CPU gaming differences

Toggle sidebar Toggle sidebar
Sort by date Sort by votes
The fx 8350 and 8320 will perform comparably with the 4670K in gaming except mmos where intel trashes them with a single gpu. If you plan on crossfire or sli you need an intel processor to properly push the cards
 
Upvote 0 Downvote


Pretty sure you could have read through one of the ~50 threads already in this forum with people arguing about this. Then at least you could have found out that your question has no answer - It is highly debabable which is best and it is highly irrelevant which you pick performance wise they are comparable.
 
Upvote 0 Downvote


Fx-8350 is 4 physical and 8 virtual cores.
 
Upvote 0 Downvote


You can run a CF/SLI setup on the 8320/8350...There is a video out there of a guy running Metro 2033 on 4 HD 7970s in CF with an 8350, and CPU usage is at 30-40% the entire time.

The 8350 does not bottleneck any GPU in existence in this generation of GPUs, and likely won't next generation either since it doesn't bottleneck a GTX 690 or HD 7990, or a 4 way CF of HD 7970 GHz editions.
 
Upvote 0 Downvote


No... Hyperthreading on I7s are virtual cores. AMDs multiple cores share SOME of the less used resources, its still a physical 8 core CPU that behaves LIKE a REAL 8 core CPU.

BUT Intel is the best gaming solution if you are not on a tight budget.
The reason why Intel is generally better is that Intel has much better IPC, so a CPU with 1 core clocked at lets say 3GHz from Intel is MUCH faster than a current AMD CPU with 1 core also clocked at 3GHz due to better scheduling among other reasons.
 
Upvote 0 Downvote


It does in select games that are not well multithreaded and CPU intensive like SC2. But its very rare and can be remedied by a mild OC.
 
Upvote 0 Downvote


Okay thanks but performance per core of amd is crappy
 
Upvote 0 Downvote


I didnt say you couldnt run an SLI/Xfire setup on the FX chips. I said you'll get better FPS with running one on an Intel Setup.

Toms test. 7970's in Crossfire

BF3MediumCPUBottleneck2013.png

SkyrimHiCPUBottleneck2013.png

F12012UltraCPUBottleneck2013.png


Now if an 8350 at 4.4Ghz cant keeep up with a 3770K at 4.4Ghz with 2 7970's. Imagine how far behind it is with 4
 
Upvote 0 Downvote
Intel God, actually since there is no bottleneck, the AMD will do BETTER at 4 way SLI/Xfire since it provides a hell of a lot more PCI-e bandwidth. You would need a X79/2011 platform to get similar PCI-e bandwidth, or the comparison makes no sense.
Stop making assumptions and get your facts right.
 
Upvote 0 Downvote
http://www.overclock.net/t/1318995/official-fx-8320-fx-8350-vishera-owners-club/6650#post_18997642

LL


LL


LL


There's a 4 way CFX with HD 7970 GHz Editions.

A follow up to an earlier post about Vishera and multi GPU use. The question has been raised by many members and those dropping in considering the FX-83XX for a multi GPU setup. (and they usually get bad info by someone using an Intel CPU) So I wanted to set the record straight on this.
You will notice about 28% CPU use and all four of the GPU's are being fully utilized.
This is at a standard 1080P resolution not an 'Ultra high res' The Vishera with multi GPU's is as fast or faster (in this case faster than a heavily OC'd i7 3960X with 4 x GTX 680 quad SLI)
The same usage characteristics hold true for games in general dispelling the myth of Vishera being a bottleneck in multi-GPU setups.
I have now graphed 26 modern titles with the same result and I will make them available upon request.
If I can be of help or answer questions about FX and multi-GPU setups, let me know.
Click to expand...

