Sign in with
Sign up | Sign in
Your question

Why are the amd fx 6300 and fx 8300 worse for gaming than an i3 or i5?

Last response: in CPUs
Share

What CPU would you chose for gaming?

Total: 29 votes

  • i5 3570k
  • 64 %
  • i3 3220
  • 0 %
  • fx 6300
  • 7 %
  • fx 8300
  • 31 %
September 17, 2013 4:07:06 PM

I have been reading around lately and noticed people saying that the i3 and i5's are better for gaming than the 6 core fx 6300's and the 8 core fx8300, how is this possible? cosidering the cores and the speed of the fx 6300 and fx 8300.
a b 4 Gaming
a b à CPUs
September 17, 2013 4:14:12 PM

The sure thing is that intel are considerably better at all around perfomance and power consumption(at least at stock setings). As for the gaming part, amd compete very well in terms of price/perfomance with intel.
In short, if you want striclty a gaming machine and you want to save money, go with amd, otherwise go with intel...
a b 4 Gaming
a b à CPUs
September 17, 2013 4:22:03 PM

and has power consumption of an i7
Related resources
a b 4 Gaming
a c 214 à CPUs
September 17, 2013 4:28:39 PM

Why did you put the i5-3570k and not the i5-4670k ?

Until games can use more cores the i5's will do better in games, if games started using 4,5,6 or more cores the the AMD 8350 will be a better choice then an i5 and you will have to get the i7's with Hyper Threading to compete and at that point the cost will become a big factor with the i7 costing over $100 more then the 8350.
a b à CPUs
September 17, 2013 4:33:29 PM

I run an fx-8350 and enjoy gaming just fine.
I have gamed on an i5 rig as well just fine.
the amd and intel chips are both great and I love them both.

but gaming at high settings and great FPS really boils down to a great GPU.
I am not saying you can run crysis 3 on max on an i3 and a fifty dollar card.
lol i3's
a b à CPUs
September 17, 2013 4:33:44 PM

Performance at a slimlar clock speed in single threaded software will generally show the the Intel chips coming out on top due to Ivy being an overall better architecture, but the FX chips tend to fair well in software that can put those extra cores to work. Also the FX chips suck down more power and output more heat as a result.

When it comes to gaming, most modern games generally minimally use at least 4 cores these days and a good few others will use above 4, this works to the advantage of the FX chips in most cases. The FX 6300 is also cheaper, and if you overclock then it's a done deal in my view.
September 17, 2013 5:26:29 PM

inzone said:
Why did you put the i5-3570k and not the i5-4670k ?

Until games can use more cores the i5's will do better in games, if games started using 4,5,6 or more cores the the AMD 8350 will be a better choice then an i5 and you will have to get the i7's with Hyper Threading to compete and at that point the cost will become a big factor with the i7 costing over $100 more then the 8350.


I used it as an example because it is the CPU that I have seen the most people comparing with the 6300's and 8300's, otherwise I would have compared the i5 4670k as it is a more fair comparison.
a b 4 Gaming
a c 79 à CPUs
September 17, 2013 11:25:35 PM

SR-71 Blackbird said:
FX6300 is better than an i3.


how so? for some games the i3 is better.
But yeah, the i5's are better for gaming overall than the fx8350 as ther process more instructions per core and have better Floating Point Units and faster cache. the fx8350, while being 8 core, only has 4 FPU's, and its cache is comparably slow, and its cores can process less per clock cycle. The fx8300 is better for some well-threaded apps, but not for games. Games also have a very variable load, they can not load up 8 cores with fixed functions, as can a video encoding app. calculations within a game constantly change, the load on the cpu is always varying, so its hard to load up lots of cores. It also comes down to how the games are coded and optimized, it would take a lot more time and tweaking to better optimize a game for an 8 core cpu. Generally as cores increase, the performance percentage does not scale with it, after 4 cores you get significantly diminishing returns in games.
September 19, 2013 1:07:14 PM

[/quotemsg]
after 4 cores you get significantly diminishing returns in games.[/quotemsg]

Thanks this is what I wanted to know!!

