Intel Core i3 4340 Review – Beating AMD At Their Own Game

[drunkenmaster]

Locked vs unlocked, who cares, the numbers are bullcrap and it takes a numpty to not see it.

Xbitlabs, who I would trust show a i5 4430(3.2Ghz turbo, 1.1Ghz turbo on gpu) vs a 6800k and fps in Metro last light is 22.9 fps vs 30fps respectively, the 6800k is almost 40% faster.

In this review a i5 4670(3.6Ghz turbo, 1.2Ghz on the gpu) gets 38.5fps vs the 6790k's 23fps.

It's funny that on a dodgy site with BS numbers, incredibly similar cpu's the reputable seen for years site has the AMD igpu 40% faster, yet this site has the Intel igpu almost 70% faster.

That is quite the swing, lets just say, of all the numbers I've seen on various websites, the Xbit numbers match what I see almost everywhere else, this reviews numbers I've never seen come close on any other review ever.

The entire thing is BS start to finish, Boom posted it because.... god knows, it's clear as day what BS it is.

6790k vs 6800k, it's 100Mhz slower and has the same gpu and same gpu clocks. The 4770k has higher base and turbo speed, higher gpu turbo speed and double the threads vs a i5 4670. So the 6790/6800k are almost identical, the 4770k is categorically faster than the i5 4670.....

http://www.anandtech.com/show/7032/...-gpu-on-the-desktop-radeon-hd-8670d-hd-4600/2

The 6800k spanks the 4770k igpu, yet this magical review has a slower cpu/gpu combo smashing the AMD igpu to pieces.

This review is 100% bs start to finish, the really hilarious thing is, Boom STILL went out of his way to post the power comparison in Cinebench without posting the fx8350 spanking the i3 Cinebench benchmark image to go along with it.

 

[Boomstick777]

More trolling, but with a disclaiming of "i'm not trolling" which apparently makes trolling immune,also if you say something like "I'm not trolling, I want to buy produce xxxx from AMD therefore I can't be a troll".

Yeah because anything that doesn't paint AMD CPU's in the best light is trolling.

Just posted the review matey, you do know I like AMD right, have many AMD components? This exact review has been posted on other forums as well. The OP's their aren't being accused of trolling. This is your issue, being defensive again.

An i3 shouldn't be beating a FX 8350 in gaming, it shows just how far AMD are behind. No point covering it up, let's hope they release new chips and don't just put all there hopes all Mantle.

 

[Boomstick777]

AMD really need to get competitive in the CPU market again :\

Intel are gonna rip all our balls off if they don't.

This ^^.

I'd never pick an i3 over an fx8320, that said.
People rant on and on about now much cheaper it costs for an fx8320 over clocked system is over an over clocked i5....

But then when it comes down to the fx8320 being the more expensive in the way of against the i3, it's ignored and the fx8320 needs to be over clocked and at a higher base set up to 'fairly' compare? Pull the other one.

But what do I know, I'm blates a fanboy, because I have an Intel CPU, whereas everyone with their AMD chips are totes unbiased.

Skewed opinions and hypocrisy galore more like.

That said, I'm with cat somewhat on the results lol, I know that an i3 2120 is capable of, and I see that bottlenecking a 7870.

Haha, 'Skewed opinions and hypocrisy galore more like' totally agree with your post. I'm with you on the older first gen i3's maybe holding things back. I don't know how that i3 2120 would perform so can't say.

It's really the Haswell low end CPU's that have become so good that they don't really bottleneck single GPU's at all. I don't know whether the newer mobo's have a part to play, but all the Pentium Haswell builds I have done have being surprisingly quick, power use is low as well. That really they are a good option for budget gaming rigs

 

[mmj_uk]

Comparing a locked CPU with a factory clocked unlocked CPU = rubbish. Where was the part where they overclocked the 8350 to 4.8ghz or more?

That's what I always wonder when the local AMD crowd talk down the 4670K based on its non-overclocked performance.

 

[CAT-THE-FIFTH]

GameGPU have just tested the latest version of BF4:

http://gamegpu.ru/images/remote/htt...attlefield_4_China_Rising_-test-bf_4_proz.jpg

FX6300 is around Core i5 2500K level performance. HT improves performance quite a bit and pcgameshardware already tested the effect of HT on framerates.

In AC4,the FX6300 handily beats the Core i3 2100 in minimums:

http://gamegpu.ru/images/remote/htt...ssassins_Creed_4_Black_Flag-test-ac4_proz.jpg

It means an IB Core i3 should be around the same,and the Haswell a bit faster overall.

