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

[de5_Roy]

Where i live guys with P4D's and (smthing like 8600GT) try/to play BF3 on low res and settings
Doh..

you should look into toms bf3 benchmarks where dual core cpus were tested and try to estimate P4D performance from there. due to it's age, p4d performance should be quite limiting.

 

[IRONBATMAN]

seems like a comment got removed

 

[TheBigTroll]

only spam is removed

 

[ingtar33]

AMD FX 8350 vs Intel 3570K vs 3770K vs 3820 - Gaming and XSplit Streaming Benchmarks

Interesting results.

http://www.youtube.com/watch?v=eu8Sekdb-IE

I had seen that a while back..I found it interesting as well.Although i am not sure what to think of it. As no other professional site agrreed at least not that i am aware of.

it wasn't that no one agreed. simply put no one benches games WHILE streaming video. That test shows results while streaming video without a capture card. Which is an entirely cpu bound exercise.

All those extra cores on the 8 core fx unsurprisingly helped it deal with the extra load better then the 4 core w/ HT intels. It was a very cpu heavy tasks that needed a lot of cores.

Its a unique test that no one else did or tried and it's an intersting one. I mean a LOT of people do stream their game sessions without capture cards. If you're one of those people you will see a striking advantage with gaming on an 8350.

 

[InvalidError]

Its a unique test that no one else did or tried and it's an intersting one. I mean a LOT of people do stream their game sessions without capture cards. If you're one of those people you will see a striking advantage with gaming on an 8350.

No need for a 'capture card', just need to have video streaming software that supports QuickSync or can delegate rescaling, filtering and most of the encoding to GPGPU to the IGP.

There are more than sufficient additional compute resources available to cover streaming/capture software developers' needs using lower-end CPUs/APUs if they would bother implementing support for them.

 

[stingstang]

Yea, i just got the 6300fx to upgrade over the 4100fx so Rome 2 could play better. It...gives me about 7 more frames. Not very happy about that, even if three game is single-threaded. I'm thinking that my old q9550 would still be faster than this. I "upgraded" to the 4100 from that and didn't show a performance gain.

 

[logainofhades]

Yea, i just got the 6300fx to upgrade over the 4100fx so Rome 2 could play better. It...gives me about 7 more frames. Not very happy about that, even if three game is single-threaded. I'm thinking that my old q9550 would still be faster than this. I "upgraded" to the 4100 from that and didn't show a performance gain.

What graphics card are you running? If your card is the bottleneck, CPU will not matter much, if at all.

 

[Jhutson25]

i3 > 8350 for gaming? Just wow. Tom's Hardware is so paid. There is absolutely no way in hell that this chart is even close to being acurate. Just look at every single comparison from any other sight (specifically tek syndicates review on amd vs. intel). The more accurate description would be 8350 = Haswell i7k. Why can't review sights just be honest and unbiased? I guess when Intel pays you, you would say anything too, I guess. i3 > 8350 is just disgusting, blatant bias.

 

[artk2219]

i3 > 8350 for gaming? Just wow. Tom's Hardware is so paid. There is absolutely no way in hell that this chart is even close to being acurate. Just look at every single comparison from any other sight (specifically tek syndicates review on amd vs. intel). The more accurate description would be 8350 = Haswell i7k. Why can't review sights just be honest and unbiased? I guess when Intel pays you, you would say anything too, I guess. i3 > 8350 is just disgusting, blatant bias.

If you take all those games that only use one or two threads (skyrim, starcraft II, WoW, Diablo III, blizzard in general) and look at the results, the I3 is ahead because of its superior single thread performance. If you have enough of those types of games, or if those games are weighted heavily enough in the bench totals, then yes. You will definitely see the I3 being ahead of the 8350 for the most part (honestly even the Pentium dual core and Celeron can be as well, but no way in hell would I recommend them as gaming CPU's, and personally I believe all of those games should be stripped from the benchies, but thats me, they're still massively popular games). When it comes to anything at all even remotely multithreaded though, the 8350 trounces the I3. Especially if you're gaming and streaming/browsing/have a VM or 3 up/have tons of apps up/etc. Personally when im recommending hardware i try not to recommend I3's, they're chips that work pretty ok now but they will age badly, and you'll probably be wanting another chip in 18 to 24 months. If you're the type of person that upgrades that often, or are the type of person that upgrades the CPU only, then its no problem. But if you want a stable gaming build to last more than 2 years, they're a no go (In my opinion even the FX 4000 series and 4 core APU's are a no go, but they should live a little longer than the I3). In that case I'd at the very least recommend an FX 6300, I5, or better. Anyway back to the original comment, I can see why they place them above the 8350, but in the future, you'll be wishing you'd have gone with a better processor. Also quit with the whole paid off by Intel bit, it makes you sound fan boy and trollish. Im sure they get some sort of kick back from them, but i doubt its more than they get from any other manufacturer/sponsor.