Username is: Red1776

His FX 8350 is watercooled and OC'ed
 
Upvote 0 Downvote


:lol:

Wow that's a pretty hilarious statement. Proof? I just proved that a 3770K will perform better then an 8350 with 2 cards yet you think an amd setup will perform better due to more lanes? What about the Z87 boards with PLX chips that can do X8 across all the lanes? You might want to get your facts straight before throwing that bad info out again.




http://www.overclock.net/t/1235557/official-top-30-heaven-benchmark-4-0-scores

Heaven benchmark thread. Look at the top 30 graph. See reds name? He pulls 141.9fps with a 8350. Now look above him. See King4x4? He's running a 3770K and pulls 148.1fps with a 7950 quadfire setup. Still want to keep saying an 8350 doesnt bottleneck 4 way 7970's? :lol: How could a 3770K with quad 7950's possibly pull better scores then a quad 7970 setup? Easy. Major cpu bottleneck with the amd setup

Both of you need to stop with the amd bias and get your facts straight. Intel > Amd everyday of the week in multi card setups
 
Upvote 0 Downvote


That's not a bottleneck...I don't think you understand the definition of bottleneck.

Let me elaborate...I am a game developer and I am intimately familiar with what a bottleneck actually is. You're describing "variance" which is a different result entirely.

A CPU is a bottleneck of a GPU when the CPU usage is greater than 95% and the GPU usage is less than 95%.

In the case listed above, I just showed you his screen shot with 28% CPU usage with 4x HD 7970s. That is not a bottleneck, especially considering all 4 7970s were running over 95% utilization.

A difference of 7 FPS is not a bottleneck, it's a variance.

EDIT: Additionally, his FPS is higher than similar Intel builds in some benchmarks posted. So by cherry picking a specific result...you are not negating the facts...you are simply picking a result that happened to show Intel ahead under the given circumstances of the test.

We don't know if the HD 7950s on the Intel rig are overclocked, (since it is an overclocking forum it's very likely that's the case)...which would also skew performance since his HD 7970s are not overclocked only his CPU.
 
Upvote 0 Downvote


PLX solutions are not perfect. And how in the world does less than 9FPS difference quantify to a "major" CPU bottleneck... Those few FPS could be attribted to slow PER CORE PERFORMANCE on the AMD, because a CPU bottleneck would cause a MUCH larger difference in FPS than a mere what? 4%?

EDIT : PLX bridges actually increase latency, worsening microstutter, but the PLX on boards add a hell of a lot more cost to a motherboard. You are comparing a board/CPU solution that is close to double the cost of a 8350 and a decent 990FX board.
Your point is moot and your fanboyism is OBVIOUS! This from someone that uses Intel.
 
Upvote 0 Downvote


:lol: Wow look at the excuses keep coming. A variance huh? Variances dont exist when the same test settings are ran by everyone. The 7950 setup has 1024 less stream processors then the 7970 setup yet still loses by 7fps. :lol: You're my new favorite amd fanboy :lol: There's no excuse you can make that will make anyone think you're right
 
Upvote 0 Downvote


No, they can't be running the same benchmarks on the same hardware...

1.) Intel and AMD CPUs and MBs cannot be the same for the test.

2.) The hardware from 2 different individuals would be different and not the exact same component even if they are the same make and model of component excluding CPU and MB. Because manufacturers have "variance" inside their components themselves. You witness this constantly with certain people achieving better/worse overclocks on the same model of CPU from the same manufacturer.

3.) Even tests with the exact same components will not reproduce exactly duplicate results, this is called "Margin of Error" and in hardware where everything is the same for 2 tests you see a minimum 2% variance or Margin of Error.

4.) Per different component you introduce roughly 1-2% greater variance if the components are not exactly the same. (i.e. same serial number to the letter). So, even if you just changed CPU + MB to run the tests you would introduce a MoE of roughly 5% into the test because of the changed components.

5.) You would already know this if you had ever had any statistical analysis background, but you clearly don't.

I am no fanboy...but I know what I am talking about...while you clearly think yourself infallible, even though you have less knowledge regarding the subject at hand. You are clearly not infallible.
 
Upvote 0 Downvote


Great, you just confirmed that I can disregard anything you contribute in the future. Makes things sooo much easier...
 
Upvote 0 Downvote




:lol: Thats perfectly fine by me considering you dont know anything about processors, or graphics cards, or pcie lanes, or scaling or anything in general but dont worry. I'll be watching out for you and cleaning up your mess of bad information
 
Upvote 0 Downvote


No one messes with intel god! LOL!
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom