I don't think 2.6Ghz CPU is worth 100$

[airlied]

I've got 2.2Ghz and 2.6Ghz at the same time. I ran geekbench multiple times but the score between them are so close. Check out the screenshot.

Of course 2.6Ghz is higher in general, The highest score I got so far (5371, 22461) is from 2.6Ghz, but the second highest (2011, 22311) is from 2.2Ghz.

I think you can't feel much differences in real world performance. Perhaps 2.6Ghz will save you few seconds at some 10+ minutes video exporting.

If you don't mind CTO and waiting, get the base model 15' then upgrade the graphic card to 560X, and 512GB SSD would be a solid choice (If you want 32GB ram you are gonna CTO anyway). The extra 100$ seems not too much, but you can get a 500GB external SSD drive for 100+$ these days.

BTW I didn't say 2.2Ghz is 'better'. Of cource 2.6Ghz is better. Im saying that if the performance boost 2.6 gives you overceeds 100$ you have to pay, of course go for it, it's totally up to you. But for most of us saving 100% then get a extra SSD or something might benifit you the most.

I hope this could help you making purchase decision.

 

[exogeneity]

Thanks for affirming my choice, I have a 2.2/16/560x/512 coming end of the week. I can spare a few minutes (not that I am rendering videos but maybe regression models and dataset), so I channeled the savings to my iPhone 2018 fund pot. Hahah!

 

[Howard2k]

I hope this could help you making purchase decision.

It would, if people read the forum before posting their 2.2 vs 2.6 or 13" vs 15" threads.

 

[airlied]

Thanks for affirming my choice, I have a 2.2/16/560x/512 coming end of the week. I can spare a few minutes (not that I am rendering videos but maybe regression models and dataset), so I channeled the savings to my iPhone 2018 fund pot. Hahah!

I think you won't regret your choice. And another bonus for 2.2Ghz is, it runs cooler than what 2.6Ghz does.
Of course when two cpus running at high frequency they both become pretty hot, but at general usage (like idle, browsing webs, 4k playback), 2.2Ghz is definitely cooler.

 

[exogeneity]

I think you won't regret your choice. And another bonus for 2.2Ghz is, it runs cooler than what 2.6Ghz does.
Of course when two cpus running at high frequency they both become pretty hot, but at general usage (like idle, browsing webs, 4k playback), 2.2Ghz is definitely cooler.

Cheers!

 

[DeanPSN]

B

I've got 2.2Ghz and 2.6Ghz at the same time. I ran geekbench multiple times but the score between them are so close. Check out the screenshot.
View attachment 778512

Of course 2.6Ghz is higher in general, The highest score I got so far (5371, 22461) is from 2.6Ghz, but the second highest (2011, 22311) is from 2.2Ghz.

I think you can't feel much differences in real world performance. Perhaps 2.6Ghz will save you few seconds at some 10+ minutes video exporting.

If you don't mind CTO and waiting, get the base model 15' then upgrade the graphic card to 560X, and 512GB SSD would be a solid choice (If you want 32GB ram you are gonna CTO anyway). The extra 100$ seems not too much, but you can get a 500GB external SSD drive for 100+$ these days.

I hope this could help you making purchase decision.

Honestly, benchmarks means nothing in real world performance and the i9 has its own benefits for those who need the extra horsepower. ;-)

To the one who mentioned the 2.2 runs cooler, i think there are plenty of videos on YouTube showing that they are same and in some cases the i9 will run cooler since it has higher clock speed and doesn't need to turbo as much as the 2.2 for example.
Check yourself the videos if you want, i only have the i9 so cannot test against another MBP.

 

[happyslayer]

This was my thought as well--CPUs seem to be really close in most benchmarks and such--and it's I why I went with the base model 15" instead of the next level 2.6GHz model. I wanted the coolest running, highest battery life I could get out of the larger screened Mac laptop. On the other hand, I was using a 12" MacBook before this so even the base 2.2GHz hexacore CPU is still a pretty dramatic increase in power, so I may not be the best judge.

Edit: Also, thanks for the table at the top of the post. The engineer in me loves all that info!

 

[leman]

Yeah, if you buy a computer primarily for running a limited-use synthetic benchmark, its probably not worth it. If you actually look at doing proper work, it might be worth it.

And another bonus for 2.2Ghz is, it runs cooler than what 2.6Ghz does.

Why would it? Most 2.6Ghz models are higher binned units that are marketed as running at higher clocks with identical energy consumption. Furthermore a faster machine will complete the work faster and thus spend less time in high-energy state. Of course, it depends on myriad small things...

but at general usage (like idle, browsing webs, 4k playback), 2.2Ghz is definitely cooler.

I think you might misunderstand what the base clock represents. It's not the clock at idle. It's the minimal guaranteed clock when the CPU is running a sustained intensive workload on all cores when consimung 45 watt of energy. All of these CPUs will automatically clock up and down depending on what you are doing. If the machine is idling, the CPU will be sent in a power-saving state and set the clock well below its base frequency.
[doublepost=1535557927][/doublepost]

I wanted the coolest running, highest battery life I could get out of the larger screened Mac laptop.

For maximising battery life, your best bet is going with the highest-binned CPU, as they tend to be most efficient (that would be the i9 model). That in theory, of course. Practical differences will probably be negligible.

