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

[Kibaruk]

I don't get all the hostility, like I said earlier I have a 8320 (not the E) and I find it to be a wonderful chip. But the reality is it's pretty even and can even beat intel's Sandy bridge and sometimes Ivy bridge in multi-threaded tests, but with the enhancements in IPC and power efficiency in haswell and intel has even begun to drop prices down a bit on haswell chips, its really starting look like AMD needs to figure something out. I would say any 4-8 core chip from any of the the two companies from 2012 on up (especially if unlocked) is more then enough for just about any common computer tasks we run today so in he end who really cares.

You hit the nail, it's a good chip no one says different, but it's hardly a competition to intels solutions, on the comparisson was more even with the dual core i3 than the i5 -even with 8 cores- and this has nothing to do with it being BAD, it's just that intels solutions tend to be faster and more efficient.

 

[PatcheZ]

Great review Steve.
I've been waffling at the idea whether I had went with the right processor or not during BF.
FX-8320 vs i5-4440 processors with their respective motherboards were only ~$10-20CAD difference... and I went with the "better" deal. LOL

The i3 vs i5 debate is a bit more complicated I think, as other users have mentioned, there is(are?) game(s) such as Dragon Age - Inquisition that require a quad core processor just to load (my only reason for getting quad+ core). If this is going to be the trend moving forward on EA/Bioware/Frostbite engine, then this might actually put dual cores out of consideration for gaming.

PS Those Intel Haswell's are crazy efficient. When I installed my processor, I didn't notice that the CPU fan was stuck and not spinning. When I prime stress tested the CPU, it only went up with 85-88°C on essentially PASSIVE cooling.

 

[SuperVeloce]

DA:I will not know i3 is a dualcore. Game sees 4 threads, the same as on i5. And game starts normally on i3, that is a fact.

 

[Jugoslav]

Fx 8320E overclocked to 4.6ghz scores 604 points on Cinebench benchmark, while My Fx 8320 overclocked to 4ghz scores 648 points??!!

FX 8320E is probably not overclocked to 4.6ghz on all cores, but have only turbo frequency set to 4.6ghz. Thats why I have much better score with 4ghz. Properly 4.6ghz overclocked Fx processor scores 720 points in Cinebench R15.


To properly overclock FX processor you need to disable Turbo, and then set multiplier to 23x, with that all processor cores will run at 4.6ghz, and we will avoid full load and half load turbo*.


*Stock FX 8320E turbo explanation

Half load turbo goes to max frequency(4ghz) but only with half of cores. (4ghz turbo load also happens very rarley)
Full load turbo uses all cores but goes only to 3.5ghz(happens most of the time.)


So this test and probably many of other tests where FX processors are overclocked are not even correctly executed... well they are, but only for Intel fans..

 

[Jugoslav]

So actualy you ran most of benchmarks on half of cores, and on half turbo.and thats for both stock and overclocked 8320E.

And thats why i3 beats 8320E in your benchmarks. lol

 

[Steve]

So actualy you ran most of benchmarks on half of cores, and on half turbo.and thats for both stock and overclocked 8320E.

And thats why i3 beats 8320E in your benchmarks. lol

I made it pretty clear that we understand all this in my last post but for you I will explain it again. Turbo was disabled and all cores were synced.

If what you are saying is true which it isn't then the power consumption results would be even more horrendous.

Funny how you talk about Intel fanboys but you are clearly an AMD fanboy and the foremost expert on FX overclocking.

Side note, if you analyse the rest of the data your point becomes invalid anyway. Many other tests show large gains from the overclock. 7-zip for example shows a 39% performance boost once overclocked. If the stock 8320E was running at just 3.2GHz in this test then we boosted the frequency by 43% and got 39% more performance. Do you think the 8320E wasn’t using its Turbo feature in this test?

So actualy you ran most of benchmarks on half of cores, and on half turbo.and thats for both stock and overclocked 8320E.

And thats why i3 beats 8320E in your benchmarks. lol

It beats it in some tests because its core efficiency is vastly superior, sorry if that makes your purchase somehow feel inadequate. lol

 

[Jugoslav]

Sure its the truth.

Why does my 4ghz 8320 get better score in Cinebench R15, then 8320E at 4.6ghz??!!

FX 8320E overclocked on 4.6ghz, should get 720 points! Thats 104 points more, then what you get in your benchmark...

 

[Jugoslav]

So there must be something wrong... or you didnt disabled turbo, even when you said you did.

 

[OneB1t]

606 for cinebench is very low for 4.6ghz

your VRM trotling or APM is not disabled...
just search on the internet for correct results
5ghz is over 800 points no problem
4.5ghz and fast memory = 720

also overclock CPU-NB it will give u nice boost in memory performance

 

[OneB1t]

So there must be something wrong... or you didnt disabled turbo, even when you said you did.

So actualy you ran most of benchmarks on half of cores, and on half turbo.and thats for both stock and overclocked 8320E.

And thats why i3 beats 8320E in your benchmarks. lol

I made it pretty clear that we understand all this in my last post but for you I will explain it again. Turbo was disabled and all cores were synced.

If what you are saying is true which it isn't then the power consumption results would be even more horrendous.

Funny how you talk about Intel fanboys but you are clearly an AMD fanboy and the foremost expert on FX overclocking.