a b 4 Gaming
a b à CPUs
September 19, 2013 1:23:39 PM

I wouldn't say an i3 is better than a 6300, it's kind of a 2 way street. The i3 will give you 0-5 more FPS in about 60% of games. The 6300 will give you 5-20 more FPS in about 80% of online multiplayer games as the CPU becomes more of a factor and the i3's 2 cores just don't hack it. Also the 6300 is considerably more powerful in terms of audio/video work as most of these programs are able to utilize all 6 cores. When it all boils down the gaming performance 75% of the time is so close you couldn't even tell the difference, productivity performance solidly favors the FX 6300, which is why I chose it over the i3.
a b 4 Gaming
a c 210 à CPUs
September 19, 2013 3:01:05 PM

hafijur said:
inzone said:
if games started using 4,5,6 or more cores the the AMD 8350 will be a better choice then an i5

This is wrong as if the fx8350 has 7 cores being utilised the i5 3570k 4670k at stock will beat the fx8350. It will need all 8 cores of the fx8350 at 90% plus to beat the i5's slightly. Intel haswell has 2.1x ipc advantage over the piledriver fx cpus hence why with half as many cores at same clock intel wins at multithreaded performance usually and intel cpus have more memory bandwidth and ipc making intels great for gaming.


Wrong, the 8350 running anything that takes 5+ threads >3570k.

There is no doubt about it.
September 19, 2013 4:09:05 PM

hafijur said:
8350rocks said:
hafijur said:
inzone said:
if games started using 4,5,6 or more cores the the AMD 8350 will be a better choice then an i5

This is wrong as if the fx8350 has 7 cores being utilised the i5 3570k 4670k at stock will beat the fx8350. It will need all 8 cores of the fx8350 at 90% plus to beat the i5's slightly. Intel haswell has 2.1x ipc advantage over the piledriver fx cpus hence why with half as many cores at same clock intel wins at multithreaded performance usually and intel cpus have more memory bandwidth and ipc making intels great for gaming.


Wrong, the 8350 running anything that takes 5+ threads >3570k.

There is no doubt about it.


So you are saying the fx8350 is 65-75% faster then an i5 3570k.


Read this thread again carefully. A 8350, on average, would have to be overclocked at 4.5 /4.6g to compete with a i5 at stock speed.
a b 4 Gaming
a c 210 à CPUs
September 20, 2013 9:44:15 AM

toostrike said:
hafijur said:
8350rocks said:
hafijur said:
inzone said:
if games started using 4,5,6 or more cores the the AMD 8350 will be a better choice then an i5

This is wrong as if the fx8350 has 7 cores being utilised the i5 3570k 4670k at stock will beat the fx8350. It will need all 8 cores of the fx8350 at 90% plus to beat the i5's slightly. Intel haswell has 2.1x ipc advantage over the piledriver fx cpus hence why with half as many cores at same clock intel wins at multithreaded performance usually and intel cpus have more memory bandwidth and ipc making intels great for gaming.


Wrong, the 8350 running anything that takes 5+ threads >3570k.

There is no doubt about it.


So you are saying the fx8350 is 65-75% faster then an i5 3570k.


Read this thread again carefully. A 8350, on average, would have to be overclocked at 4.5 /4.6g to compete with a i5 at stock speed.


Not in things that use 5+ threads.

Like rendering, encoding, encryption, compression, video editing, etc.

In fact, gaming is about the *only* thing, outside of iTunes specifically, where the 8350 is not on par or better than the 3570k flat out.

If you're *just* gaming, the 3570k is great. If you heavily multitask, or run CPU intensive programs that are well threaded...then the 8350 is often a *better* choice for *less* money. In those types of scenarios, it's hot on the heels of the i7 class CPUs from Intel.
a b 4 Gaming
a c 455 à CPUs
September 20, 2013 9:51:53 AM

FX 6300 > i3. FX 6300 can overclock and the i3 cannot. The FX 6300 will multitask better and do better in games that are able to take advantage of its number of cores. FX 8320 would be my pic for a gaming rig over an i5 at this point as more money can be spent on GPU. Had the current FX lineup been available when I needed a new board and CPU, I would have bought that over my i5 3570k. Thanks to Microcenter, I could have had and FX 8320 with motherboard for far less than what I paid for my 3570k and Z77 pro.
a b à CPUs
September 20, 2013 11:01:52 AM

The problem with comparing i5 "K" chips to the FX-8350 is that people don't buy "K" chips just to run them at stock 3.4GHz as they are typically asymmetrically benchmarked (3.4GHz vs 4.0GHz). So "like for like" as they're bought to be used (both run at the same 4.0-4.6GHz clock speeds) = "Add 10% onto the FX-8350 stock score, but add typical 20-35% onto the i5 stock score"...

