The Best CPU for the Money: AMD FX vs. Intel Budget Shootout

This article makes no sense. It's called "Budget Shootout", yet it uses a high-end GPU to emphasize AMDs weaknesses.

A "budget" oriented buy would also use this budget CPU with a bugdet GPU. So none of this 980 nonsense. The gaming benches should've used an R9 270X or the GTX 760/960.

Budget gaming setups usually max out GPU performance way before CPUs even matter.

So today that flagship graphics card is being matched by a $120 budget GPU. So what happens in 2-3 or even 4 years when those who bought a Core i3-4360 or FX-8320E want to game with a GPU that is nearly as powerful or perhaps even more powerful than the GTX 980?

What happens is those that bought the Core i3-4360 will likely be able to extract the maximum amount of power out of their new GPU. Those that bought the FX-8320E will be overclocking for the moon trying to extract every last MHz just to try and get near the limits of their new GPU.

What happens? Easily. With a 980, in 3-4 years time, trying to max out games *will* max out GPU load, therefore using an i3 at 2 GHz or i7 at 4 GHz becomes irrelevant. I am using a 560 Ti with an FX 6300 @ 4.5 GHz. By all means, my CPU is considered *slower* than any i3s on the market. Maybe even slower than the Pentium 3280 or whatever the unlocked version is. Does it affect me? Absolutely not. I'm GPU bound in absolutely every game I play, when I'm at 99% usage every time, all the time.

Also, you are slightly short sighted in close term. You think the 980 in this case, will be hampered by the poor CPU performance of an FX 8xxx, even at 4.5-5 GHz, even compared to an i3. With the dawn of DirectX12 on us, which dramatically reduced CPU dependency, you still think it will matter if you have an FX 8 core or an i3-i5-i7 ?

I'm not ridiculous. I won't ever suggest anyone running SLI/CrossFire to something faster than the level of an 280X/770 a current-gen FX CPU. Not even a 5 GHz one. But if budget is a problem, by all mean, get the better GPU and the best CPU possible. And an FX 8000 is a better choice than an i3 in the long term, with DX12, better multithreading in games and all. Also, a lot of people here seem to acknowledge the whole article is kinda NULL because your FX was underperforming for whatever reason.
 
This article makes no sense. It's called "Budget Shootout", yet it uses a high-end GPU to emphasize AMDs weaknesses.

A "budget" oriented buy would also use this budget CPU with a bugdet GPU. So none of this 980 nonsense. The gaming benches should've used an R9 270X or the GTX 760/960.

Budget gaming setups usually max out GPU performance way before CPUs even matter.

So today that flagship graphics card is being matched by a $120 budget GPU. So what happens in 2-3 or even 4 years when those who bought a Core i3-4360 or FX-8320E want to game with a GPU that is nearly as powerful or perhaps even more powerful than the GTX 980?

What happens is those that bought the Core i3-4360 will likely be able to extract the maximum amount of power out of their new GPU. Those that bought the FX-8320E will be overclocking for the moon trying to extract every last MHz just to try and get near the limits of their new GPU.

What happens? Easily. With a 980, in 3-4 years time, trying to max out games *will* max out GPU load, therefore using an i3 at 2 GHz or i7 at 4 GHz becomes irrelevant. I am using a 560 Ti with an FX 6300 @ 4.5 GHz. By all means, my CPU is considered *slower* than any i3s on the market. Maybe even slower than the Pentium 3280 or whatever the unlocked version is. Does it affect me? Absolutely not. I'm GPU bound in absolutely every game I play, when I'm at 99% usage every time, all the time.

Also, you are slightly short sighted in close term. You think the 980 in this case, will be hampered by the poor CPU performance of an FX 8xxx, even at 4.5-5 GHz, even compared to an i3. With the dawn of DirectX12 on us, which dramatically reduced CPU dependency, you still think it will matter if you have an FX 8 core or an i3-i5-i7 ?

