Intel Core i5-3570K vs. i7-3770K Ivy Bridge

[niko084]

Aw, thank you! Something else I can point out to show people they don't need an i7 to play games.
What I would be interested in is how an ivy bridge i3 stacks up

 

[Nordic]

Even for what those extra threads give you, an i7 just does not seem worth it.

 

[Sah7d]

Nice review, what cooler did you use for the test ?
Becouse I imagine that with a $40 cooler the problem would be more than solve isn´t ?

 

[FireKillerGR]

nice review, what cooler did you use for the test ?
Becouse i imagine that with a $40 cooler the problem would be more than solve isn´t ?

i used prolimatech megahalems with a sunon 120mmx38mm

^^

 

[Mals]

Aw, thank you! Something else I can point out to show people they don't need an i7 to play games.
What I would be interested in is how an ivy bridge i3 stacks up

You say this...but these extra threads/cores give you a HUGE advantage in many of the games we play each day. Benchmarks alone are testing a very specific environment that are not very CPU dependent. Games like MMO's, online-FPS's, etc all require significant CPU grunt and the difference in both threads, clocks, cache, etc make significant gains, ESPECIALLY in minimum FPS. The old "put a rig together and throw a benchmark run of a singleplayer game" only tests the GPU horsepower, it does almost nothing to test real-world CPU stresses.

 

[manofthem]

I just grabbed a 3570k and a 212 Evo for my brother's PC build, and I'm looking forward to seeing what it can do. I'm hoping temps aren't too bad as I'd like to get a 24/7 stable mild overclock.

Nice review, it was good stuff

 

[HammerON]

Great review and keep them coming

 

[Mals]

I just grabbed a 3570k and a 212 Evo for my brother's PC build, and I'm looking forward to seeing what it can do. I'm hoping temps aren't too bad as I'd like to get a 24/7 stable mild overclock.

Nice review, it was good stuff

Temps are honestly totally fine, especially with a cooler like that, until you start getting heavy OC's. You have to really start pushing it to be worried about temps. On my okay-binned i5-3570k I hit 4.2ghz before I even had to bump up the voltage. Follow an overclock guide and you will be fine. I stabilized my 3570k at 4.4ghz with barely a bump in voltage and with a Corsair A70 never saw over 80 degrees while stress testing.

 

[Ikaruga]

any other games that you can recommend for cpu testing?

Crysis3 for sure, perhaps Left4Dead2 (but it runs too fast on these new systems, so it's for bench only)

The thing is that there are many game engines out there which uses HT efficiently (especially with ivy bridge where some of the resources are not shared anymore): Dunia (FarCry2), Frostbite (BF:BC2), Rage Engine (GTA4, but this one is far from being efficient in anything expect killing the CPU), etc.., but the problem is that GPUs are too weak at 1080p with ultra settings, so games are GPU limited.
If you test any of those games in 1024x768, you will see that the difference is quite "significant" (10-20fps or more), it's just meaningless because nobody cares if something runs 160 or 185fps. Perhaps future games (AMD gaming evolved titles, XBOX720 and PS4 ports will run much better on i7s, at least that's my guess.

 

[nickbaldwin86]

Good review... little dated and I am sure more would like more data too

SLi/crossfire/High res/120hz see a bigger need for OC in games.... but if you are just gaming at 1080p why even waste you money on either of these CPUs, just to say you have it? you could do with a low end non-K model IVY Bridge and get the same gaming results.

 

[MaKCuMyC]

What's wrong with WinRAR?
3770K without oc in my system shows around 8 MiB/s with 4.20 version.

 

[Nordic]

Good review... little dated and I am sure more would like more data too

SLi/crossfire/High res/120hz see a bigger need for OC in games.... but if you are just gaming at 1080p why even waste you money on either of these CPUs, just to say you have it? you could do with a low end non-K model IVY Bridge and get the same gaming results.

In my case, natural selection 2 requires a lot of single threaded cpu grunt. Most games don't though.

 

[Sasqui]

Even for what those extra threads give you, an i7 just does not seem worth it.

I'm sitting happy with a 3750k, knowing it wasn't as good, but at $169 the price/perforamance was better and the benches confirm that.