End result, eg, x264 encoding both at 4.4GHz:-
http://www.anandtech.com/show/6396/the-vishera-review-a...

FX-8350 = 62.2fps stock becomes 68.4fps OC'd
i5-3570K= 54.1fps stock becomes 70.3fps OC'd
a b 4 Gaming
a c 210 à CPUs
September 20, 2013 12:07:15 PM

BSim500 said:
The problem with comparing i5 "K" chips to the FX-8350 is that people don't buy "K" chips just to run them at stock 3.4GHz as they are typically asymmetrically benchmarked (3.4GHz vs 4.0GHz). So "like for like" as they're bought to be used (both run at the same 4.0-4.6GHz clock speeds) = "Add 10% onto the FX-8350 stock score, but add typical 20-35% onto the i5 stock score"...

End result, eg, x264 encoding both at 4.4GHz:-
http://www.anandtech.com/show/6396/the-vishera-review-a...

FX-8350 = 62.2fps stock becomes 68.4fps OC'd
i5-3570K= 54.1fps stock becomes 70.3fps OC'd


Interestingly enough, The 8350 typically overclocks 400 MHz higher than that on a 990 board.

My 8350 is running 4.4 at stock voltage, on a 970 board, using air cooling with low temps. I am confident had I bought something like a sabertooth or crosshair board, I would easily hit 5.0 GHz.

Which a 5.0 GHz 8350 would easily, once again, reclaim the performance advantage.

Unless you want to discuss recompiled for architecture Linux benchmarks, where the 8350 defeats even a 3930k at rendering...?

The issue with trying to use benchmarks as concrete evidence is simple, there's too great a margin of error in the tests, there's not a true consensus among any 5 websites that shows the same exact performance, and many of them use compilers or other optimizations that favor one architecture specifically, and have little to no benefit at all to the other.

So, when it comes to benchmarks, they're useful to get a "very rough" idea...but they are not a definitive word on "this is how it will be always".

TL;DR: Take synthetic benchmarks of any kind with a grain of salt, and many of them out there are optimized to favor an architecture over the other, even if only subtly.
a b 4 Gaming
a c 210 à CPUs
September 20, 2013 12:39:18 PM

BSim500 said:
8350rocks said:
Which a 5.0 GHz 8350 would easily, once again, reclaim the performance advantage.

Unfortunately, at 5GHz it also doubles as a +270w room heater...


Considering the fact that my 8350 is running at stock voltage @ 4.4 and is very stable on a 4+2 VRM board...I find it extremely hard to believe that the numbers you quote are accurate across the board.

In fact, I would wager that's not the fact at all...

Back to having a heaping tablespoon of salt again...
a b 4 Gaming
a c 214 à CPUs
September 20, 2013 1:23:20 PM

I guess we're getting off track, might be a good time to close this thread ss the Fan Boys seem to be taking over as usual.
a b 4 Gaming
a b à CPUs
September 21, 2013 6:46:55 PM

inzone said:
I guess we're getting off track, might be a good time to close this thread s the Fan Boys seem to be taking over as usual.


Ya well ya can't really say AMD in these forums now a days without being a fanboy can ya ? Just go with the common belief, don't upset the apple cart. That kind of thing right ?
a b 4 Gaming
a c 79 à CPUs
September 22, 2013 5:29:29 PM

8350rocks said:
BSim500 said:
The problem with comparing i5 "K" chips to the FX-8350 is that people don't buy "K" chips just to run them at stock 3.4GHz as they are typically asymmetrically benchmarked (3.4GHz vs 4.0GHz). So "like for like" as they're bought to be used (both run at the same 4.0-4.6GHz clock speeds) = "Add 10% onto the FX-8350 stock score, but add typical 20-35% onto the i5 stock score"...

End result, eg, x264 encoding both at 4.4GHz:-
http://www.anandtech.com/show/6396/the-vishera-review-a...

FX-8350 = 62.2fps stock becomes 68.4fps OC'd
i5-3570K= 54.1fps stock becomes 70.3fps OC'd


Interestingly enough, The 8350 typically overclocks 400 MHz higher than that on a 990 board.