 

[Sakkura]

i3 > 8350 for gaming? Just wow. Tom's Hardware is so paid. There is absolutely no way in hell that this chart is even close to being acurate. Just look at every single comparison from any other sight (specifically tek syndicates review on amd vs. intel). The more accurate description would be 8350 = Haswell i7k. Why can't review sights just be honest and unbiased? I guess when Intel pays you, you would say anything too, I guess. i3 > 8350 is just disgusting, blatant bias.

Tek Syndicate are the ones whose reviews clash with almost all the reliable sites out there. So if you're looking for corruption, that's where I'd look first.

 

[Onus]

i3 > 8350 for gaming? Just wow. Tom's Hardware is so paid. There is absolutely no way in hell that this chart is even close to being acurate. Just look at every single comparison from any other sight (specifically tek syndicates review on amd vs. intel). The more accurate description would be 8350 = Haswell i7k. Why can't review sights just be honest and unbiased? I guess when Intel pays you, you would say anything too, I guess. i3 > 8350 is just disgusting, blatant bias.

If you take all those games that only use one or two threads (skyrim, starcraft II, WoW, Diablo III, blizzard in general) and look at the results the I3 is ahead because of its superior single thread performance. If you have enough of those types of games, or if those games are weighted heavily enough in the bench totals, then yes. You will definitely see the I3 being ahead of the 8350 for the most part (honestly I believe all of those games should be stripped from the benchies, but thats me, they're still massively popular games). When it comes to anything at all even remotely multithreaded though, the 8350 trounces the I3. Especially if you're gaming and streaming/browsing/have a VM or 3 up/have tons of apps up/etc. Personally when im recommending hardware i try not to recommend I3's, they're chips that work pretty ok now but they will age badly, and you'll probably be wanting another chip in 18 to 24 months. If you're the type of person that upgrades that often, or are the type of person that upgrades the CPU only, then its no problem. But if you want a stable gaming build to last more than 2 years, they're a no go (honestly in my opinion the FX 4000 series and 4 core APU's are a no go as well, but they should live a little longer than the I3). In that case I'd at the very least recommend an FX 6300, I5, or better. Anyway back to the original comment, I can see why they place them above the 8350, but in the future, you'll be wishing you'd have gone with a better processor. Also quit with the whole paid off by Intel bit, it makes you sound fan boy and trollish...

This. This illustrates very well the danger of basing purchasing decisions on a single data point, out of context. Someone who enjoys older and/or single-threaded games, relying on TekSyndicate (by my understanding an AMD fanboy site, if not an outright shill) to buy an FX, will be disappointed because for those titles, the i3 is better. OTOH, someone relying solely on a site claiming the i3 beats the FX, but planning to play only the latest and greatest as they appear, will also have good cause to be very angry. Anyone can cherry-pick benchmarks to "prove" what they want; you need to select the ones most relevant to you before you buy.

 

[ajcroteau]

I'm seriously considering not even reading this article anymore... nor the GPU Hierarchy either... Sometimes i think it's based on price, sometimes i think it's based on performance... Why the 8350/8320 is not in Tier 1 is beyond me... You just can't beat the price vs performance on an AMD FX-8350...

 

[logainofhades]

I'm seriously considering not even reading this article anymore... nor the GPU Hierarchy either... Sometimes i think it's based on price, sometimes i think it's based on performance... Why the 8350/8320 is not in Tier 1 is beyond me... You just can't beat the price vs performance on an AMD FX-8350...

Sure you can, FX 8320. 😛

 

[nokiddingboss]

I'm seriously considering not even reading this article anymore... nor the GPU Hierarchy either... Sometimes i think it's based on price, sometimes i think it's based on performance... Why the 8350/8320 is not in Tier 1 is beyond me... You just can't beat the price vs performance on an AMD FX-8350...

just wait early to mid next year. that's when the i3 will be shuffled to fit your expected placement. it wasn't too long ago when nearly all of amd's cpu were at the same tier as a pentium SB.