Side note, if you analyse the rest of the data your point becomes invalid anyway. Many other tests show large gains from the overclock. 7-zip for example shows a 39% performance boost once overclocked. If the stock 8320E was running at just 3.2GHz in this test then we boosted the frequency by 43% and got 39% more performance. Do you think the 8320E wasn’t using its Turbo feature in this test?

So actualy you ran most of benchmarks on half of cores, and on half turbo.and thats for both stock and overclocked 8320E.

And thats why i3 beats 8320E in your benchmarks. lol

It beats it in some tests because its core efficiency is vastly superior, sorry if that makes your purchase somehow feel inadequate. lol

I know intel pays u alot for such rubbish testing but next time try to select better games idealy from 2014/1015 era you will find that in multithreaded games fx-8320 simply destroy every i3
(watch dogs, assasins creed unity patch 1.4, crysis 3 welcome to the jungle, battlefield 4 MP)

 

[Kibaruk]

Guys... you can reply in a single post, or if you forgot you can edit your post and then add the information, when you want to quote reply someone you can do the same thing. Stop the multiple posts for a single idea, this isn't a chat. Thanks!

Also, stop looking at "your" results, and start looking into the average.

EDIT: Back on topic:
http://www.guru3d.com/articles-pages/amd-fx-8320e-processor-review,11.html
http://cpuboss.com/cpu/AMD-FX-8320E
http://cpuboss.com/cpus/Intel-Core-i3-4360-vs-AMD-FX-8320E

Now please STAHP with all the fanboysm. Thanks!

 

[OneB1t]

Guys... you can reply in a single post, or if you forgot you can edit your post and then add the information, when you want to quote reply someone you can do the same thing. Stop the multiple posts for a single idea, this isn't a chat. Thanks!

u cant be serious using such website for CPU performance comparsion

 

[Jugoslav]

http://saveotic.com/images/2015/02/11/4OADK.png

if only am3+ board had a better memory controller

 

[Kibaruk]

u cant be serious using such website for CPU performance comparsion

In that "can't be serious" page they have all the benchmarks posted, cinebench among others.

 

[Jugoslav]

In that "can't be serious" page they have all the benchmarks posted, cinebench among others.

Here are cinebench 15 benchmark.

Fx 8320 4ghz

Fx 8350 4.6ghz

 

[cliffordcooley]

Here are cinebench 15 benchmark.
Fx 8320 4ghz
Fx 8350 4.6ghz

What is your point? The "E" version is less powerful than your FX 8320.

http://www.guru3d.com/articles_pages/amd_fx_8320e_processor_review,11.html

 

[Jugoslav]

OMG. cant you read. I score 648 with 4ghz, and techspot score 604 on 4.6ghz, if he overclocked as he should he would score 720 points.

 

[Kibaruk]

OMG. cant you read. Its not the same proc, the "e" vs the normal one.

 

[OneB1t]

OMG. cant you read. Its not the same proc, the "e" vs the normal one.

these 2 cpus are same on different basic frequency

so

fx-8350@4.6ghz is totally same as fx-8300@4.6 or fx-8320E/8370E@4.6ghz
they all need to score around 720 points in cinebench at 4.6ghz

 

[OneB1t]

It is same on same frequency results from benchmarks should be exactly same

 

[Kibaruk]

It is same on same frequency results from benchmarks should be exactly same

If you saw cinebench bencmark it's clearly NOT the same.

 

[GeforcerFX]

I would say there is something possibly wrong with the R15 scores or that anandtech's could be wrong.
http://www.anandtech.com/show/8864/amd-fx-8320e-cpu-review-the-other-95w-vishera/4
they scored 767 @4.8ghz. But they used a 12+2 phase 990FX board and this review used MSI's 8+2 970 board that might be the culprit for the difference in scores? Granted I have seen the 700 scores coming out of other 8+2 boards and I don't know why a 970 would limit the cpu performance compared to a 990FX (at least it shouldn't).

 

[Jugoslav]

It doesnt matter. All AMD FX 83xx and 9xxx processors are the same, and if you overclock them to the same clock, you will get the same performance and wattage usage with every processor( 8320,8320E,8350, 8370E etc) (if I overclock my 8320 to 4.7ghz, I will virtualy get FX 9590 with 220w TDP.


E versions of FX 83xx line processors have low clock, only 3.2ghz, thats why it eats up only 95W, if I downclock my fx 8320 to 3.5ghz, I can run it at 1.118v, and will use only 90 watts.

So all the difference between FX 83xx and 9xxx line is in their clocks.

 

[OneB1t]

It is same on same frequency results from benchmarks should be exactly same

If you saw cinebench bencmark it's clearly NOT the same.

it is

I have 6+2 phase board M5A99FX with useless 1333mhz CL9 memory and it scores 729 @ 4.6ghz so techspot is wrong and they got APM enabled or VRM overheat

 

[Jugoslav]

Bottom line: Tester said that he is aware of FX turbo loads, one is Full turbo load(happens on all of the cores, but doesnt use max cpu frequency) and one is called Half turbo load(happens on half of the cores, but at full speed, happens very rarley), so he disabled turbo, and raised the multiplier, so the clock is the same on all of the cores.


But based on CB r15 test he used 3.2ghz clock and not 4ghz