Best Gaming CPUs For The Money: January 2012 (Archive)

[tanh]

I'm with blaz on that it's too risky to purchase an Athlon II X3 and expect that it's locked core works perfectly.

You can buy Athlon x3 slightly more $ but with 100% guarantee unlock.

 

[de5_Roy]

i've seen two crysis 3 cpu benches(techspot and that russian website) so far, both report average fps, both use a high end nvidia gpu(680 and 690) ironically crysis 3 being part of amd's games bundle promotion. both showing fx8350 keeping up with core i5 3570k/3770k well. i keep wondering why they didn't use a 7970 ghz ed.
if the fx8350 stutters, there will be crying over not testing 4module/4core mode which will miraculously improve smooth gameplay.
still, at least the game seems to use 8 integer clusters. only the russian site showed task manager usage.

 

[SuperVeloce]

Is there even a possibility to turn off only one core in a module? As far as I know, you can only turn off whole modules

 

[mousseng]

Is there even a possibility to turn off only one core in a module? As far as I know, you can only turn off whole modules

Depends on your motherboard and BIOS - many of them have the ability to select individual cores to disable.

 

[Intel_shills_uncovered]

I actually performed the tests and results are as reported. I've tested FX vs Sandy and Ivy bridge since they were introduced. I have never seen the 8350 outperform the i5-3570K in games.

[EDIT: I should mention that our own Crysis 3 performance article is coming out in the near future and, indeed, it's one of the very few games where the FX-8350 produces similar average frame rates compared to the i5-3570K.

**HOWEVER**, the minimum frame rates are very different (something Tek Syndicate doesn't publish), with Intel having a significant lead while the FX gets pulled down to stuttering. Furthermore, this is one of the only games I've ever seen where the Core i7 has a significant advantage over the Core i5. Crysis 3 is not your typical game, and I don't expect other titles to reflect this state of affairs on average for a few years, if ever.

As for Far Cry 3, our own bench results showed that the FX-8350 performed very close to the i5-3550, which has virtually the same clocks as the 3570K at stock.

To summarize, Tek Syndicate has used the two games most favorable to the FX series in this particular article. The irony is, their results don't stray far from our own in those two games, which are statistical outliers. The difference is, we tested more games than they did, they were only looking at Crysis 3 and Far Cry 3.

After performing even cursory fact checking, your suggestion of impropriety appears even more ludicrous. /EDIT]

And you installed the amd hotfix right? And you weren't running at 800x600 right?
Logan has already tested many different games to you know. http://teksyndicate.com/videos/amd-fx-8350-vs-intel-3570k-vs-3770k-vs-3820-gaming-and-xsplit-streaming-benchmarks

 

[cleeve]

And you installed the amd hotfix right? And you weren't running at 800x600 right?
Logan has already tested many different games to you know. http://teksyndicate.com/videos/amd-fx-8350-vs-intel-3570k-vs-3770k-vs-3820-gaming-and-xsplit-streaming-benchmarks

Hotfix installed, at 1080p. You can see the resolutions and settings for yourself soon when it's published.

I don't know how to say it other than those results go against years of consistent game testing results from multiple websites.

I have no idea how they've managed to get poor relative performance from a i5.
If I had to guess, something might be wrong with their 1155 platform.

If you can afford it, get the gear yourself and let us know your own results. Because they will not show the FX faster than an i5 at games.
It's just not going to happen.

If you can't afford it, and you're still convinced there's a conspiracy afoot, you're invited to come to my lab, check out the rigs, and try the tests for yourself.
I'll let you choose the motherboard settings and reinstall the OS if you like. We can use the cheapest budget Intel board and the most expensive premium AMD board. It won't make any difference. I've done it dozens of times myself with the same results on multiple platforms.

FX is slower than the i5 at gaming. Plain and simple.

 

[bigcyco1]

As always intel ftw!

 

[Intel_shills_uncovered]

both of which use very similar engines.

Stopped reading right there.

 

[Sakkura]

Stopped reading right there.

Far Cry 3 uses the Dunia 2 engine, which is derived from the Cryengine series that Crysis 3 is based on. The engines are related.

 

[Intel_shills_uncovered]

Far Cry 3 uses the Dunia 2 engine, which is derived from the Cryengine series that Crysis 3 is based on. The engines are related.

And if you did your research Far cry 2's Dunia 2 engine only had less then 2% of code reused from the original Cryengine 1.

