• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Seemingly inexplicable poor FPS in games

teppic said:
This thread has become really unnecessarily rude and unpleasant.

@Th3M8dH8tt3r - from what we were saying yesterday I think you understand that you wouldn't be getting a huge performance boost, and in the few games that only use 1 core, not very much.
Click to expand...

Yup, I've bought the 6350. I'm sure ill notice a gain but I don't think the difference between the 6350 and 8350 would be so noticeable.
 
Th3M8dH8tt3r said:
From what I was reading, the 63 has better single core performance which suits me does it not as I'm only using up to 4 cores for 99% of my games?
Click to expand...
I can understand what you mean, but I think Martini is merely talking from value point of view.

To clarify, if you are overclocking the Piledriver CPUs, then the default factory clock (high or low) would means very little (as their max overclock would be around the same). All of them are essentially the same CPU, but the difference is only on how many cores it got.

Going by OcUK's price:
A FX6300 cost £90
A FX6350 cost £108
A FX8320 cost £125

If it was me looking having a FX8150 and AM3+ board, I would thinking along the lines of:

"hm...FX6350 is just a FX6300 with higher clocked, and since I would definitely be overclocking beyond the 3.90GHz anyway, so the higher stock clock of 3.90GHz on the FX6350 is meaningless to me so it don't deserve me paying the extra money for. So between £90 FX6300 and a £125 FX8320 huh...? At the moment I'm only playing mmos and older games such as Killer Floor (which uses 1-2 cores), and have 8 cores instead of 6 cores won't benefit me at all, so why should I pay the £35 extra for the FX8320?...But then again, I may be playing light-threaded games now, but what about the future? At the moment I am have no plan on playing games such as BF and Crysis where 8 cores will have advantage over 6 cores and give better frame rate, but what about in a few months time when BF4 is out, or may be other future titles? May be I should pay the £35 extra for the two extra cores just in case".

"What about the Intel option? A new Haswell CPU and a Z87 socket 1150 board would cost around £270+, that's far more than what I'm willing to spend. What about 2nd hand Sandy i5 option? A 2nd hand i5 2500K only cost around £100 now, and I can get a new decent Gigabyte budget Z68 socket 1155 board for just £46. It will cost me just under £150 in total, and will give me much better performance in games that use 4 cores or less, and I can keep my AM3+ board and the FX8150 for spare PC build in the future, since I would probably not get much money back selling them.

Decisions, decisions".

The above is just one of the scenario I come up with, imaging I already have a AM3+ board rather than already have a i5. Hope it helps put things into perspective for you.

teppic said:
You'll notice a gain. Not many games only use 1 core, and ones that use 2-4 will get the most from the upgrade.
Click to expand...
Please please PLEASE stop pulling things out of thin-air....you don't know what you are talking about. Between the Bulldozer and Piledriver the extra performance only comes from the higher IPC and may be some of the instruction sets...how many cores a game uses, may it be 1 core or up to 6 cores would have nothing to do with it (if anything, for a game properly uses up to 8 cores, the Piledriver 6 cores would perform worse than the Bulldozer 8 cores when both are on the same clock speed.)
 
Last edited:
Marine-RX179 said:
Stop pulling things out of thin-air....you don't know what you are talking about. Between the Bulldozer and Piledriver the extra performance only comes from the higher IPC and may be some of the instruction sets...how many cores a game uses would have nothing to do with it (if anything, for a game uses up to 8 cores, the Piledriver 6 cores would perform worse than the Bulldozer 8 cores when both are on the same clock speed.)
Click to expand...

While I agree he's pulling things out of thin air, I don't understand the 1 core games would show less gain thing at all.

The FX8150/FX6300 gaming performance would depend purely on how much the engine can use the cores.
 
For 8 core games...I'll turn on my Xbox One. I don't really like FPS on PC because of the controller config. Yes I could get a controller but then I have to have the PC next to my TV, which it already is, but I just prefer having the console for those games...racing, shooters, sports, and the PC for RTS and sim.

I bought the 6350, knowing it's the same as the 6300, because in my head, AMD will pick the better quality chips to overclock stock. Thus, the lottery of getting a good overclocker odds are shortened. It may be insane logic but for £15, I'm happy. I don't see the point in another 8 core as my games don't play it so I'll take the £10 or £13, even though it's more because of my superstition, I'd have to get the 8350 which is £162, and go to the pub!!!

