GTA V CPU benches - AMD gets hammered (again)

[Ramses]

Well the pentium III isnt Intel's current top of the line processor, so that is hardly a relevant analogy. The FX, while it may be old, is the best AMD has to offer. So while I agree the title of the thread is a bit harsh, being old is no excuse for the FX, in fact it is more of an indictment of AMD that they have lagged so far behind.

I strongly disagree with that. If AMD never makes another desktop CPU will we be comparing the FX to whatever intel has in 2025 and being astounded all anew at how badly it performs in comparison? Where does one draw the line?
I say that line is at product line updates, intel has had what now two? AMD isn't competing, I don't think it's fair to drag the poor things through the muck for no reason. The results speak for themselves. I understand CPU land is pretty barren and boring these days and some junk talking might be better than standing around saying how slow a old dual core something is, but I think we as a community should be better.

 

[TurtleCrusher]

GTA5 runs better on high settings than GTA4 on the same hardware. That's piss-poor coding on the former game, while the current performance is a head scratcher. I get it that it's a port of a port, but still. This also isn't a shooter of any kind. 40fps min and an average of near 60? Hammered I think not. Did you all forget we played GTA4 at 20-30fps even on high end hardware when it was released?

The 8310/20/50/70 hits just over $100 at least once a month and they're getting bested by 15% by newer CPUs that cost 2-3x. Saving $150 and playing on high settings is a compromise all but a few snobs have no problem doing.

High instead of very high on a setting or two may make any of these 4 thread or better CPUs 60+ fps or better.

 

[Abwx]

GTA5 runs better on high settings than GTA4 on the same hardware. That's piss-poor coding on the former game, while the current performance is a head scratcher. I get it that it's a port of a port, but still. This also isn't a shooter of any kind. 40fps min and an average of near 60? Hammered I think not. Did you all forget we played GTA4 at 20-30fps even on high end hardware when it was released?

The 8310/20/50/70 hits just over $100 at least once a month and they're getting bested by 15% by newer CPUs that cost 2-3x. Saving $150 and playing on high settings is a compromise all but a few snobs have no problem doing.

High instead of very high on a setting or two may make any of these 4 thread or better CPUs 60+ fps or better.

Hey, dont bring the Fps/$ on the equation, that would completely destroy the purpose of the thread since in this matter Intel s CPU are not only hammered but litteraly reduced to nanometric thick plates, according to some opinions AMD chips should perform the same at half the price to be valuable buys....

 

[TurtleCrusher]

Also, look at what video cards are needed for 60fps minimum. Have fun paying for a Titan. Unless you dropped an unreasonable amount of money, you're much more likely top be GPU limited than CPU limited across the board. On very high settings your wallet gets hammered if you need 60fps minimum.

 

[frozentundra123456]

Not really. The results for amd are even worse compared to Intel in the PC lab test using a GTX 970. So you hardly need a titan to see better performance from Intel.

 

[greatnoob]

Just a heads up, an open-world game like this likely uses worker threads which means any one core can be utilised to its max.

The more worker threads -> more cores being used and/or more load on a core. This means any worker thread can be run on any core. So there could be 10 worker threads running on just one core rather than 2 on each core (it all depends on the work dispatcher and OS). No use trying to find a trend when it doesn't exist!

 

[Abwx]

Not really. The results for amd are even worse compared to Intel in the PC lab test using a GTX 970. So you hardly need a titan to see better performance from Intel.

Can i use a PClab style site to display the numbers that suits me..?.

Because this site is specialized in the lowest possible scores when it comes to AMD, for whatever game or bench generaly...

 

[TuxDave]

Can i use a PClab style site to display the numbers that suits me..?.

Because this site is specialized in the lowest possible scores when it comes to AMD, for whatever game or bench generaly...

You must not have been around in the P4 days... Intel reviews were pretty brutal on this forums.

 

[TurtleCrusher]

Not really. The results for amd are even worse compared to Intel in the PC lab test using a GTX 970. So you hardly need a titan to see better performance from Intel.

A $330 GPU for your $135 CPU (8320E). Makes sense.

 

[erunion]

Also, look at what video cards are needed for 60fps minimum. Have fun paying for a Titan. Unless you dropped an unreasonable amount of money, you're much more likely top be GPU limited than CPU limited across the board. On very high settings your wallet gets hammered if you need 60fps minimum.

Even a rather pedestrian GPU can achieve 60fps. You just use the appropriate graphics preset.

But CPU scaling isn't so easy.

 

[frozentundra123456]

A $330 GPU for your $135 CPU (8320E). Makes sense.

Wow, thought I saw the goalposts, but they shifted again. Maybe we should all use HD6450s so the FX would look equal to intel.

 

[TurtleCrusher]

Wow, thought I saw the goalposts, but they shifted again. Maybe we should all use HD6450s so the FX would look equal to intel.

Would you pair an i3 with a 980/Titan? Goalposts didn't move at all. No one would build a system like that.