Speaking of which, too bad no price performance charts like with the GPUs. Though it would be hard to compile the different measures to put something meaningful together.

 

[FireKillerGR]

What's wrong with WinRAR?
3770K in my system shows around 8 MiB/s with 4.20 version.

Seemed to be an OS bug, cause it limits the load of winrar on cpu :/

 

[Mals]

Good review... little dated and I am sure more would like more data too

SLi/crossfire/High res/120hz see a bigger need for OC in games.... but if you are just gaming at 1080p why even waste you money on either of these CPUs, just to say you have it? you could do with a low end non-K model IVY Bridge and get the same gaming results.

I hate to say it but I completely disagree. Framerates, especially minimum, are heavily dependent on core clock. Clocking from stock 3.5 to the 4.6 that my 3770k is at increased my minimum FPS by over 20 in BF3. There are SIGNIFICANT improvements to be had.

There are tons of people testing this information and lots of sources to back it up.

Having an unlocked CPU makes for big gains in the gaming world of framerates.

However, if you just play games like say.. league of legends on a 1080p 60hz monitor, then there is no need.

These processors are meant for playing BF3, Farcry, Crysis, etc, at their best framerates and to be paired with high end video cards.

 

[BiggieShady]

any other games that you can recommend for cpu testing?

Grand Theft Auto 4 (more if vehicle count and object view distance is maxed).
Microsoft Flight Simulator
Battlefield 3 MP 64 players
Total war series, Shogun 2, Supreme commander (huge battles)

... and Skyrim for CPU bottlenecked fun with cheese http://www.youtube.com/watch?v=fCrhDamN82k ... I suggest this test to be called "a million cheese wedges test"

 

[Enmitynz]

Good review

I do want to point out tho that I have used both of these processors and the in-game benchmarks make a very misguided opinion of their gaming performances.

I switched from the i5 to the i7 after some extensive in-game testing of games like BF3 on multiplayer maps.

In standard benchmarks, they see no difference, but games like BF3 utilize those hyper-threaded cores in a huge way.

My minimum FPS over a period of time in BF3 rose by over 25 Frames.

I game at 120FPS on a 120hz monitor. With the i5-3570k and GTX 670's in SLI I saw significant, frequent dips to below 80 frames.

Now, with the exact same setup and only changing to an i7-3770k I NEVER see my FPS go below 100. Lowest I've seen is 102 or so.

Just an FYI, it really makes a huge difference in online, multiplayer games. MMO's, FPS's, etc.

This is a very good point, a lot of reviewers seem to only record the average frame rates, or the maximum achieved framerates when it comes to cpu reviewing, testing games.

I think when it comes to gaming, it would be very beneficial to us all to see the minimum, average and the maximum frames achieved. I can understand that testing bf3 in multiplayer is pretty much impossible as you cannot reproduce the same bench...but min/max/avg would be great

Really nice review man! keep up the great work and i'm looking forward to seeing your next review to see how you've improved

 

[Mals]

This is a very good point, a lot of reviewers seem to only record the average frame rates, or the maximum achieved framerates when it comes to cpu reviewing, testing games.

I think when it comes to gaming, it would be very beneficial to us all to see the minimum, average and the maximum frames achieved. I can understand that testing bf3 in multiplayer is pretty much impossible as you cannot reproduce the same bench...but min/max/avg would be great

Really nice review man! keep up the great work and i'm looking forward to seeing your next review to see how you've improved

You can test BF3 multiplayer, but it takes a lot of time on the same maps to really get a better idea.

You play the same map for a few hours and then test A/B you will receive a decent idea of the frames. As long as playercount is the same, it evens out over time. It's not perfect, but for an A/B test it makes a pretty good assessment.

 

[deadmansclick]

any other games that you can recommend for cpu testing?

X3 Terran Conflict rolling demo http://www.egosoft.com/download/x3tc/demos_en.php?list=92

This is really all about the cpu. o/cing a cpu will yield a massive fps bump.

It might be a few years old now but still looks stunning. Also X Rebirth is due out this year & from what I've read will be much of the same with its newer engine in regards to the cpu, but utilise multicores better

 

[Melvis]

Thanks for the review.

(for example Metro2033 is flat out entirely GPU-bond).