They have now added the Core i3 4330 to their new articles:

http://gamegpu.ru/images/remote/http--www.gamegpu.ru-images-new-msi-test-wl_2_proz.jpg
http://gamegpu.ru/images/remote/htt...GPU-MMO-World_of_Tanks_0.8.10-wot_proze_3.jpg

Wasteland 2 uses 3 threads heavily.

Even WoT,the FX6350 should be around the same as the FX8350 and costs the same as a Core i3 4330,and that game uses 2 threads.

 

[Boomstick777]

So if I'm looking at this right, in WOT the i3 4330 lower clocked than the 4340 in this thread, beats the FX 8350 and FX 6300 etc, but in Wasteland i3 max fps is same as FX 8350 but has lower bottom -3fps. That concurs with the review in this thread, the 4340 is a bit higher clocked so has more of a lead.

This i3 is pretty impressive tbh, consider how little power use it uses to achieve those fps, whereas the amount the FX consumes. Intel's chips are very impressive. I wander if 1150 mobo's come into play with improving performance..

 

[CAT-THE-FIFTH]

If a Core i3 in games which suit its advantages(heavily loaded two cores) is not even that much faster than an FX6300/FX6350,then in multi-threaded games it is not really going to do much better is it?? Moreover,those BF4 results contradict the ones in your article,and are closer to what people are saying. The thing is the FX6300 is cheaper than a Core i3 4330 and the Core i3 4130 is going to be a few percent slower still.

I am a SFF PC fan,and still those power consumption arguments are not really that relevant for full sized rigs(I was running an overclocked high VID Q6600 on a 975X in a Shuttle and a overclocked 8800GTX 512MB for example). Remember a FX8350 consumes around 30W to 95W more than a IB or Haswell Core i5 or Core i7 at the wall under gaming conditions. That is on a 990FX chipset which consumes a lot of power(not the midrange 970).

You would need to game 40 hours a week,for all 52 weeks of the year to even break £2/month more in extra power consumption at 15P to 16P a Kilowatt hour at the upper range(95W).

40 hours is the equivalent of 4 hours each weekday and 10 hours each day of the weekend,and you would need to do that for 52 weeks. Game less than 40 hours a week and not every week of the year and it is even less. That is assuming a heavily threaded game too. Lightly threaded games will show less of difference IMHO.

Its most likely,that with the amount of time many of us have,20 hours a week would be doable TBH!!

 

[Boomstick777]

If a Core i3 in games which suit its advantages(heavily loaded two cores) is not even that much faster than an FX6300/FX6350,then in multi-threaded games it is not really going to do much better is it?? Moreover,those BF4 results contradict the ones in your article,and are closer to what people are saying.

Contradicts? The benchys you showed were for the lower clocked chip, that chip still beats the FX 8350 in 1 benchy and matches it in the other and beats the FX 6300 in both? The chip from this thread would be further ahead? That concurs with not contradicts.

More fps at less than half the power use, this is def a win for Intel anyway you look at it. I hope the under £50 Intel Pentium G3420 gets reviewed soon, like to see how it matches the FX 8350 in gaming.

 

[linktoinsanity]

Even WoT,the FX6350 should be around the same as the FX8350 and costs the same as a Core i3 4330,and that game uses 2 threads.

WoT?

If you are referring to World of Tanks, no, that game uses a single thread.

If you are referring to another game then please let us know.

I also have to laugh at the hypocrisy evident in this thread.

 

[drunkenmaster]

Yeah because anything that doesn't paint AMD CPU's in the best light is trolling.

Just posted the review matey, you do know I like AMD right, have many AMD components? This exact review has been posted on other forums as well. The OP's their aren't being accused of trolling. This is your issue, being defensive again.

An i3 shouldn't be beating a FX 8350 in gaming, it shows just how far AMD are behind. No point covering it up, let's hope they release new chips and don't just put all there hopes all Mantle.

You do know an i3 shouldn't be almost matching an i5/i7 in gaming... but it can........

Once again, I posted the image you left out.

The power usage chart you showed is power usage in CINEBENCH, NOT in gaming.

In Cinebench the fx8350 was almost 100% faster than the i3. Comparing power levels in a benchmark in which the FX8350 not only utterly utterly trashed the i3, but beat the i5 by 10-20% as well, then claiming it uses double the power for less gaming performance is one of two things, trolling, or ignorant.

If you assume a chip always uses the same amount of power, you're trolling or ignorant. If you can't see that the particular games they are showing are basically not using more than 2 threads, based on the fact that the i5 with 2 more cores is performing almost identically.... you get the picture.