 

[soccerballtux]

When looking at the CPu usage the 2600K is loaded at 72.5% for the main cores while the 8350 is at 61%, the 20% higher loading of the former translate in 20% higher Fps, how surprising that a CPU that is underutilised will perform not as well as a one wich is better maxed out....

its likely more gets done on the main 4 threads and the secondaries have less capacity hyper threaded as they are and load up faster than the dual core modules which have slower single thread capacity but higher multithread capacity

 

[TurtleCrusher]

its likely more gets done on the main 4 threads and the secondaries have less capacity and load up faster

I agree with this. If accounting for higher IPC this is plausible.

 

[Abwx]

I agree with this. If accounting for higher IPC this is plausible.

IPC has nothing to do with it, to summarize the FX8350 at 61% of its (integer) throughput is on par with a 2600K that is at 72% of its throughput, so the two CPUs are providing the same amount of computation yet the results are largely different.

Of course this imply that the FX has 20% higher integer throughput than a 2600K, wich is the case.

 

[frozentundra123456]

I strongly disagree with that. If AMD never makes another desktop CPU will we be comparing the FX to whatever intel has in 2025 and being astounded all anew at how badly it performs in comparison? Where does one draw the line?
I say that line is at product line updates, intel has had what now two? AMD isn't competing, I don't think it's fair to drag the poor things through the muck for no reason. The results speak for themselves. I understand CPU land is pretty barren and boring these days and some junk talking might be better than standing around saying how slow a old dual core something is, but I think we as a community should be better.

And I just as strongly disagree with you. AMD doesnt get a free pass because their processor is old, especially since we have been hearing on these forums since before the current gen consoles came out how "future proof" the FX is and how when games started to use more cores, FX was suddenly going to exhibt this marvelous performance in gaming.

 

[monstercameron]

And I just as strongly disagree with you. AMD doesnt get a free pass because their processor is old, especially since we have been hearing on these forums since before the current gen consoles came out how "future proof" the FX is and how when games started to use more cores, FX was suddenly going to exhibt this marvelous performance in gaming.

maybe you could be upset after we the fx performs poorly in a dx12 title...yes,yes I know it is always later...

 

[AMD64Blondie]

Thank god I have a Intel Core i5 4670.

(ditched my old AMD FX-8150.)

 

[StinkyPinky]

The lack of hyperthreading murders those pentiums.

Just goes to show that while most people focus on GPU's, CPU's still matter a great deal.

 

[Phynaz]

Can i use a PClab style site to display the numbers that suits me..?.

Because this site is specialized in the lowest possible scores when it comes to AMD, for whatever game or bench generaly...

As per usual, it's never AMD. Always somebody else's fault.

 

[Abwx]

Thank god I have a Intel Core i5 4670.

(ditched my old AMD FX-8150.)

You will be good in games only, despite being more recent the 4670 has lower integer throughput, and notably, than a FX8150 of old, while being completely obliterated by a FX8350...

The lack of hyperthreading murders those pentiums.

Just goes to show that while most people focus on GPU's, CPU's still matter a great deal.

It s not like people were unaware, myself i pointed unrelentlessly the poor value of thoses pentiums (actualy the 3258) even in games, seems that this one did bring the definitive evidence for thoses who were still skeptical...

 

[Phynaz]

maybe you could be upset after we the fx performs poorly in a dx12 title...yes,yes I know it is always later...

Who is this "we"?

 

[CHADBOGA]

IPC has nothing to do with it, to summarize the FX8350 at 61% of its (integer) throughput is on par with a 2600K that is at 72% of its throughput, so the two CPUs are providing the same amount of computation yet the results are largely different.

Of course this imply that the FX has 20% higher integer throughput than a 2600K, wich is the case.

So does this mean that even when AMD loses, they actually are the real winners?

 

[TurtleCrusher]

IPC has nothing to do with it, to summarize the FX8350 at 61% of its (integer) throughput is on par with a 2600K that is at 72% of its throughput, so the two CPUs are providing the same amount of computation yet the results are largely different.

Of course this imply that the FX has 20% higher integer throughput than a 2600K, wich is the case.

Games aren't just integer performance or else FX would be right up there with Haswell i7s. FX is an interger moster. I'm talking about instructions per clock, not integer performance clock.

It's not absurd to think the 2600k does more with 69% across all 8 threads (not 71%) than almost 5 cores of an FX-8350 (61%). Likely the code is being held back from a single thread that branches off and assigns other tasks. Core one has a 77% load and the other 23% is spread across the "secondary" cores of the modules, but is still held back from the performance of one core. Mantle/DX12 couldn't come soon enough for titles like this.

 

[Abwx]

As per usual, it's never AMD. Always somebody else's fault.

So does this mean that even when AMD loses, they actually are the real winners?

No need of two posts to quote what are twin straws....

So set apart putting words into other people mouths what could be your contribution to this debate..?..