13" 2.66 Core 2 Duo Vs 15" 2.4 i5

[Ace134blue]

Benchmarks of the i-series have already proven your speed increase wrong, so I'm not going to argue that point any more.

Your right about turbo boost, but the i-series processors get very hot, very fast, so the turbo boost isn't very aggressive, it will only kick in when absolutely necessary.

And as for hyper-threading, I'm not saying the speed is decreased when there are multiple threads, I'm saying you only have so many execution cores that can be used. If one thread is using 80% of the ECs, the "hyper-thread" is only able to use a maximum of 20%, and only if its instructions can be processed by those specific cores. Like I said, its just a method of using up wasted cpu cycles.

Since there is no app to test just the cpu in osx, no one can show *real* benchmarks. Geekbench isnt a good app to test out the cpu, it adds all the points for the entire system together.
Best bet is to go into windows and test out 3dmark vantage 06 or similar.

This " If one thread is using 80% of the ECs, the "hyper-thread" is only able to use a maximum of 20%, and only if its instructions can be processed by those specific cores." is wrong. If the app is indeed multi thread aware, than no thats wrong.. period. If its multithread aware than it splits the processing into 4 threads. So lets say you got a cpu app thats only single-threaded away. Your 1st thread will be at 100%, now say it just got updated and is now 4-threaded aware , it than splits it up by 4. So each thread will run at 25%. Now this isnt always the case, not all apps can use 4 threads to its advantage and sometimes the most those apps will use is 2 threads.

The turbo boost kicks in when the cpu is being used, it doesnt matter how many threads are being used or how much the cpu is being stressed.
Now the cpu will throttle if it gets way too hot, but than that would be apples own fault.

Both core 2 dous and the i7s have the same tdp of 35 watts. The i7 is not gonna be much hotter, if at all. Hyperthreading might make it slightly hotter tho.

 

[alust2013]

Both core 2 dous and the i7s have the same tdp of 35 watts. The i7 is not gonna be much hotter, if at all. Hyperthreading might make it slightly hotter tho.

Actually, the C2Ds in the 13" are the P8xxx, with a TDP of 25 watts. The previous gen C2D 15 and 17s were the 35 watt TDP though.

 

[Ace134blue]

Actually, the C2Ds in the 13" are the P8xxx, with a TDP of 25 watts. The previous gen C2D 15 and 17s were the 35 watt TDP though.

No, 13in has nothing to do with any of this since it does NOT have an i7/i5. This is all about the 15in mbp which have the T9xxx

 

[kny3twalker]

No, 13in has nothing to do with any of this since it does NOT have an i7/i5. This is all about the 15in mbp which have the T9xxx

Are you sure both the 2.53 and 2.66 were 35w? I could believe the 2.66 was, but I'm pretty sure the 2.53 was the p8700. The p8800 I believe was the 2.66, and there are the p9xxx series which have double the cache and are still 25w and 28w. There was a 2.66 and 2.53 p9xxx CPUs as well.

 

[Ace134blue]

Are you sure both the 2.53 and 2.66 were 35w? I could believe the 2.66 was, but I'm pretty sure the 2.53 was the p8700. The p8800 I believe was the 2.66, and there are the p9xxx series which have double the cache and are still 25w and 28w. There was a 2.66 and 2.53 p9xxx CPUs as well.

Yes, the 15in/17in macbook pros, last gen, had the T9xxx series. My old 2.8ghz has a T9600, the upgraded cpu 3.06 has a T9900. The Txxx series have a 35watt TDP. If you dont believe me look it up.
Also, the 2.53 is a T9400 and the 2.66 is a T9550, which all have a tdp of 35watts. They did NOT have the P series

 

[mikeo007]

Since there is no app to test just the cpu in osx, no one can show *real* benchmarks. Geekbench isnt a good app to test out the cpu, it adds all the points for the entire system together.
Best bet is to go into windows and test out 3dmark vantage 06 or similar.

This " If one thread is using 80% of the ECs, the "hyper-thread" is only able to use a maximum of 20%, and only if its instructions can be processed by those specific cores." is wrong. If the app is indeed multi thread aware, than no thats wrong.. period. If its multithread aware than it splits the processing into 4 threads. So lets say you got a cpu app thats only single-threaded away. Your 1st thread will be at 100%, now say it just got updated and is now 4-threaded aware , it than splits it up by 4. So each thread will run at 25%. Now this isnt always the case, not all apps can use 4 threads to its advantage and sometimes the most those apps will use is 2 threads.

The turbo boost kicks in when the cpu is being used, it doesnt matter how many threads are being used or how much the cpu is being stressed.
Now the cpu will throttle if it gets way too hot, but than that would be apples own fault.

Both core 2 dous and the i7s have the same tdp of 35 watts. The i7 is not gonna be much hotter, if at all. Hyperthreading might make it slightly hotter tho.

I understand what you're trying to say, but by your explanation, hyper threading would offer zero benefit (4 @ 25 % = 1 @ 100%). In reality, hyper threading offers a 5-30% benefit under specifit circumstances. Remember, HTT is basically just a 2nd set of registers, no other part of the core is duplicated.

My explanation of 80% / 20% may not have made sense when you think about it in terms of ECs, so picture it as clock cycles instead. One processor still only has so many clock cycles. If there is a cache miss, or something else causes a thread to stall, that's when hyper threading kicks in. I'm sure you already know how HTT works, but your performance expectations are pretty exaggerated.

In terms of heat, the i7 MBP has been pretty highly criticized for the amount of heat it produces. Intels HTT has also been criticized for wasting a lot of power and thus causing even more heat.

 

[Ace134blue]

I understand what you're trying to say, but by your explanation, hyper threading would offer zero benefit (4 @ 25 % = 1 @ 100%). In reality, hyper threading offers a 5-30% benefit under specifit circumstances. Remember, HTT is basically just a 2nd set of registers, no other part of the core is duplicated.

My explanation of 80% / 20% may not have made sense when you think about it in terms of ECs, so picture it as clock cycles instead. One processor still only has so many clock cycles. If there is a cache miss, or something else causes a thread to stall, that's when hyper threading kicks in. I'm sure you already know how HTT works, but your performance expectations are pretty exaggerated.

In terms of heat, the i7 MBP has been pretty highly criticized for the amount of heat it produces. Intels HTT has also been criticized for wasting a lot of power and thus causing even more heat.

Ok i see what your seeing, yeah your right there would be zero performance benefits with all threads at 25%, but lets say you got two threads at 100%. Sure if HT was enabled there wouldnt be any performance difference having 4 threads at 25%, BUT that just adds more peformance available to the app.
Say your handbraking a dvd to your phone, it takes 40 mins to handbrake, say avatar, with just two threads. The 2 threads will be maxed at 100% Ok so you now enable HT, and all 4 threads are maxed too but the time it takes is now 25 mins.

i7 macbook pro is noticably cooler than my old 2.8ghz mbp.
Increased heat is due to the crappy application of tim on the cpu/gpu.