My 8350 is running 4.4 at stock voltage, on a 970 board, using air cooling with low temps. I am confident had I bought something like a sabertooth or crosshair board, I would easily hit 5.0 GHz.

Which a 5.0 GHz 8350 would easily, once again, reclaim the performance advantage.

Unless you want to discuss recompiled for architecture Linux benchmarks, where the 8350 defeats even a 3930k at rendering...?

The issue with trying to use benchmarks as concrete evidence is simple, there's too great a margin of error in the tests, there's not a true consensus among any 5 websites that shows the same exact performance, and many of them use compilers or other optimizations that favor one architecture specifically, and have little to no benefit at all to the other.

So, when it comes to benchmarks, they're useful to get a "very rough" idea...but they are not a definitive word on "this is how it will be always".

TL;DR: Take synthetic benchmarks of any kind with a grain of salt, and many of them out there are optimized to favor an architecture over the other, even if only subtly.


you need good water cooling to get the 8350 to 5ghz, and it would not get there easily. Did some research on this when recently looking into water cooling. many people struggle to hit 5ghz 24/7 stable with 8350's.
But hey, we will just take everything with a grain of salt and just listen to your opinion.
September 22, 2013 6:27:54 PM

6300> i3. intel is always faster when it comes to single thread performance. clock per clock intel wins. but in gaming 6300 would be a better choice because of more cores. crysis 3 and bf3 multiplayer utilize up to 6 cores while the i3 only has 2 physical cores and 2 threads. so its wiser to go with the 6300. when it is an i5 vs 83xx go with the i5 unless you have a liquid nitrogen and overclock your fx to 8ghz. i5 has 4 cores and 4 threads. you wont use anything more than 8 cores. so its no contest between an i5 and 83xx
a b 4 Gaming
a b à CPUs
September 23, 2013 5:01:44 AM

http://translate.google.com/translate?act=url&depth=1&h...



What it comes down to is this: Intel's cores do significantly more work then AMD's do. So even though AMD has more cores and has a higher base clock, Intel is still faster. The difference only grows as you OC, due to Intel OC'ing more (due again to AMD's higher base clock).

As for the "AMD does better as you use more cores" argument, understand how the hardware works: If Intel does twice as much work per core, then a workload that brings an AMD core to 100% would cause an Intel core to go to 50%. Guess what? That core can handle a second thread without affecting processing time. So how much work each core is doing is more important then the number of threads being worked on. This again works against AMD, since a single, high workload thread has a much greater chance of slowing the application as a whole compared to Intel.

As for Crysis 3:

http://pclab.pl/art52489-9.html



i3-3220 roughly equal to the FX-6300.



Roughly equal to the FX-965.



Equal to the FX-6300, within one FPS of the FX-8350.

The reason why the i3 does so badly in "Welcome to the Jungle" is because of a decision made by the devs to offload work from the GPU to the CPU. And the type of work they offloaded (specifically: a subprocess that affects the rendering of grass) is especially in use on this level, hence the i3 doing worse then it does in the "Root of all Evil" and "Post Human" levels. But on the whole, even in games that thread well (Remember: GPUView indicates Crysis 3 uses 11 threads that do significant work), the i3 is only slightly worse then the FX-6300.

Hence why I've long questioned AMD's approach for using FX as their high tier gaming CPU, since the architecture is not particularly good for those types of workloads.
a b 4 Gaming
a c 210 à CPUs
September 23, 2013 6:25:34 AM

iam2thecrowe said:
8350rocks said:
BSim500 said:
The problem with comparing i5 "K" chips to the FX-8350 is that people don't buy "K" chips just to run them at stock 3.4GHz as they are typically asymmetrically benchmarked (3.4GHz vs 4.0GHz). So "like for like" as they're bought to be used (both run at the same 4.0-4.6GHz clock speeds) = "Add 10% onto the FX-8350 stock score, but add typical 20-35% onto the i5 stock score"...

End result, eg, x264 encoding both at 4.4GHz:-
http://www.anandtech.com/show/6396/the-vishera-review-a...

FX-8350 = 62.2fps stock becomes 68.4fps OC'd
i5-3570K= 54.1fps stock becomes 70.3fps OC'd


Interestingly enough, The 8350 typically overclocks 400 MHz higher than that on a 990 board.