 

[ingtar33]

i3 > 8350 for gaming? Just wow. Tom's Hardware is so paid. There is absolutely no way in hell that this chart is even close to being acurate. Just look at every single comparison from any other sight (specifically tek syndicates review on amd vs. intel). The more accurate description would be 8350 = Haswell i7k. Why can't review sights just be honest and unbiased? I guess when Intel pays you, you would say anything too, I guess. i3 > 8350 is just disgusting, blatant bias.

Tek Syndicate are the ones whose reviews clash with almost all the reliable sites out there. So if you're looking for corruption, that's where I'd look first.

not to defend teksyndicate but their methodology, while unorthodox seemed pretty solid. The problem is the AMD & Intel fanboys took their testing to say something it didn't.

All teksyndicate was doing was running a very specific test, which was live streaming of game benches in certain games... so they weren't just benching the games with their rigs, they were streaming those benching sessions at the same time without a video capture card.

So they weren't just playing crysis3 or farcry3, they were streaming it too.

As everyone here knows who's tried it, video streaming without a capture card is a very cpu intensive task that eats up full cores. Well guess what. those intel chips don't have a lot of cores to dedicate to video streaming. And the games they were benching were mostly multithreaded, which already leans a bit more in AMD's favor.

Teksyndicate wasn't claiming that the fx8350 was better then intels. if you watch the commentary in the original and follow up video he makes it clear a number of times. All he was saying was the fx8350 was a VIABLE gaming cpu, that generally plays about as good as an intel and in some cases is even a better option.

Which is something people on this site say all the time. The fx lineup, even the athlon II or old core2 lineup or i3 lineup or even the apu lineup all have viable gaming chips that will game fine depending on needs, budget and games.

I think people really were only going nuts about those videos because they didn't notice he was streaming video as part of the bench. so they thought he had magically found some secret way to make an 8 core fx beat some intel cpus. Well that 8core fx DOES beat intels in some tests and games... it's always been the case there were a few situations it would beat an intel. he wasn't saying anything ground breaking.

Its just he made the mistake (IMHO) to talk about the scheduling fix in windows for the fx cpus when talking about his testing methodology. It made some implication to people who weren't familiar with that patch that a long running windows patch was somehow missing from all the other benching sites reviews and that was why he has the fx winning the benches. And for those who couldn't bother to listen to the commentary, his results seemed to "expose" other benching sites as having flawed methodologies.

however he never made any of those claims and as a result those videos which i think are boarder-line entertaining/informative have become synonymous with poor methodology. Its a shame too... cause i've found a number of his videos and reviews to be rather entertaining and at times informative.

 

[Onus]

Ingtar, those are all very good points. When you consider the addition of the streaming workload, the "moar cores" approach of AMD works well.

 

[nokiddingboss]

i3 > 8350 for gaming? Just wow. Tom's Hardware is so paid. There is absolutely no way in hell that this chart is even close to being acurate. Just look at every single comparison from any other sight (specifically tek syndicates review on amd vs. intel). The more accurate description would be 8350 = Haswell i7k. Why can't review sights just be honest and unbiased? I guess when Intel pays you, you would say anything too, I guess. i3 > 8350 is just disgusting, blatant bias.

Tek Syndicate are the ones whose reviews clash with almost all the reliable sites out there. So if you're looking for corruption, that's where I'd look first.

not to defend teksyndicate but their methodology, while unorthodox seemed pretty solid. The problem is the AMD & Intel fanboys took their testing to say something it didn't.

All teksyndicate was doing was running a very specific test, which was live streaming of game benches in certain games... so they weren't just benching the games with their rigs, they were streaming those benching sessions at the same time without a video capture card.

So they weren't just playing crysis3 or farcry3, they were streaming it too.

As everyone here knows who's tried it, video streaming without a capture card is a very cpu intensive task that eats up full cores. Well guess what. those intel chips don't have a lot of cores to dedicate to video streaming. And the games they were benching were mostly multithreaded, which already leans a bit more in AMD's favor.

Teksyndicate wasn't claiming that the fx8350 was better then intels. if you watch the commentary in the original and follow up video he makes it clear a number of times. All he was saying was the fx8350 was a VIABLE gaming cpu, that generally plays about as good as an intel and in some cases is even a better option.