That is totally incorrect!! Ive benchmarked this game using two different CPU's using the same everything else, and the frame rate near doubled from a Phenom 965 to a 8350 at stock settings.

Metro 2033 ill tell you this right now loves moar cores period!!

 

[nickbaldwin86]

This is a very good point, a lot of reviewers seem to only record the average frame rates, or the maximum achieved framerates when it comes to cpu reviewing, testing games.

I think when it comes to gaming, it would be very beneficial to us all to see the minimum, average and the maximum frames achieved. I can understand that testing bf3 in multiplayer is pretty much impossible as you cannot reproduce the same bench...but min/max/avg would be great

Really nice review man! keep up the great work and i'm looking forward to seeing your next review to see how you've improved

What? did you see the gaming benchmarks??? they are 1-2FPS different.... no one cares what you are getting on your system, we aren't talking about your system, we are talking about the benchmarks in this review on a controlled environment system just like every other review on the internet.

Sorry but as far as gaming at 1080p is concerned the cores/threads/clock speed just doesn't matter, this review shows that:
http://www.techpowerup.com/reviews/Intel/Core_i5_3570K_and_i7_3770K_Comparison/8.html

maybe you skipped that page? or didn't care to read the graphs? sorry but 78-80FPS isn't noticeable or worth $100+++ and plus $100 for a cooler

 

[Ikaruga]

That is totally incorrect!! Ive benchmarked this game using two different CPU's using the same everything else, and the frame rate near doubled from a Phenom 965 to a 8350 at stock settings.

Metro 2033 ill tell you this right now loves moar cores period!!

I also did my tests ofc, but why not compare the 8350 to a 486DX and see how much better result you will get?

The thread is about i5 vs i7, I'm trying to maintain context with the subject:
i7-980XE vs X4965
average fps with 4 core 81.34 vs 81.81 with 6 cores:

etc.... As soon as you reach 4 cores (in Metro2033), only the minimum fps is going to be affected a little bit, but the overall gaming experience is about the speed of the GPU.

 

[Melvis]

I also did my tests ofc, but why not compare the 8350 to a 486DX and see how much better result you will get?

The thread is about i5 vs i7, I'm trying to maintain context with the subject:
http://www.legionhardware.com/articles_pages/metro_2033_performance_guide,9.html
average fps with 4 core 81.34 vs 81.81 with 6 cores: http://forums.anandtech.com/showpost.php?p=31972973&postcount=6

etc....

Now that's just plain stupid.

Im just telling you that your statement is wrong, nothing more.

No your not your saying that Metro 2033 is all about GPU and it isnt, hell even this video backs up my claims.

AMD FX 8350 vs Intel 3570K vs 3770K vs 3820 - Gami...

Just make sure in future you look into these things before saying stuff like that ok

 

[Ikaruga]

Now that's just plain stupid.

Im just telling you that your statement is wrong, nothing more.

No your not your saying that Metro 2033 is all about GPU and it isnt, hell even this video backs up my claims.

AMD FX 8350 vs Intel 3570K vs 3770K vs 3820 - Gami...

Just make sure in future you look into these things before saying stuff like that ok

Oh, I already commented on that video ones, so I just gonna link that post here.
Please don't get me wrong, I'm not doing this to disrespect you, or to offend you (definitely not how you just tried to offend me btw), but because that video doesn't really worth another unique post from me tbh.

http://www.techpowerup.com/forums/showpost.php?p=2836720&postcount=92

 

[Melvis]

Oh, I already commented on that video ones, so I just gonna link that post here.
Please don't get me wrong, I'm not doing this to disrespect you, or to offend you (definitely not how you just tried to offend me btw), but because that video doesn't really worth another unique post from me tbh.

http://www.techpowerup.com/forums/showpost.php?p=2836720&postcount=92

Well i guess you haven't done personal testing like i have then? I have screen shots to prove those results you posted are wrong, what can i say?

Im not trying to offend you either im just saying that its a false statement from personal (not from some website) that the game does indeed like more cores (for me anyway) i got told the same thing by many that 2033 was all gpu bound till i tested it personally and found that to be false.

Just for you ill make a thread to show you the results i got with my own system and you can see for your self, what else can i say or do?