My 8350 is running 4.4 at stock voltage, on a 970 board, using air cooling with low temps. I am confident had I bought something like a sabertooth or crosshair board, I would easily hit 5.0 GHz.

Which a 5.0 GHz 8350 would easily, once again, reclaim the performance advantage.

Unless you want to discuss recompiled for architecture Linux benchmarks, where the 8350 defeats even a 3930k at rendering...?

The issue with trying to use benchmarks as concrete evidence is simple, there's too great a margin of error in the tests, there's not a true consensus among any 5 websites that shows the same exact performance, and many of them use compilers or other optimizations that favor one architecture specifically, and have little to no benefit at all to the other.

So, when it comes to benchmarks, they're useful to get a "very rough" idea...but they are not a definitive word on "this is how it will be always".

TL;DR: Take synthetic benchmarks of any kind with a grain of salt, and many of them out there are optimized to favor an architecture over the other, even if only subtly.


you need good water cooling to get the 8350 to 5ghz, and it would not get there easily. Did some research on this when recently looking into water cooling. many people struggle to hit 5ghz 24/7 stable with 8350's.
But hey, we will just take everything with a grain of salt and just listen to your opinion.


You follow the part I bolded for anything and you're ahead of the pack.

As for my chip...I probably got a "good one". I have a buddy who is OC'ed to 5.556 GHz 24/7 stable on a Crosshair MB. Now, he runs an elaborate evaporative cooling setup at the clock he runs at...however, he hit 4.994 GHz on air and that's close enough to 5.0 GHz for me.
a b à CPUs
December 30, 2013 2:06:39 AM

I am beginning to loathe this discussion. The intel fanboys cry about "performance per core", and the AMD fanboys cry about "performance per dollar".


Like it or not, just because the 3-400$ intel machines are great, doesn't mean the 200$ AMD machines are garbage. Its like comparing a 100,000$ Corvette ZL1 to a 350,000$ Lamborghini Murcielago. Yeah... the lamborghini beats the corvette in everything but a straight line drag race... but guess what? If all you do is drag race, why the hell would you spend 3 times as much on the lambo? Oh yes... to say "I have a lambo"... this transposes nicely into the I5/I7 vs FX8/9 debates. Focus on what you're doing , what your budget is, and what you expect out of your rig.

200$ goes a LOT further in the AMD world, but in a limitless budget, we would all be running 6 core intel i7's. They each have their merits and purposes. Find out what you want and expect, and go from there. The natural presumption is to go straight for the i5-3770k. Is it REALLY that much better? Benchmarks of newer games say other wise. A lot of retort seems to be "in older games..."; "in single core blah blah's..."; "hyper threading..."... lets focus on whats out now, and whats going to come. An 8350 will play skyrim JUST FINE. I PROMISE.

budget to performance ratio, thats all this is. Without AMD, we'd all be forking over lots of money for sub par processors. With intel, AMD wouldn't be advancing multi core solutions.






Why is the I3 even in this conversation?


a c 158 à CPUs
December 30, 2013 2:09:59 AM

AMD has made the perfect gaming chip, but games are seriously slow to catch up and still using horribly optimized code. The majority of them don't even use 64bit, that's how slow they are to adapt. So when you have an old game that is only using 2 cores (suck as Skyrim for instance), the intel with it's much stronger ipc will perform better.

On the flipside, if you have a game that can use all 8 cores, the AMD fx-83 will perform better and even rival the much more expensive i7 4770k. This is why Mantle is incredibly exciting.
April 10, 2014 11:14:00 AM

the way I see it is: AMD processors are meant for overclocking. even the cheap ones, and all you have to do is spend a little money on a decent cooling solution. if you want to overclock an Intel processor, you have to get an i7, which will run you at least $300. I.E., if you get an AMD and don't overclock, you're not really getting your money's worth, and might as well have gone with Intel. that just goes for the AMD CPUs, not APUs. those can also be overclocked some, but not ideal for gaming.
a b 4 Gaming
a b à CPUs
April 10, 2014 11:18:38 AM

You dont need an i7 to overclock. core i5-4670k as an example
a b 4 Gaming
a c 455 à CPUs
April 10, 2014 11:27:02 AM

Yea, not sure what made you think an i7 was needed to overclock. o.O Overclocking in general costs more than it is worth anymore. Hence why most of my recommendations are for i5 4440/4570 and the 1230 v3.
!