Will the FX bottleneck?

Tazz4795

Reputable
Apr 19, 2014
84
0
4,630
Hi, i need to know if I would see any noticeable bottle neck from an FX 6350 OC'd at 4.5ghz if i were to upgrade to a GTX 970? I play mostly AAA titles including AC4 black flag, BF4, Watch Dogs, and more.

The reason I want to upgrade from my gtx 770 is because it only has 2gb of vram and i see alot of newer games will need 4gb+. Also the features that come with the 900 series intrigue me. I will be purchasing soon so i need to know thanks!
 
Solution
You will see a VERY big increase in performance. I would certainly go for it.

Your bottleneck will not be NEARLY as bad or noticeable as people are leading you to believe.

For craps sake, I used a 965BE with SLI 660tis (about as powerful as a 780) and I had next to ZERO visible bottleneck in 90% of games.

You will be perfectly fine.
Some CONFUSION here:

Your bottleneck will VARY quite a bit by the game. Some will have none, and some will be up to 40% roughly such as Skyrim, Starcraft 2 and other programs that aren't well threaded.

There's plenty of proof. I'm a little frustrated that people just keep stating no bottleneck will happen with the FX CPU's when they don't know what they are talking about:

Going forward Mantle and DX12 (and well optimized DX11) games will use your FX CPU much better so bottlenecks should drop off a lot for those games.

4GB:
The immediate need of this is questionable (Watch Dogs aside which is mostly a programming issue) but the GTX970 is a great card.

I would suggest the MSI TwinFrozr V version. I've reviewed every card and it IMO has the best combination of fan control for noise, overclocking capability and quality.

The Gigabyte card and Asus Strix cards are nice as well but the Gigabyte won't throttle it's fans down as much AFAIK, and the Asus Strix can't get as high of an overclock.
 
You cannot claim Skyrim will bottleneck when it is easily locked at 60fps by an FX CPU. Yes an intel CPU will run it better (maybe 100fps vs 80fps), but the FX will still lock it at 60fps. I do not consider this type of bottleneck ANY kind of issue as you CANNOT EVEN SEE IT.

I had a 965BE build up until recently and it locked Skyrim at 60fps at 1080p with graphical mods easily. Never dipped.
 


Yes and No.
1) You will drop frame rates depending on the MODS you use.

2) Not everyone locks to 60FPS if they have a better monitor that can do 120Hz or 144Hz.

3) You absolutely will get occasional dips below 60FPS due to the CPU. That's well tested. It may not be frequent or obvious to most though.

4) Not my main point anyway. I'm talking in GENERAL for performance.

Probably not worth a further discussion.

It's pretty close to 60 FPS before AA and mods though (FX-8350 at 4GHz):
http://www.bit-tech.net/hardware/2012/11/06/amd-fx-8350-review/6

It works out to 38% average FPS difference between the i5-3570K and FX-8350 @4GHz. Again, let's not get hung up on the 60FPS aspect.
 
Thank you for all your replies so fast! I know there will be some bottleneck in certain games, but for the vast majority i will see a performance increase from my gtx 770 right? Btw i game at 1080p 60hz. I dont heavily mod skyrim or play starcraft all too much but i do play MMOs and i can see how my cpu bottlenecks me there for sure.
 
You will see a VERY big increase in performance. I would certainly go for it.

Your bottleneck will not be NEARLY as bad or noticeable as people are leading you to believe.

For craps sake, I used a 965BE with SLI 660tis (about as powerful as a 780) and I had next to ZERO visible bottleneck in 90% of games.

You will be perfectly fine.
 
Solution


You meant GTX 770 correct?

You can use THIS as a guideline: http://www.techpowerup.com/reviews/ASUS/GTX_970_STRIX_OC/25.html

It's difficult to say how much advantage the GTX970 would be for some MMO's since they might vary a lot in how well they use your CPU.

For example, in Guild Wars 2: http://www.tomshardware.com/reviews/guild-wars-2-performance-benchmark,3268-7.html

One frustrating thing is that many reviewers use LOW RESOLUTIONS to emphasize the difference between CPU's because higher resolutions tend to shift the bottleneck towards the graphics card.

That's fine, but if they don't also run tests at 1920x1080 we don't know how much performance difference there is between CPU's due to bottlenecking.