The fx8350 is not miles behind the i3 OR the i5 and IIRC, costs less than both. You're comparing an i3 at 100% load, to a FX8350 in which 50-75% of the core is UNUSED and claiming the i3 is faster..... once again, trolling/ignorant, interchangeable aren't they. Do you know how much faster a 4770k is with one thread used vs an identically clocked i3 Haswell........ within a few percent. Do you know how much faster that 4770k is when it uses 4 threads in something that scales well, pretty much 100% faster, in something that uses 8 threads it might end up 150% faster. By your own argument the 4770k is no faster than a i3 4430 in gaming because in a game which uses 2, maybe 3 threads, they'd be the same speed at the same clocks....... but you also want to say the 4770k trounces a 4670k, and i3's, and is the best value chip around and the best chip for gaming.

This review categorically isn't showing what you think it is and as shown with Anandtech and Xbit labs the IGPU results are 100% made up on that review, which also throws in to question the results in every other benchmark there.

 

[CAT-THE-FIFTH]

Contradicts? The benchys you showed were for the lower clocked chip, that chip still beats the FX 8350 in 1 benchy and matches it in the other and beats the FX 6300 in both? The chip from this thread would be further ahead? That concurs with not contradicts.

More fps at less than half the power use, this is def a win for Intel anyway you look at it. I hope the under £50 Intel Pentium G3420 gets reviewed soon, like to see how it matches the FX 8350 in gaming.

Some of use(like me) have used Intel for years(poor AMD support for the ITX form factor for the last 6 years,essentially meant that),but still perhaps you should look at those BF4 results and the results of multiple BF4 reviews. See how the Core i5 4670K obliterates the Core i3. The FX6300 matches a Core i5 2500K. That is how it should be for the CPU intensive parts of the BF4 single player game. I also run Crysis3 and those results in your review come from corridor areas,and not the wide open maps. Moreover those 2 games(Wasteland 2 and WoT),are the best possible situation for a Core i3,and the performance difference is not as big as I would expect.

Morever,going on about power usage while have a Core i7 4770K and a R9 290 is hilarious. I should be the one bitching and moaning about power consumption(considering I only use SFF PCs),but it really is hardly anything in reality for anything other than people who spend 40+ hours a week gaming for the entire year. Not even my most gaming mad mate with 600+ games on Steam can do that consistently.

How many hours do you play games each week on average??

WoT?

If you are referring to World of Tanks, no, that games uses a single thread.

If you are referring to another game then please let us know.

I also have to laugh at the hypocrisy evident in this thread.

Look next time and check those test results. The FX8350 runs at 4.2GHZ single core Turbo. The FX6350 runs at 4.2GHZ single core Turbo. The A10 6800K runs at 4.4GHZ Turbo. The FX4320 runs at 4.2GHZ single core Turbo. They will all have similar single thread performance(CB 11.5 score are similar),which would place them all at 6.3% and 10% lower minimums than a Core i3 4330:

http://gamegpu.ru/images/remote/htt...GPU-MMO-World_of_Tanks_0.8.10-wot_proze_3.jpg

That is in a very lightly threaded game which is an advantage for a Core i3.

The FX6350,A10 6800K and FX4320 are all Core i3 4330 level pricing. The FX6300 is below usual Core i3 level pricing at many retailers.

When Kaveri is out,I suspect it will do better than of the current AMD CPUs in lightly threaded situations(even if it is only 10%).

WoT is the best possible scenario for a Haswell Core i3. I would expect more.

Moreover,I suspect you are considering this earlier situation with the game:

http://gamegpu.ru/images/stories/Test_GPU/MMO/World of Tanks 0.8.5.0/wot proz.jpg

However,it looks like better optimisations for AMD CPUs in the latest patch. The performance gap is much lower than it used to be.

I remember,another game(one which I played a lot),called Planetside 2. It used to use mostly two cores,and meant higher clockspeed Intel CPUs pulled ahead(but performance was still dire). Two patches later,and now it easily uses 4+ threads,with over a 40% increase in framerates for my Xeon E3,and better optimisations for AMD CPUs as a result. People need to keep upto date with the patches with games. They can massively change the peformance scaling with different CPUs.

Edit!!

Anyway,I will have to agree to disagree with you both,otherwise we will just go around in circles.

 

[Boomstick777]

Morever,going on about power usage while have a Core i7 4770K and a R9 290 is hilarious. I should be the one bitching and moaning about power consumption(considering I only use SFF PCs),but it really is hardly anything in reality for anything other than people who spend 40+ hours a week gaming for the entire year. Not even my most gaming mad mate with 600+ games on Steam can do that consistently.

How many hours do you play games each week on average??

Yeah Cat, I was referring to the architecture, what I was saying is that the performance relative to power consumption is miles ahead of AMD, and in that sense the i3 is a lot better. Higher fps for lower power use. If I was building a budget gaming PC I would def go with the Haswell Pentium. If I was building a high end gaming rig I would go with 4770K. The architecture is just so far ahead at this point..

I do agree that in the future we will see more optimization for load spread across more cores, right now though, in the latest games you have an dual core i3 beating an FX 8350 in fps while using less power. That's very impressive for Intel, not so great for AMD. I hope future hopes aren't just on Mantle, I would like to see AMD launch more power efficient hardware, higher IPC, CPU parts. That's just my opinion though, I know some are happy with the current FX chips...

 

[R3X]

I hope AMD come back in style with Mantle but ain't banking on it.

Ones thing is for sure we can't just have intel, AMD needs to exist and I hope they release something for everyone.

 

[Jarrastafari]

Perhaps the most surprising is the i3 beating AMD's Richland APU in BF4

Huh Isn't it the other way around?

 

[Boomstick777]

I hope AMD come back in style with Mantle but ain't banking on it.

Ones thing is for sure we can't just have intel, AMD needs to exist and I hope they release something for everyone.

Same here, I'm secretly hoping they just don't want to hype their next release to much, so are remaining silent, come on AMD give us 8 core SR chips that can put a real whooping on Intel

 

[bru]

You guys do realise that if your going to say the fx8350 should have been overclocked, then it would only be fair to run the i3 4340 with the optional 4 additional turbo boost bins taking it up to 4GHz.

You still sure the FX would be so much better?

 

[ALXAndy]

You guys do realise that if your going to say the fx8350 should have been overclocked, then it would only be fair to run the i3 4340 with the optional 4 additional turbo boost bins taking it up to 4GHz.

You still sure the FX would be so much better?

I3 do not have turbo.

http://ark.intel.com/products/77480

Only I5s and I7s had the limited unlock and I'm not sure Haswell support the limited unlock any more either. I3s are completely and utterly derped unless you touch the FSB, and good luck with that.

And yes, I'm still sure it would be that much better in heavily threaded apps.

 

[drunkenmaster]

Huh Isn't it the other way around?

Yes, every other respectable review shows the 6800k gpu smashing the HD4600 to pieces, full stop, no exceptions. The 4770k has a faster igpu, faster cpu and more threads available and it gets smashed, not a chance in hell the slower i5 4670 could be faster, and no question the i3 4430 isn't faster either, hence the entire review is BS.

Boom likes to ignore the bits that don't fit in with his version of events though.

He seems to think that the identical core in a 4770k that is in the 4430 shouldn't have higher single thread performance than the FX8350.

He seems to believe that if the 4770k beats the FX8350 in single thread performance, that the 4430 should be miles behind it(despite using identical technology) and thus the FX8350 is being embarrassed somehow and it "should not happen".

Yeah, you get a game that uses 8 threads well, the fx8350 will spank the i3 4430 silly, as will the 4770k as will the i5 4670. But Boom wants to make whatever point it is and ignores these simple facts. A couple low threaded games, some made up benchmarks and using the higher power usage in Cinebench NOT gaming, as proof the FX8350 is crap.... despite the FX8350 being almost 100% ahead of the 4430 in Cinebench where it's power usage is significantly higher.

Fact's and logic have never really gotten in Boom's way before and are unlikely to in the future.

Some people seem to forget the very fundamental design of Bulldozer was adding the second module for a 5-10% die space increase for the module, hence the second core every every module is insanely cost effective. It's that 5% die space increase that allows in 8 threaded situations, Bulldozer and even more often Piledriver to beat a 4 thread 4 core i5 in MANY situations and MANY games.

 

[ALXAndy]

I hope AMD come back in style with Mantle but ain't banking on it.

Ones thing is for sure we can't just have intel, AMD needs to exist and I hope they release something for everyone.

Jesus man, don't be so pessimistic !

The simple fact is (and it is fact, as much as people don't want it to be) there is nothing wrong with AMD FX CPUs. There never really was tbh (even the Bulldozers). The problem was always poor software support for them. If you take a look at Bulldozer reviews you'll note that in some tests (like Winzip and encoding) they were actually really good. In games? less so, but that was then.

Since then AMD have addressed the issue with low thread counts by increasing the IPC enough to make PDs fine even in games that do not use all of their threads. In stuff like Cinebench the FX 8s actually outperform the 4670k (even overclocked) so again PD is more of the same. It needs software support ! so what do you do as a company? well instead of wasting more and more money investing in CPU technology you simply get the software to support what you already have, queue Mantle.

But it's not just Mantle. HSA is also something they're working on all of which will make a better future for them. The CPU IPC "core race" is over and has been for a while.

 

[bru]

I3 do not have turbo.

Fair enough I wasn't aware that was the case.