120's, Win7

Link to comment
Share on other sites

Link to post
Share on other sites

The one thing that has mainly confused me with benchmarks such as this, the 9590 just being a cherry picked 8350 Oc'd to 4.7 with the turbo of 5 being pitted against a non-overclocked 3970x, of course the 9590 is still sorta classed as it's own CPU but I would've much preferred if these benchmarks were done with most if not everything overclocked, (with the exception of the i3's, etc). As many people would say if you own a 3970x and you aren't overclocking it, then son, you're doing it wrong, you get the point anyway. All in all, another set of these benches with all the CPU's that are able to overclock, overclocked, then we MAY have a different story.

I agree, that's why we have Linus, he overclocks everything, that's why I love his benchmarks !

Link to comment
Share on other sites

Link to post
Share on other sites

Those benchmarks are false.

 

doesn't even matter... What I'm saying is that it is a good cheap processor that' ll give you great performance and the comment from that guy was total fanboy crap.

Link to comment
Share on other sites

Link to post
Share on other sites

I agree, that's why we have Linus, he overclocks everything, that's why I love his benchmarks !

Pretty much mate.

Cpu: i5-2500k @4.8Ghz, MB: Asus Maximus V Formula, CPU cooler: Be quiet! Dark rock pro 2, GPU: Evga Gtx660 FTW@1.24ghz. Ram: Corsair Vengeance 8GB 1866Mhz, PSU: Be quiet! 730w Semi modular, SSD: Corsair force 3 240Gb, HDD: WD Green 1TB, Case: Nzxt H2 with 4 Corsair SP120's, Win7

Link to comment
Share on other sites

Link to post
Share on other sites

Okay there is something called Mantle. It is a low level API that can utilize all of the cores you throw at it. Also there is optimizations made specifically for AMD CPU's/GPU's. Typically the i5's, and the i7's are faster but with Mantle the 8350 will pull ahead. Not too mention Hyperthreading only produces a tangible benefit to your FPS, so going with an i5 seems more logical doesn't it? Not really since they only have four cores, and AMD has 8 cores in the 8350. Overclock it to 5 Ghz and it will beat an i5-3570K(Running at stock that is). What matters is the draw calls made, and when you have 8 cores instead of 4 cores working at executing draw calls as fast as possible, you gain an advantage. Even if the game does not support 8 cores, if the game uses Mantle, Mantle will utilize all 8 cores to assist the GPU with rendering graphics. Combine this with an AMD based GCN GPU and you have a real killer. The new cards are coming out soon. Not too mention it makes sense that graphical applications such as games have most of the burden placed upon the GPU not the CPU. So the 8350 will preform just as fast, unless you have an Intel Extreme edition 6 core.

Yes. We all know that, it's why I'm considering AMD for this build instead of just going for a 4770K or something.

AMD FX 6300 | ASUS M5A97 R2.0 ATX AM3+ | HITACHI ULTRASTAR 1TB 7200RPM HDD | XFX RADEON HD 7950 3GB | 8GB DDR3-1600 MUSHKIN BLACKLINE | OCZ ZT 550W 80+ BRONZE PSU | ZALMAN Z5 PLUS CASE | LITE-ON iHDS118-04 DVD/CD DRIVE | BENQ GW2255 21.5" MONITOR


Link to comment
Share on other sites

Link to post
Share on other sites

CPU Intel I5-3470 4.0Ghz  Mobo  Asrock Z77 ExtremeGPU Asus HD 7870 2GB 1175/1450 Mhz Ram Corsair Vengeance 8 Gb 1600 Mhz HSF Cooler Master Hyper 212 Evo Case Fractal Desing Arc Midi R2 PSU Aerocool Strike-X 600w 80 + bronzer ( to be replaced ) HDD Western Digital Caviar Blue 1 TB SSD Kingston Vseries 60 GB 6GB/s

Link to comment
Share on other sites

Link to post
Share on other sites

Guys. I'm going to say this in huge font so maybe you will listen.