I'm not ridiculous. I won't ever suggest anyone running SLI/CrossFire to something faster than the level of an 280X/770 a current-gen FX CPU. Not even a 5 GHz one. But if budget is a problem, by all mean, get the better GPU and the best CPU possible. And an FX 8000 is a better choice than an i3 in the long term, with DX12, better multithreading in games and all. Also, a lot of people here seem to acknowledge the whole article is kinda NULL because your FX was underperforming for whatever reason.
also DX12 and mantle brings multithreaded rendering into play and then fx will be much better than i3

Thief-High.png


look at minimal fps
 
Wow guys you were right the Cinebench multi-thread result for the overclocked 8320E was wrong!

Before you jump up in the air and celebrate a victory for FX owners everywhere this is not because of an overclocking mistake. I hit 6 instead of 7 when entering the data into Excel :S

Stupid mistake I know, the score should have been 706pts. The thing is you guys could have picked up on this since the single thread score of 109 is accurate. So with the exception of the Cinebench R15 multi-threaded score which I will now fix every other result at 4.6GHz has been confirmed.

Rather than post screen shots of your Cinebench score, a synthetic benchmark no less, why didn't you confirm the 7-zip and WinRAR tests or anything else for that matter?

Nonetheless thank you for spotting the typo.

What happens? Easily. With a 980, in 3-4 years time, trying to max out games *will* max out GPU load, therefore using an i3 at 2 GHz or i7 at 4 GHz becomes irrelevant. I am using a 560 Ti with an FX 6300 @ 4.5 GHz. By all means, my CPU is considered *slower* than any i3s on the market. Maybe even slower than the Pentium 3280 or whatever the unlocked version is. Does it affect me? Absolutely not. I'm GPU bound in absolutely every game I play, when I'm at 99% usage every time, all the time.

Also, you are slightly short sighted in close term. You think the 980 in this case, will be hampered by the poor CPU performance of an FX 8xxx, even at 4.5-5 GHz, even compared to an i3. With the dawn of DirectX12 on us, which dramatically reduced CPU dependency, you still think it will matter if you have an FX 8 core or an i3-i5-i7 ?

I'm not ridiculous. I won't ever suggest anyone running SLI/CrossFire to something faster than the level of an 280X/770 a current-gen FX CPU. Not even a 5 GHz one. But if budget is a problem, by all mean, get the better GPU and the best CPU possible. And an FX 8000 is a better choice than an i3 in the long term, with DX12, better multithreading in games and all. Also, a lot of people here seem to acknowledge the whole article is kinda NULL because your FX was underperforming for whatever reason.

Okay buy an FX processor, core efficiency (or inefficiency I should say) has been proven and we should hold out to see DX12 save the series almost 5 years after it was first introduced. You call me short sighted yet your theory is yet to be seen, hell how far off are decent DX12 games anyway? - Side note will DX12 fix the application and encoding performance?

You seemed to have missed the point of my first post and didn't understand the GTX 580 example which was a real scenario that took place in recent history.

Finally the whole article is kinda not NULL because the FX wasn't under-performing. There was a typo in one synthetic test, that's it. Every other test is accurate and the FX 8320E was running all cores clocked at 4.6GHz.
 
To be honest Steve, I was finding it quite funny how much these AMD guys needed to defend their "top end" AMD chip against Intel's lowest Core series. It was manly of you to go over the numbers again and fix it. But still, at the end of the day, the fact you can compare a duel core to an 8 core proccessor, that's pretty poor performance no matter what angle you look at it from AMD. Do we know if they have any new architecture coming soon? The rumour mill must have something ;)
 
Your encoding performance is rubbish Steve
my cpu:
pass 1 @ 4.6ghz
encoded 11812 frames, 67.96 fps, 7754.28 kb/s

please retest or remove this useless review ;)

Encoding_02.png
 
Wow guys you were right the Cinebench multi-thread result for the overclocked 8320E was wrong!

Before you jump up in the air and celebrate a victory for FX owners everywhere this is not because of an overclocking mistake. I hit 6 instead of 7 when entering the data into Excel :S

Stupid mistake I know, the score should have been 706pts. The thing is you guys could have picked up on this since the single thread score of 109 is accurate. So with the exception of the Cinebench R15 multi-threaded score which I will now fix every other result at 4.6GHz has been confirmed.

Rather than post screen shots of your Cinebench score, a synthetic benchmark no less, why didn't you confirm the 7-zip and WinRAR tests or anything else for that matter?

Nonetheless thank you for spotting the typo.

What happens? Easily. With a 980, in 3-4 years time, trying to max out games *will* max out GPU load, therefore using an i3 at 2 GHz or i7 at 4 GHz becomes irrelevant. I am using a 560 Ti with an FX 6300 @ 4.5 GHz. By all means, my CPU is considered *slower* than any i3s on the market. Maybe even slower than the Pentium 3280 or whatever the unlocked version is. Does it affect me? Absolutely not. I'm GPU bound in absolutely every game I play, when I'm at 99% usage every time, all the time.

Also, you are slightly short sighted in close term. You think the 980 in this case, will be hampered by the poor CPU performance of an FX 8xxx, even at 4.5-5 GHz, even compared to an i3. With the dawn of DirectX12 on us, which dramatically reduced CPU dependency, you still think it will matter if you have an FX 8 core or an i3-i5-i7 ?

I'm not ridiculous. I won't ever suggest anyone running SLI/CrossFire to something faster than the level of an 280X/770 a current-gen FX CPU. Not even a 5 GHz one. But if budget is a problem, by all mean, get the better GPU and the best CPU possible. And an FX 8000 is a better choice than an i3 in the long term, with DX12, better multithreading in games and all. Also, a lot of people here seem to acknowledge the whole article is kinda NULL because your FX was underperforming for whatever reason.

Okay buy an FX processor, core efficiency (or inefficiency I should say) has been proven and we should hold out to see DX12 save the series almost 5 years after it was first introduced. You call me short sighted yet your theory is yet to be seen, hell how far off are decent DX12 games anyway? - Side note will DX12 fix the application and encoding performance?

You seemed to have missed the point of my first post and didn't understand the GTX 580 example which was a real scenario that took place in recent history.

Finally the whole article is kinda not NULL because the FX wasn't under-performing. There was a typo in one synthetic test, that's it. Every other test is accurate and the FX 8320E was running all cores clocked at 4.6GHz.
Thanks for finding the issue, the r15 score was the only one that seemed off to me, now we know why :)
 
To be honest Steve, I was finding it quite funny how much these AMD guys needed to defend their "top end" AMD chip against Intel's lowest Core series. It was manly of you to go over the numbers again and fix it. But still, at the end of the day, the fact you can compare a duel core to an 8 core proccessor, that's pretty poor performance no matter what angle you look at it from AMD. Do we know if they have any new architecture coming soon? The rumour mill must have something ;)
Zen archetecture around 2016. The article to me shows how silly most chips over a quad core are for most users. But the handbrake benchmark shows why I went 8320. In my larger batch encodes I do it cuts hours of time compared to other chips in the price range I paid, if I have to burn some extra electricity in that time so be it for me its cheap. I also did it for the unlock aspect this was a fully unlocked chip compared to locked intel chips and I wasn't gonna go pentium anniversary. Amd does neee to remove therr head from there behinds though, u want us to move to fm2+ fine , but you need to give me a reason to do so.
 
Last edited:
Your encoding performance is rubbish Steve
my cpu:
pass 1 @ 4.6ghz
encoded 11812 frames, 67.96 fps, 7754.28 kb/s

please retest or remove this useless review ;)

LOL you are a crack up. Thanks for providing the free entertainment.
 
I doubt it would be any good, those are APUs and lack raw power, the i3 will beat it to it's grave.
Yes, however its the new steamroller architecture which is great for budget builds and overclock nicely. Just thought it would be a nice addition though its not completely necessary as 760K and others in the past have been decent budget chips.

I don't get what all the arguing is about (Other than the typo part), most programs and games favor single threaded performance over multithreaded. Its a fact of life hence why core to core performance is more important than a straight up more cores. When deciding on a test, its better to alleviate all potential bottlenecks as best as possible so you get clear facts. If one processor performs 15% better on the best GPU at the same price as another, its probably going to be your better choice in the end plain and simple as putting a GPU that shows the gap closer to 5% would not help decide much.
 