 

[cyberskull]

I suggest first test processors then make a chart, new celeron g1610 and g1620 are faster in every game I tried (over 20 games) than pentium g630

 

[chewy1963]

[citation][nom]Intel_shills_uncovered[/nom]And if you did your research Far cry 2's Dunia 2 engine only had less then 2% of code reused from the original Cryengine 1.[/citation]

Who said anything about cryengine 1? What Sakkura said was that Far Cry 3 uses Dunia 2 engine from the same series that Crysis 3 is based on. So the real question is how much code is in common between Durnia 2 and the Crysis 3 engine? Answer is a heck of alot more than 2%.

Trolls like the trolling way they troll.

 

[NdieZ]

i think i5-3570K is still the best bang for the buck in terms of gaming cpu.

 

[ml413]

Nice article--think I noticed a few glitches: the i5 3350p is shown in the chart as having HD 2500 graphics at 77-watts (the article says it slips in under 70w with no HD graphics). Also, the i7 3770k is listed as 4(4) cores/threads, which should be 4(8).

 

[coolronz]

still loving my i7-3820, 50-60fps Crysis, high settings in eyefinity.... but im 3way xfire HD 7970 😀 I was waiting for the ivy bridge-E, but I don't think its coming honestly...

 

[Onus]

The bars show the FX-4300 is weaker than the PhII X4 965, yet is a tier higher?

 

[de5_Roy]

(Ed.: Tom's Hardware editor Paul Henningsen is working on a story that will either confirm Don's opinion or contest it).

could paul test the competing amd cpus (if there is any) for overvolting by the motherboard? i've had a suspicion that overvolting and multicore enhancement might be the reason why amd cpus (and to a lesser extent intel core i5) performed so well (especially in latency) in the sub $200 gaming a/cpu roundup and pentiums and core i3 got the unfair deal. i am all for 4 core cpus taking charge of gaming pcs on every price levels, but on a more even ground.

 

[edlivian]

The i7-3770k finally got some recognition!!!!!

 

[barto]

Maybe I'm dumb or can't read, but doesn't the 3770k have 8 threads not 4? Isn't that the difference between the i5's and i7's?

 

[UltimateDeep]

Can you add Mobile Processors to the list please? I want to gauge how far or competitive the Mobile CPUs is to the Desktop Counterparts.

 

[blazorthon]

[citation][nom]UltimateDeep[/nom]Can you add Mobile Processors to the list please? I want to gauge how far or competitive the Mobile CPUs is to the Desktop Counterparts.[/citation]

To be fair, it should be easy to gauge where they are if you compare core count and clock frequency with a desktop CPU of the same architecture.

 

[blazorthon]

[citation][nom]de5_Roy[/nom]could paul test the competing amd cpus (if there is any) for overvolting by the motherboard? i've had a suspicion that overvolting and multicore enhancement might be the reason why amd cpus (and to a lesser extent intel core i5) performed so well (especially in latency) in the sub $200 gaming a/cpu roundup and pentiums and core i3 got the unfair deal. i am all for 4 core cpus taking charge of gaming pcs on every price levels, but on a more even ground.[/citation]

How could overvolting do anything? Voltage shouldn't affect performance. Clock frequency is what matters when it comes to that and overvolting alone (even if that's what happened) shouldn't affect that except when overclocking. Furthermore, in relevance to the quote that you replied to from the article, I don't understand why the i5s would have any such advantage.

Within this context, they only have an advantage in core count, cache, Turbo, and Turbo/multiplier overclocking (where applicable for the multiplier overclocking since not all i5 CPUs have unlocked nor even partially unlocked multipliers), not an advantage in motherboards.

 

[dscudella]

At what price point does the i5-2500K beat the i5-3570K on a performance to cost ratio? $15 difference? $30 difference?

Not trolling, just wondering.

 

[blazorthon]

At what price point does the i5-2500K beat the i5-3570K on a performance to cost ratio? $15 difference? $30 difference?

Not trolling, just wondering.

Their performance is extremely similar, at least for overclocking. I'd say that the price difference is just how much less you want to pay to not think about getting the few extra features and somewhat lower power consumption.

 

[captain_jonno]

Still don't see the need to upgrade from my i7-930. Since Charts only show it at 2.8 ghz and it has been happily at 4 ghz for quite some time now.