The Watch Dog requirements recommending an 8 core CPU were fake. The developer said on his twitter, in French, that the official ones would be coming out soon, and would not be as extreme as these. 


Hopefully you all saw it that time.
 

† Christian Member †

For my pertinent links to guides, reviews, and anything similar, go here, and look under the spoiler labeled such. A brief history of Unix and it's relation to OS X by Builder.

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

Another thing to might consider is mantle working with a 6 core intel processor. Perhaps the multithreaded capability, 12 cores including the virtual, will be a benefit. Then again I could be very very wrong. I just figured mantle will be amd gpu specific not amd cpu. No source, just speculation.

Sent from my SAMSUNG-SGH-I717 using Tapatalk 2

Link to comment
Share on other sites

Link to post
Share on other sites

Another thing to might consider is mantle working with a 6 core intel processor. Perhaps the multithreaded capability, 12 cores including the virtual, will be a benefit. Then again I could be very very wrong. I just figured mantle will be amd gpu specific not amd cpu. No source, just speculation.

Hyper threading won't help if the Core is at 100% load. It only makes a difference if the Core is being utilized, but not at 100%.

† Christian Member †

For my pertinent links to guides, reviews, and anything similar, go here, and look under the spoiler labeled such. A brief history of Unix and it's relation to OS X by Builder.

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

Hyper threading won't help if the Core is at 100% load. It only makes a difference if the Core is being utilized, but not at 100%.

You are completely incorrect with that. This is probably because you dont understand how hyperthreading works. You actaully get the most benefit from hypertheading when your pc is maxed out, its all about keeping the physical cores busy.

Link to comment
Share on other sites

Link to post
Share on other sites

 

 

Depends on the GPU you want to pair it with currently. I5/i7 will do the best work in my opinion, just look at the current drivers for Radeon and GeForce:
 
It's a particularly big problem with Battlefield 4's beta but this has been case for years now, GeForce works better with slower multicore CPUs due to multithreaded drivers, while Radeon 7970 needs a quick single thread.
 
CPUs + AMD Radeon HD 7970:
 
CPUs + Nvidia GeForce GTX 770:
 
Just look at FX 83xx! With GTX 770, at stock clocks, it pulls 7 (!!!) FPS more  than OVERCLOCKED FX 83xx with Radeon 7970!
 
OVERCLOCKED FX 83xx with GTX 770 gets 17 (!!!) FPS more than OVERCLOCKED FX 83xx with Radeon 7970!
 
We'll have to wait for Mantle before answering your questions but right now:
 
If you go AMD, go Nvidia GPU.
 
If you go Intel, you can choose either AMD or Nvidia GPU.

 

Thank god, someone that knows what there talking about.

Sponsored by Twitch and DxRacer.

Link to comment
Share on other sites

Link to post
Share on other sites

doesn't even matter... What I'm saying is that it is a good cheap processor that' ll give you great performance and the comment from that guy was total fanboy crap.

Fair enough.

Sponsored by Twitch and DxRacer.

Link to comment
Share on other sites

Link to post
Share on other sites

Guys. I'm going to say this in huge font so maybe you will listen.

The Watch Dog requirements recommending an 8 core CPU were fake. The developer said on his twitter, in French, that the official ones would be coming out soon, and would not be as extreme as these. 

Hopefully you all saw it that time.

 

I dont think most people believed the requirements. Even minimum settings were too extreme already, recommended hardly anyone cos play. Course they weren't serious.

AMD FX 6300 | ASUS M5A97 R2.0 ATX AM3+ | HITACHI ULTRASTAR 1TB 7200RPM HDD | XFX RADEON HD 7950 3GB | 8GB DDR3-1600 MUSHKIN BLACKLINE | OCZ ZT 550W 80+ BRONZE PSU | ZALMAN Z5 PLUS CASE | LITE-ON iHDS118-04 DVD/CD DRIVE | BENQ GW2255 21.5" MONITOR


Link to comment
Share on other sites

Link to post
Share on other sites

I am running a I7 at 3.2Ghz while playing BF4 and getting 85%+ and capping out at 100% quite often. Optimizations will help bring this down when it's released but this game seems to be really CPU heavy.

                                                                                              Sager NP9370EM - I7 3630QM - 680m 1045Mhz - 8gb 1600mhz ram - 240gb msata 750gb hdd

Link to comment
Share on other sites

Link to post
Share on other sites

You are completely incorrect with that. This is probably because you dont understand how hyperthreading works. You actaully get the most benefit from hypertheading when your pc is maxed out, its all about keeping the physical cores busy.

@Glenwing, you explained how HT works once to me. Based on that, I think what I said was correct. Am I right? Does it even matter?

† Christian Member †

For my pertinent links to guides, reviews, and anything similar, go here, and look under the spoiler labeled such. A brief history of Unix and it's relation to OS X by Builder.

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

@Glenwing, you explained how HT works once to me. Based on that, I think what I said was correct. Am I right? Does it even matter?

 

Ehhh whether the CPU is at 100% or not is a bit too broad of a situation to say whether HTT will be doing a lot or not much.

Link to comment
Share on other sites

Link to post
Share on other sites

Ehhh whether the CPU is at 100% or not is a bit too broad of a situation to say whether HTT will be doing a lot or not much.

Alright, 4770k running a video game (which uses lots of Floating Point operations obviously).

† Christian Member †

For my pertinent links to guides, reviews, and anything similar, go here, and look under the spoiler labeled such. A brief history of Unix and it's relation to OS X by Builder.

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

Alright, 4770k running a video game (which uses lots of Floating Point operations obviously).

 

Not much, if anything.

 

Hyperthreading is basically, each core is instructed to have a second thread ready, and whenever the main thread is stalled, it works on the secondary thread in the meantime.  Threads can be stalled by various things, like a cache miss, or a mishap in scheduling where a thread needs the result of another thread's calculation to continue and needs to wait for it to finish.  The biggest source of stalled threads and idle cores though is RAM operations; every time the CPU makes a request to the RAM, it takes several cycles for the response to come back, during which time a core without HTT would be sitting there, waiting for the information it needs to continue.  Prefetching mitigates this somewhat although even then, there is still latency even between the cores and the cache.  Programs heavy on RAM I/O usually see some nice improvement from HTT.  Games don't take much benefit from HTT because of that (they aren't RAM intensive) as well as other things; most don't use many cores at the moment as we all know, so the improvement is zero in that case, but even if they are optimized to use more cores and actually gain some benefit, it won't be much.  You can look at Pentium vs i3 benchmarks, there are plenty of games that we know can use at least 4 cores and even on those the fps difference rarely breaks single digits.  Then when you move up to i5, basically exchanging the i3's two hyperthreaded cores for physical ones, there's an enormous fps gap.

Link to comment
Share on other sites

Link to post
Share on other sites

Not much, if anything.

 

Hyperthreading is basically, each core is instructed to have a second thread ready, and whenever the main thread is stalled, it works on the secondary thread in the meantime.  Threads can be stalled by various things, like a cache miss, or a mishap in scheduling where a thread needs the result of another thread's calculation to continue and needs to wait for it to finish.  The biggest source of stalled threads and idle cores though is RAM operations; every time the CPU makes a request to the RAM, it takes several cycles for the response to come back, during which time a core without HTT would be sitting there, waiting for the information it needs to continue.  Prefetching mitigates this although there is still latency even between the cores and the cache.  Programs heavy on RAM I/O usually see some nice improvement from HTT.  Games don't take much benefit from HTT because of that (they aren't RAM intensive) as well as other things; most don't use many cores at the moment as we all know, so the improvement is zero in that case, but even if they are optimized to use more cores and actually gain some benefit, it won't be much.  You can look at Pentium vs i3 benchmarks, there are plenty of games that we know use 4 cores and even on those the fps difference is rarely breaks single digits.  Then when you move up to i5, basically exchanging the i3's two hyperthreaded cores for physical ones, there's a enormous fps gap.