Cmon Steve show us that u can clock :D and repair that enconding results (also you 702 in cinebench is still low for 2133mhz memory try harder)
 
Cmon Steve show us that u can clock :D and repair that enconding results (also you 702 in cinebench is still low for 2133mhz memory try harder)

This is the last time I am acknowledging any of the trash you post so don’t bother trying harder.

The tone of the article wouldn’t change even if the overclocked FX 8320E was 10% faster in every test.

Having been a massive fan of the AMD K6, K6-2, K6-III, K7 Athlon and K8 range I hate to see where things are at now. Nothing would please me more than to see AMD kicking some Intel backside in the CPU world, we could actually write CPU related articles on a regular basis again.

Anyway if you can’t handle the truth that you are better off buying a base model Core i5 than you are spending similar money on the FX-8320E and cooler then that is your problem.
 
You cant even count correctly Steve 67.96fps is about 57% faster then your measured 42.9fps
as OC FX-8xxx is 2times faster in nearly every multithreaded use than i3 at same price :)

Moderator note: Multiple accounts not permitted.
 
Last edited:
Wow guys you were right the Cinebench multi-thread result for the overclocked 8320E was wrong!

Before you jump up in the air and celebrate a victory for FX owners everywhere this is not because of an overclocking mistake. I hit 6 instead of 7 when entering the data into Excel :S

Stupid mistake I know, the score should have been 706pts. The thing is you guys could have picked up on this since the single thread score of 109 is accurate. So with the exception of the Cinebench R15 multi-threaded score which I will now fix every other result at 4.6GHz has been confirmed.

Rather than post screen shots of your Cinebench score, a synthetic benchmark no less, why didn't you confirm the 7-zip and WinRAR tests or anything else for that matter?

Nonetheless thank you for spotting the typo.



Okay buy an FX processor, core efficiency (or inefficiency I should say) has been proven and we should hold out to see DX12 save the series almost 5 years after it was first introduced. You call me short sighted yet your theory is yet to be seen, hell how far off are decent DX12 games anyway? - Side note will DX12 fix the application and encoding performance?

You seemed to have missed the point of my first post and didn't understand the GTX 580 example which was a real scenario that took place in recent history.

Finally the whole article is kinda not NULL because the FX wasn't under-performing. There was a typo in one synthetic test, that's it. Every other test is accurate and the FX 8320E was running all cores clocked at 4.6GHz.

Cool, thanks for owning it Steve. I wish AMD was more competitive, but that's not the case these days. I still think they are nice for low to mid builds around their FM2+ APUs, but the bulldozers aren't that great, that's why my ESXi box is still rocking a 1090T.
 
The tone of the article wouldn’t change even if the overclocked FX 8320E was 10% faster in every test.

Having been a massive fan of the AMD K6, K6-2, K6-III, K7 Athlon and K8 range I hate to see where things are at now. Nothing would please me more than to see AMD kicking some Intel backside in the CPU world, we could actually write CPU related articles on a regular basis again.
Unfortunately, when the Empire hit it out of the park with Core, AMD have spent most of the next decade looking like the catcher here
post-28317-Darth-Vader-Baseball-gif-homer-sbtX.gif
 
Ok so ONE score was off, the rest was all good, Steve acknowledge it and they keep trolling??

You really don't know when to stop do you? Get it right, IT IS NOT AN i7 FOR AS MUCH AS YOU WOULD LIKE! THANKS!
 
Simply put, if this was going to be a budget build, let it stand alone. Take the extreme gaming GTX980 GPU out of the mix and re-run the tests.
This is a CPU review, not a GPU review. A high-end GPU was used to ensure there was no bottle-neck in CPU performance. The use of the GTX 980 should be ignored.

Who needs a video card to run office software or to watch video on a computer? So subtract the price of the video card and use the integrated IGP on the CPU. To make a realistic comparison open a database application, open your e-mail groupware app like outlook, open a web browser to a few different sites and watch a video and do some searching online and listen to the radio over the Internet and then do a benchmark on everything running at once.
 
