3570k or 2600k

[J_tyo]

Can't decide between the two of them but I am planning on overclocking. Current mobo, P8Z68-v/gen 3

 

[popatim]

what will you be doing? anything that will use the HT cores that the i5 doesnt have?

 

[redeemer]

Sandy Bridge>Ivy Bridge

 

[J_E_D_70]

Reason u cant go with 3770k on a z77?

Edit: what is your goal with the overclock?

 

[neodymium]

If you plan to use programs that can only use a single thread, go with the i5. Other than that, the i5 uses less power. However, the i7 allows for a higher stable maximum overclock, but the Ivy bridge (i5) can get more processes done per cycle.

 

[J_tyo]

I plan on gaming and recording at the same time

 

[imomun]

2600k for you

 

[goonbar79]

Ahh.. Sandy Bridge vs Ivy Bridge.. This has been questioned a zillion times in the past 6~9 months..

My take on 2600K vs. 3570K.

At stock speed of 3.4ghz:
1. 3570K is about 6% faster using 4 threads or less, and uses about 20~30 watt less.
2. 2600K is about 28% faster using all 8 threads.

At reasonable OC'd speed (for example, a good value is 2600K @ 4.4ghz, 3570K @4.2ghz)
1. both are equal using 4 threads or less, but 3570K uses 20~30 watt less.
2. 2600K is about 33% faster using all 8 threads.

Price Difference (i.e. newegg.com): 2600K= $320, 3570K= $230

My opinion: Value wise, 3570K is better for most tasks including gaming, due to current pricing. If you multitask and transcode video like crazy, 2600K is better, but FX-8xxx could be an option for under $200. 2600K is better if you can find it at a discounted price; at regular price 3770K is better value. I bought 2600K for $200 + $40 discount on mobo at microcenter, and because of that it was much better value than 3570K or 3770K at the time.

 

[chesteracorgi]

Sandy Bridge>Ivy Bridge

Nonsense a IB clocked at 4.0 GHz does the same work as a SB clocked at 4.5 GHz. So OC your SB ro 5.175 GHz and I'll run my IB at 4.5 GHz and we end in a tie.

 

[deiidarax]

save money with FX 4100, the best

 

[goonbar79]

Nonsense a IB clocked at 4.0 GHz does the same work as a SB clocked at 4.5 GHz. So OC your SB ro 5.175 GHz and I'll run my IB at 4.5 GHz and we end in a tie.

IB is faster than SB, but not by THAT much.. IB at 4.0ghz is about the same as SB at 4.25ghz. Pretty much all tech sites have tested the performance difference, and that is the consensus.

Also, at higher clocks IB heats up much quicker, thus OC'd at higher clock is a bit more difficult. With similar cooling, my 2600K OC'd to 4.4 ghz, and my friend's 3570K OC'd to 4.2ghz, both without voltage changes, with similar hardware (not identical, but similar). At those clocks, the difference is the wattage used, not performance.

 

[chesteracorgi]

IB is faster than SB, but not by THAT much.. IB at 4.0ghz is about the same as SB at 4.25ghz. Pretty much all tech sites have tested the performance difference, and that is the consensus.

Also, at higher clocks IB heats up much quicker, thus OC'd at higher clock is a bit more difficult. With similar cooling, my 2600K OC'd to 4.4 ghz, and my friend's 3570K OC'd to 4.2ghz, both without voltage changes, with similar hardware (not identical, but similar). At those clocks, the difference is the wattage used, not performance.

Excuse me, but the reports of performance agree that there is a 15% differential in performance. The IB is 15% per cycle more efficient that the SB so a IB clockes at 4.0 GHz = SB at 4.0 * 1.15 GHz (4.6 GHz). Unless you have different performance measures that is the difference. [And if you mistakenly multiply 4.6 GHz by 0.85 the IB has an even greater margin of a comparable IB clocked at 3.91 GHJz].

BTW Hertz measures cycles per second. So if I am 15% more productive than Fred every cycle it means that for every 1 task he does I do 1.5. And your friend's 4.2 GHz IB chip is equal to your SB at 4.83 GHz.

 

[J_E_D_70]

Dogpile! The difference is 5% to 15% depending on the app and how it's programmed. So... you really won't be able to tell unless you run otherwise identical systems with stock clocks and compare.

 

[J_tyo]

So it seems like you guys are leaning towards the i5. but what are the big differences between the two?

 

[chesteracorgi]

And goonbar, I agree on the poor design of the IB heat dissapation compared to the SB. They should have used solder rather than paste. But numbers are numbers. And while the SB can be OCd much mor stably than the IB, the IB still is more productive per cycle.

 

[chesteracorgi]

A true comparison would be between the 2500K and a 3570K or a 2600K and a 3770K

 

[goonbar79]

Excuse me, but the reports of performance agree that there is a 15% differential in performance. The IB is 15% per cycle more efficient that the SB so a IB clockes at 4.0 GHz = SB at 4.0 * 1.15 GHz (4.6 GHz). Unless you have different performance measures that is the difference. [And if you mistakenly multiply 4.6 GHz by 0.85 the IB has an even greater margin of a comparable IB clocked at 3.91 GHJz].

BTW Hertz measures cycles per second. So if I am 15% more productive than Fred every cycle it means that for every 1 task he does I do 1.5. And your friend's 4.2 GHz IB chip is equal to your SB at 4.83 GHz.

Hi, chesteracorgi. Please do your research a bit more before making a bold comment that IB is 15% faster than SB. The only time any single microarchitecture (this includes AMD as well) change had more than 15% average performance improvement clock for clock is Nehalem from Penryn, when clock for clock, the increase was about 15%~25% (EDITED: Netburst to Core was over 100%, so lets not discuss that..). All other times, the improvement was about 5%~10%. For IB, the increase was about 6~7% clock for clock. Since 2600K is clocked at 3.4ghz, 3770K should be directly compared to 2700K to see the difference accurately.

http://www.tomshardware.com/reviews/ivy-bridge-benchmark-core-i7-3770k,3181-24.html
http://www.anandtech.com/show/5771/the-intel-ivy-bridge-core-i7-3770k-review/6
http://www.bit-tech.net/hardware/2012/04/23/intel-core-i7-3770k-review/1
http://www.tomshardware.com/reviews/processor-architecture-benchmark,2974-15.html

I have researched about this from over 10~15 sites worldwide, as well as conducting tests personally, and the results were solid. If you think my comment is inaccurate, please provide evidence from legitimate sites or professional reviews that shows some sort of explanation of your logic.

 

[J_tyo]

or should I wait for the new processors coming in June, I'm in no rush

 

[goonbar79]

And goonbar, I agree on the poor design of the IB heat dissapation compared to the SB. They should have used solder rather than paste. But numbers are numbers. And while the SB can be OCd much mor stably than the IB, the IB still is more productive per cycle.

Yes, you are right that IB is faster than SB clock for clock, but it is well known that IB is a bit more difficult to OC because of its high temp at higher clocks (usually about 4.2ghz and higher). IB WILL result in higher core temps than SB at high clocks. So the question is how experienced are you with OC'ing, and what cooler will you use? The answer is up to the PC user.

 

[goonbar79]

or should I wait for the new processors coming in June, I'm in no rush

I recommend that you wait for Haswell. It will give you many benefits, such as:

1. upgrade path with new socket 1150. Intel provided upgrade path to newer generation cpu (once) during the last several generations, so I don't think it will be any different for Haswell.
2. Likely cpu performance increase of 5~10%,
3. Likely igpu performance increase of near GTX 650M level. http://www.anandtech.com/show/6600/intel-haswell-gt3e-gpu-performance-compared-to-nvidias-geforce-gt-650m
4. more features... http://en.wikipedia.org/wiki/Haswell_(microarchitecture)
5. Most likely fixed IB OC heat issue, so higher OC clock expected (I read it somewhere.. can't remember where).

If you can wait, and can afford it, get haswell's new cpu replacing 3570K or 3770K. May be it will be called 4580K, 4780K? Only Time will tell.

 

[IntelEnthusiast]

The real issue is if you are gaming go with the Intel® Core™ i5-3570K because it give you about the same level of performance when gaming but it will also allow you to use the PCI-E 3.0 where the Intel Core i7-2600K will be stuck at PCI-E 2.0.

 

[chesteracorgi]

Hi, chesteracorgi. Please do your research a bit more before making a bold comment that IB is 15% faster than SB. The only time any single microarchitecture (this includes AMD as well) change had more than 15% average performance improvement clock for clock is Nehalem from Penryn, when clock for clock, the increase was about 15%~25% (EDITED: Netburst to Core was over 100%, so lets not discuss that..). All other times, the improvement was about 5%~10%. For IB, the increase was about 6~7% clock for clock. Since 2600K is clocked at 3.4ghz, 3770K should be directly compared to 2700K to see the difference accurately.

http://www.tomshardware.com/reviews/ivy-bridge-benchmark-core-i7-3770k,3181-24.html
http://www.anandtech.com/show/5771/the-intel-ivy-bridge-core-i7-3770k-review/6
http://www.bit-tech.net/hardware/2012/04/23/intel-core-i7-3770k-review/1
http://www.tomshardware.com/reviews/processor-architecture-benchmark,2974-15.html

I have researched about this from over 10~15 sites worldwide, as well as conducting tests personally, and the results were solid. If you think my comment is inaccurate, please provide evidence from legitimate sites or professional reviews that shows some sort of explanation of your logic.

No, you looked at exactly the same evidence that I did. From AnandTech " The 3770K however manages to slot in above all of the other Sandy Bridge parts at between 5—20% faster than the 2600K. The biggest advantages show up in either the lightly threaded tests or in the FP heavy benchmarks." Most applications, and particularly games, are single threaded. It is only in heavy multiple threaded apps that the IB only beats the SB by 6%. If you weight them properly they average out to 15%. However, even taking a 10% advantage (splitting the difference with you) a IB clocked at 4.0 GHz is equivalent to the SB clocked at 4.4 GHz.

I will readily admit that the IB is a poor overclocker compared to the SB, however, I still maintain that you can get a lot more bang for the buck with the IB. Your comment is not inaccurate, but you don't weigh the evidence properly. My logic is simple: I believe that the proper weighting of the IB compared to the SB is 15%, and that, I argue, is much closer to the true advantage of the IB over SB than your analysis. You may differ, but I do accept that for most applications the IB shows a greater gain for productivity that the increased clock speed of the SB. What we need is a show down comparing two comparable chips the 2500K and 3570K or 2700K and 3770K in the same apps (Oh I forgot that Anand and Tom's have done that), and end with the caveat that "your results may vary from ours."

Anyway goonbar79, I have enjoyed our exchange and hope that you understand that I appreciate your opinion. But in the end, I beg to differ. In any case. I will defend to the death your right to be wrong, and hope that your would pay me the same courtesy.

 

[chesteracorgi]

Hi, chesteracorgi. Please do your research a bit more before making a bold comment that IB is 15% faster than SB. The only time any single microarchitecture (this includes AMD as well) change had more than 15% average performance improvement clock for clock is Nehalem from Penryn, when clock for clock, the increase was about 15%~25% (EDITED: Netburst to Core was over 100%, so lets not discuss that..). All other times, the improvement was about 5%~10%. For IB, the increase was about 6~7% clock for clock. Since 2600K is clocked at 3.4ghz, 3770K should be directly compared to 2700K to see the difference accurately.

http://www.tomshardware.com/reviews/ivy-bridge-benchmark-core-i7-3770k,3181-24.html
http://www.anandtech.com/show/5771/the-intel-ivy-bridge-core-i7-3770k-review/6
http://www.bit-tech.net/hardware/2012/04/23/intel-core-i7-3770k-review/1
http://www.tomshardware.com/reviews/processor-architecture-benchmark,2974-15.html

I have researched about this from over 10~15 sites worldwide, as well as conducting tests personally, and the results were solid. If you think my comment is inaccurate, please provide evidence from legitimate sites or professional reviews that shows some sort of explanation of your logic.

No, you looked at exactly the same evidence that I did. From AnandTech " The 3770K however manages to slot in above all of the other Sandy Bridge parts at between 5—20% faster than the 2600K. The biggest advantages show up in either the lightly threaded tests or in the FP heavy benchmarks." Most applications, and particularly games, are single threaded. It is only in heavy multiple threaded apps that the IB only beats the SB by 6%. If you weight them properly they average out to 15%. However, even taking a 10% advantage (splitting the difference with you) a IB clocked at 4.0 GHz is equivalent to the SB clocked at 4.4 GHz.

I will readily admit that the IB is a poor overclocker compared to the SB, however, I still maintain that you can get a lot more bang for the buck with the IB. Your comment is not inaccurate, but you don't weigh the evidence properly. My logic is simple: I believe that the proper weighting of the IB compared to the SB is 15%, and that, I argue, is much closer to the true advantage of the IB over SB than your analysis. You may differ, but I do accept that for most applications the IB shows a greater gain for productivity that the increased clock speed of the SB. What we need is a show down comparing two comparable chips the 2500K and 3570K or 2700K and 3770K in the same apps (Oh I forgot that Anand and Tom's have done that), and end with the caveat that "your results may vary from ours."