Which is something people on this site say all the time. The fx lineup, even the athlon II or old core2 lineup or i3 lineup or even the apu lineup all have viable gaming chips that will game fine depending on needs, budget and games.

I think people really were only going nuts about those videos because they didn't notice he was streaming video as part of the bench. so they thought he had magically found some secret way to make an 8 core fx beat some intel cpus. Well that 8core fx DOES beat intels in some tests and games... it's always been the case there were a few situations it would beat an intel. he wasn't saying anything ground breaking.

Its just he made the mistake (IMHO) to talk about the scheduling fix in windows for the fx cpus when talking about his testing methodology. It made some implication to people who weren't familiar with that patch that a long running windows patch was somehow missing from all the other benching sites reviews and that was why he has the fx winning the benches. And for those who couldn't bother to listen to the commentary, his results seemed to "expose" other benching sites as having flawed methodologies.

however he never made any of those claims and as a result those videos which i think are boarder-line entertaining/informative have become synonymous with poor methodology. Its a shame too... cause i've found a number of his videos and reviews to be rather entertaining and at times informative.

i agree. i too do think that the problem was that most people didn't understood what logan was saying. they tend to take it like - if the fx8350 is as viable as... let's say an i5 in gaming, the sheeps automatically think that what he is saying is that the fx8350 is the same as an i5/i7 in terms of gaming performance in every game. which was clearly not. viable does not mean equal, it means that its a good option, a very good option IMHO. and at that price point, i cannot possibly disagree with him.

 

[caamsa]

You all ever gonna fix the comments so they only show the most current ones? Some problem for the GPU article. It's annoying.

 

[ronch79]

Is Cleeve employed by Tom's Hardware? It seems like these guys CAN'T take criticism for what they write around here. Why so defensive? Not a good attitude, guys. You shouldn't be shooting your readers down just because they criticize your articles. After all, quite honestly I've shifted away from TH because the quality of journalism has somewhat gone down compared to other tech sites, and it's not just me, mind you. I read posts by other readers on other sites sharing the same sentiments. Just saying. Take it the way you guys wanna take it.

 

[SessouXFX]

I could care less what the FPS is, if:

* My game plays smoothly w/decent settings (I don't need all of the eye candy).
* Doesn't cost me too much to implement.
* Works with existing hardware.
* Plays at a decent resolution.

right on man that is my point people over exaggerate hardware demand like i said if your game is smooth as butter and the graphics are looking great and way better than console what does it matter because if you see the game running fast and smooth what more is 144 ps going to do ? I'll enjoy My games while people spend a lot of money and time jerking off to more hardware and the guy above said your only a casual gamer if you get 30-60 fps lol what do they expect 60 fps puts the icing on the cake doesn't it lmao

so why not just play on a console?

AUURGH! BRIGHT LIGHT! BRIGHT LIGHT!

 

[overlandpark4me]

i5 3570K is 189.99 all day at Microcenter if you're fortunate to have one near you like I am.

 

[littleleo]

That is where, what, and how much I paid. They had them on sale for a short time at $169, but last time I looked they are back at $189.

 

[tsnor]

Hi, Any idea why an I7-920 is considered inferior to a Phenom II X4 Black Edition 980 ? Test results seem to say the opposite, e.g. http://alienbabeltech.com/main/introducing-the-new-phenom-ii-980-be-vs-core-i7-920-value-meets-performance-again/32/

 

[Sakkura]

Hi, Any idea why an I7-920 is considered inferior to a Phenom II X4 Black Edition 980 ? Test results seem to say the opposite, e.g. http://alienbabeltech.com/main/introducing-the-new-phenom-ii-980-be-vs-core-i7-920-value-meets-performance-again/32/

Because...

this hierarchy chart is useful as a general guideline, but certainly not as a one-size-fits-all CPU comparison resource.

 

[logainofhades]

Hi, Any idea why an I7-920 is considered inferior to a Phenom II X4 Black Edition 980 ? Test results seem to say the opposite, e.g. http://alienbabeltech.com/main/introducing-the-new-phenom-ii-980-be-vs-core-i7-920-value-meets-performance-again/32/

Because the charts are a mess and needed to be fixed months ago. They can call it a general guideline all day long. The reality is, that it is a broken mess. There isn't a single scenario where a PhII is superior to an i7 920 when both are at stock clocks.