Awesome, thanks for the detailed explanation. I thought that HT also had something to do specifically with Floating Point operations and Integer Operations being able to be done at the same time (effectively doing 2 operations in a single cycle), but I guess I'm wrong in that. 

Still, my comment before now makes sense I think.

HT's whole point is to not allow cores to waste cycles doing nothing. So it enables them to actually be 100% utilized. I was thinking of it in the sense that 100% utilization meant that every core was doing operations as fast as it possibly could, without HT. So since there aren't any free cycles to give out to threads, HT isn't being used at all and therefore doesn't have a benefit, like when playing video games.

However, a CPU without HT being utilized 100% (where no cycles are wasted) is probably only possible in situations like you mentioned. Where programs that are not RAM intensive are used. While the only way to get CPUs to 100% utilization when using RAM intensive programs (or some other way to increase CL/wasted cycles) is with HT, like when rendering a video, or something else super RAM intensive. 

So in a sense, I'm right, but in a sense I'm not. Cool. Good to know what HT is actually good for. 

† Christian Member †

For my pertinent links to guides, reviews, and anything similar, go here, and look under the spoiler labeled such. A brief history of Unix and it's relation to OS X by Builder.

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

Awesome, thanks for the detailed explanation. I thought that HT also had something to do specifically with Floating Point operations and Integer Operations being able to be done at the same time (effectively doing 2 operations in a single cycle), but I guess I'm wrong in that. 

Still, my comment before now makes sense I think.

HT's whole point is to not allow cores to waste cycles doing nothing. So it enables them to actually be 100% utilized. I was thinking of it in the sense that 100% utilization meant that every core was doing operations as fast as it possibly could, without HT. So since there aren't any free cycles to give out to threads, HT isn't being used at all and therefore doesn't have a benefit, like when playing video games.

However, a CPU without HT being utilized 100% (where no cycles are wasted) is probably only possible in situations like you mentioned. Where programs that are not RAM intensive are used. While the only way to get CPUs to 100% utilization when using RAM intensive programs (or some other way to increase CL/wasted cycles) is with HT, like when rendering a video, or something else super RAM intensive. 

So in a sense, I'm right, but in a sense I'm not. Cool. Good to know what HT is actually good for. 

 

Right.  Efficiency and performance decrease with RAM intensity because it causes unused CPU executions so in a sense you can think of HTT as a way for RAM intensive programs to reclaim that lost performance.  Whereas games don't really lose much to begin with.

Link to comment
Share on other sites

Link to post
Share on other sites

The GPU need to be the main factor for BF4 if you want any game play above 1600x900 with high settings.

A water-cooled mid-tier gaming PC.

Link to comment
Share on other sites

Link to post
Share on other sites

So virtual cores will not be used for optimization in an api like mantle? Or is that to much of an assumption.

Link to comment
Share on other sites

Link to post
Share on other sites

Probably not, but APIs, what they do and don't do, go beyond the scope of my expertise really.  Anything is possible... :P The virtual core usage really just depends on how the programs is using system resources.  Maybe they can shift how games do that.  I don't know.

Link to comment
Share on other sites

Link to post
Share on other sites

So virtual cores will not be used for optimization in an api like mantle? Or is that to much of an assumption.

 

No, Physical cores is what it is going after. If you are running XFire 290x than IMO your best bet would be an 3930k 6 core from intel if your playing games like BF4 as your main game.

                                                                                              Sager NP9370EM - I7 3630QM - 680m 1045Mhz - 8gb 1600mhz ram - 240gb msata 750gb hdd

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×