Anyway goonbar79, I have enjoyed our exchange and hope that you understand that I appreciate your opinion. But in the end, I beg to differ. In any case. I will defend to the death your right to be wrong, and hope that you would pay me the same courtesy.

 

[goonbar79]

No, you looked at exactly the same evidence that I did. From AnandTech " The 3770K however manages to slot in above all of the other Sandy Bridge parts at between 5—20% faster than the 2600K. The biggest advantages show up in either the lightly threaded tests or in the FP heavy benchmarks." Most applications, and particularly games, are single threaded. It is only in heavy multiple threaded apps that the IB only beats the SB by 6%. If you weight them properly they average out to 15%. However, even taking a 10% advantage (splitting the difference with you) a IB clocked at 4.0 GHz is equivalent to the SB clocked at 4.4 GHz.

I will readily admit that the IB is a poor overclocker compared to the SB, however, I still maintain that you can get a lot more bang for the buck with the IB. Your comment is not inaccurate, but you don't weigh the evidence properly. My logic is simple: I believe that the proper weighting of the IB compared to the SB is 15%, and that, I argue, is much closer to the true advantage of the IB over SB than your analysis. You may differ, but I do accept that for most applications the IB shows a greater gain for productivity that the increased clock speed of the SB. What we need is a show down comparing two comparable chips the 2500K and 3570K or 2700K and 3770K in the same apps (Oh I forgot that Anand and Tom's have done that), and end with the caveat that "your results may vary from ours."

Anyway goonbar79, I have enjoyed our exchange and hope that you understand that I appreciate your opinion. But in the end, I beg to differ. In any case. I will defend to the death your right to be wrong, and hope that you would pay me the same courtesy.

Hi, chesteracorgi. I am sorry to burst your bubble, but here are a few things you are not considering:

1. Your quote from Anandtech's article page 6 is only for SYSMark 2007 & 2012, which he adds "Although not the best indication of overall system performance, the SYSMark suites do give us a good idea of lighter workloads than we're used to testing". http://www.anandtech.com/show/5771/the-intel-ivy-bridge-core-i7-3770k-review/6

2. Anand means different parts of Sysmark tests are showing results between 5~20%, NOT that you want to average of 5 ~ 20% (which still doesn't get to 15%). If you look at the data closer, average for all of these are below 10%. Go ahead. Try it and see for your self.

3. This article compares 3770K vs slightly lower clocked 2600K. Thats a 3% difference already, and I kept saying "clock for clock". Take a look at Toms article again, and tell me if this is inaccurate. Industry wide used benchmark applications, and every one of these have difference of below 8%. http://www.tomshardware.com/reviews/ivy-bridge-benchmark-core-i7-3770k,3181-24.html

Look, I respect that you have a strong opinion about this, and that is fine. I have no problem that you have a DIFFERENT OPINION. But the CONSENSUS from tech professionals all over the world (I read Asian sites as well) has confirmed the "overall" performance difference, and you sir are alone with your theory of 15%+. If you say that for YOUR use, you get 15%+ performance, I will agree. But if you say that average of 15%+ is the norm, I still think you need solid evidence to back up that statement, and you still haven't shown ANY evidence.

 

[maza90210]

or should I wait for the new processors coming in June, I'm in no rush

I recommend that you wait for Haswell. It will give you many benefits, such as:

1. upgrade path with new socket 1150. Intel provided upgrade path to newer generation cpu (once) during the last several generations, so I don't think it will be any different for Haswell.
2. Likely cpu performance increase of 5~10%,
3. Likely igpu performance increase of near GTX 650M level. http://www.anandtech.com/show/6600/intel-haswell-gt3e-gpu-performance-compared-to-nvidias-geforce-gt-650m
4. more features... http://en.wikipedia.org/wiki/Haswell_(microarchitecture)
5. Most likely fixed IB OC heat issue, so higher OC clock expected (I read it somewhere.. can't remember where).

If you can wait, and can afford it, get haswell's new cpu replacing 3570K or 3770K. May be it will be called 4580K, 4780K? Only Time will tell.

Well this is awkward...