Real people use more than one application running at a time. If you do no gaming the point is worthless how fast a computer can run. You only care if the windows open quickly and the computer can keep up with what you are doing. I wonder what effect using an SSD would have? Also I never overclock, so no overclocking is allowed on a real test. I wonder what would happen if you loaded a huge database and ran a querry on say 50,000 rows in Excell SQL against a database with about 10 tables. I work at a college and use databases and large applications all the time.

To be truthful I just don't trust AMD. I have had issues with their CPU's in the past. If the processor has 8 cores, test it against an 8 core Intel Xeon.
 
Well I discarded your post, because you came into a tech site to talk about how fast it could open office. BUT then...

To be truthful I just don't trust AMD. I have had issues with their CPU's in the past. If the processor has 8 cores, test it against an 8 core Intel Xeon.

...you compared a server grade proc against a cheap budget home user friendly proc, it just can't be, ever.
 
...you compared a server grade proc against a cheap budget home user friendly proc, it just can't be, ever.
He's onto something though, Let's through out budget for a second and compare AMD's top end 8 core processor vs Intel's top end 8 core and see who wins, considering 2 of the 8 intel cores will give the 8 core AMD a run for it's money, another 6 would make for a funny comparison xD
 
I dont know whats wrong with those tests, but here it goes:

4.5ghz 8320E is faster then 4.6ghz 8320E in three more tests:

Pc Mark 8 Creative suite

8320E 4.6ghz Techspot 5529
8320E 4.5ghz Nexthardware 5820


Pc Mark 8 Work suite

8320E 4.6ghz Techspot 4202
8320E 4.5ghz Nexthardware 4731


FX 8320E 3200ghz Turbo On: I will not comment that
MSI-970-Gaming-pcmark8-ok.png



Winrar 5 Compression

8320E 4.6ghz Techspot Single threaded 1608 Multi threaded 8175
8320E 4.5ghz Nexthardware Single threaded 1691 Multi threaded 8433



MSI-970-Gaming-winrar-ok.png


Dont forget that techposts fx processor has 100ghz more.

Other tests I cant compare, but I will try to.
 
Steve just recently I was banned from a site for linking and defending your article. I also stated I saw similar results since I just upgraded from a 8350 to a i3 4360. The mod at the place is bias and a Amd fanboy(it's like Hitler).

Its funny the first thing they do is attack the results and even go as far as claiming bias benchmarks, even if that was the fact why buy hardware that might perform badly in some games? Just never made any sense to me. Also why didn't you include newer games? Many games give inconsistent benchmarks in the FX series more so then the I series CPU's.

The more we look into Amd fanboy's the more I think its like a religion one I'm happy I'm not part of.
 
Yeah I am not quite sure what this crap is, but I have an 8120 and overclocked to 4.5 it matchs a i7 3770. Can you tell who techspot is supporting, cause every time amd out did intel they then point out the power usage. When literally no one cares about power usage.
you dumb*** power consumption is a crucial thing for budget builders huge power consumption= a pricier CPU cooler + an expensive PSU + a really good case for noise management
 
you dumb*** power consumption is a crucial thing for budget builders huge power consumption= a pricier CPU cooler + an expensive PSU + a really good case for noise management


Yes, but yet FX cpus have lower cpu temps, then Intel does.

And when I set my FX 8320 procesor to 4ghz and disable turbo, I need only 1.272V, and normal consumption when playing games doesnt go above 100W.... at 3.5 ghz and 1.164V wattage usage wasnt going above 95W even when I ran Prime95 FFT test. Thats why E version is rated at 95W.

The reason for high wattage usage is Turbo, when Turbo is enabled, cpu voltage goes automatically to 1.45V, but as I said normally you need only 1.272V for 4ghz.

So people who put their minds into FX cpu voltage/clock ratio wont get high power consumption.

BTW: My max gaming temp is 55C on cpu socket and 45C on cores.
 
Last edited:
Back