I was wrong about the 60% thing, I was under the impression that the speed was the sum total of all the cores. I was sure I'd seen a program window somewhere that listed the core speeds and they weren't 4.8ghz. NVM...I know better now.

And Martini, some of your comments have been very good but you have to cut through swathe of scathing comments and attitude to get to them. Why the attitude? If this thread annoys you then don't read it. If you're on some ego trip demeaning people who don't know and making out they're tards when they get something wrong or have read someone else on another forum saying something and it's wrong, sucks to be you man! Quit been argumentative. Forums are for discussing, not bitching and d waving.
 
Last edited:
I've cleaned up the end of this thread. Please keep it nice and stop with the point scoring off eachother.
 
Th3M8dH8tt3r said:
I was wrong about the 60% thing, I was under the impression that the speed was the sum total of all the cores. I was sure I'd seen a program window somewhere that listed the core speeds and they weren't 4.8ghz. NVM...I know better now.
Click to expand...

You probably saw the core 'speed' before multipliers - 480mhtz with a 10* multiplier = 4800mhtz.
 
I wouldn't worry, just see how it goes. If you run benchmarks now at 4.5GHz on your BD CPU, then compare at 5GHz on the 6300, you'll see improvements across the board in games. It could be as low as ~15% (taking into account the higher clock speed, IPC gains and additional CPU instructions), but for games that use more than one core it will definitely be higher than that.

As you saw in the Sim City bench, going from Piledriver 4 core to Piledriver 8 core at the exact same clock speed gave a 10% improvement (regardless of the rest). As you can also see here: http://www.hardwarecanucks.com/foru...615-amd-vishera-fx-6300-fx-4300-review-4.html - Civ 5 can make good use of multiple cores too.
 
teppic said:
I wouldn't worry, just see how it goes. If you run benchmarks now at 4.5GHz on your BD CPU, then compare at 5GHz on the 6300, you'll see improvements across the board in games. It could be as low as ~15% (taking into account the higher clock speed, IPC gains and additional CPU instructions), but for games that use more than one core it will definitely be higher than that.

As you saw in the Sim City bench, going from Piledriver 4 core to Piledriver 8 core at the exact same clock speed gave a 10% improvement (regardless of the rest). As you can also see here: http://www.hardwarecanucks.com/foru...615-amd-vishera-fx-6300-fx-4300-review-4.html - Civ 5 can make good use of multiple cores too.
Click to expand...

I'll run CineBench and record some games, average out FPS. Be a good experiment.
 
teppic said:
But for games that use more than one core it will definitely be higher than that.
Click to expand...

I don't understand that.

Lets say Bulldozer does 40 FPS in a single threaded game, the PD does 44, that's a 10% gain.

In a 2 threaded game, if the BD did double that at 80, wouldn't the PD do 88 ; 10%

4 threaded, BD at 160, PD at 176, that's still 10%?
 
teppic said:
I wouldn't worry, just see how it goes. If you run benchmarks now at 4.5GHz on your BD CPU, then compare at 5GHz on the 6300, you'll see improvements across the board in games. It could be as low as ~15% (taking into account the higher clock speed, IPC gains and additional CPU instructions), but for games that use more than one core it will definitely be higher than that.
Click to expand...
Still not making sense.
 
teppic said:
Because there's an increase in clock speed along with IPC along with additional CPU instructions.
Click to expand...

:confused:

If his PD with its enhanced IPC and core speed is 10% faster in single threaded games against the BD, why would it be even faster percentage wise in 2-4 threaded things? The answer isn't an increase is clocked speed with IPC and instruction sets.
 
Last edited:
Unfortunately teppic, I think you are now a little off. It'll be 10% per core increase so it's irrelevant if 1 or 6 cores are being utilised. If the program uses 8 cores, I'll probably see a drop BUT, my Xbox is for 8 core games. As, typically, the RTS and Sims aren't 8 core.
 
Th3M8dH8tt3r said:
Unfortunately teppic, I think you are now a little off. It'll be 10% per core increase so it's irrelevant if 1 or 6 cores are being utilised. If the program uses 8 cores, I'll probably see a drop BUT, my Xbox is for 8 core games. As, typically, the RTS and Sims aren't 8 core.
Click to expand...

When each core is being affected by three factors that change at once, the overall performance taking multiple cores into account won't be flat, since they all work together. It would be though if only one factor was changing.
 
You must log in or register to reply here.
Back
Top Bottom