 

[1050792]

In the end of the day people will try to justify their purchases for everything... the I7 2.6 is better than the 2.2 period. It doesn't mean the 2.2 is bad but there's a reason it's only available in the stock model.

 

[Hackintosh Sr.]

what’s the highest clock speeds people are actually seeing 2.2 vs 2.6 ? Not sure if I’ve missed that info.

 

[Adeel_]

I ordered a MBP 2.2/560x/16GB/1TB
Can't wait to have it !

 

[maflynn]

In the end of the day people will try to justify their purchases

Yup, people (including me ) will say the 2.2 is better for x, y and z, where as others will say the 2.6 is better for A, B, C reasons. Then of course is the i9 folks saying their machines are better because of 1, 2, 3 reasons.

At the end of the day, you get the configuration that best meets your needs. I like my configuration but that doesn't mean the 2.6 or i9 is bad

 

[AppleHaterLover]

This base clock stuff is a lie. I don't know how Intel is getting away with this, they really should call their processors something more illustrative.

Get Intel Power Gadget for say the 2.2 GHz processor. It will be at literally any clock BUT 2.2 GHz at any given time. If it's idling, it will be at 1.3 or thereabouts, and will be in Turbo Boost if performance is required.

The difference to the 2.6 GHz model (all else being equal) is that it will Turbo Boost higher (but all else being equal, it will throttle; just the initial boost will be higher and then it will throttle to the thermal limit, hence extremely similar results).

My 2017 nTB ("2.3GHz", Turbo Boost to "3.5 GHz") will in reality be under 1.5 GHz in idle, Turbo Boost to 3.5 briefly and if continuous performance is required (I used prime95 to simulate a persistent high load) will stay at 3.1 GHz literally all day long, probabaly thermal-limited (it will hold 90-95C). TDP will greatly exceed 15W as well. I'd expect that the better processor available for this machine (2.6 up to 4.0GHz) would do literally the exact same thing except initial Turbo Boost would be higher.

So in short, it's the same thing, unless there is something else being upgraded with the processor (e.g. specs other than clock or improved thermal performance)

 

[Howard2k]

This base clock stuff is a lie. I don't know how Intel is getting away with this, they really should call their processors something more illustrative.

Get Intel Power Gadget for say the 2.2 GHz processor. It will be at literally any clock BUT 2.2 GHz at any given time. If it's idling, it will be at 1.3 or thereabouts, and will be in Turbo Boost if performance is required.

The difference to the 2.6 GHz model (all else being equal) is that it will Turbo Boost higher (but all else being equal, it will throttle; just the initial boost will be higher and then it will throttle to the thermal limit, hence extremely similar results).

My 2017 nTB ("2.3GHz", Turbo Boost to "3.5 GHz") will in reality be under 1.5 GHz in idle, Turbo Boost to 3.5 briefly and if continuous performance is required (I used prime95 to simulate a persistent high load) will stay at 3.1 GHz literally all day long, probabaly thermal-limited (it will hold 90-95C). TDP will greatly exceed 15W as well. I'd expect that the better processor available for this machine (2.6 up to 4.0GHz) would do literally the exact same thing except initial Turbo Boost would be higher.

So in short, it's the same thing, unless there is something else being upgraded with the processor (e.g. specs other than clock or improved thermal performance)

Sorry man, but it sounds like a misunderstanding on your side.

 

[1050792]

Yup, people (including me ) will say the 2.2 is better for x, y and z, where as others will say the 2.6 is better for A, B, C reasons. Then of course is the i9 folks saying their machines are better because of 1, 2, 3 reasons.

At the end of the day, you get the configuration that best meets your needs. I like my configuration but that doesn't mean the 2.6 or i9 is bad

Exactly what I meant, unlike this thread which is making the 2.6 look bad in comparison to the 2.2.

 

[Thysanoptera]

I think you might misunderstand what the base clock represents. It's not the clock at idle. It's the minimal guaranteed clock when the CPU is running a sustained intensive workload on all cores when consimung 45 watt of energy.

If that was the case I don't think anybody would buy anything else but the i9. It used to be like that, you had a clock and corresponding TDP, but now, especially in mobile SKUs the base clock looks like a marketing label and nothing else.

For maximising battery life, your best bet is going with the highest-binned CPU, as they tend to be most efficient (that would be the i9 model). That in theory, of course. Practical differences will probably be negligible.

Package TDP vs clock in 2.2, in hot server room, I'll repeat when I get to the hotel. Primes in windows, power limit set using Intel XTU:

2.2 - 32W
2.6 - 38W
2.9 - 45W (what a surprise! - but I was already hitting thermal limit, at home I could run it at 3.2)

Now it would be really nice to gather a large number of such measurements, to see if Intel is really binning those SKU's. My bet is they don't do it, there is so much spare room on 2.2 and 2.6 that there is no need for it. The most logical and cost effective method would be to use the yields without cache errors as i9 and just label 2.2 and 2.6 based on demand.

 

[leman]

This base clock stuff is a lie. I don't know how Intel is getting away with this, they really should call their processors something more illustrative.

They do document what clocks mean, but its very convoluted and no user bothers to read or understand the technical aspects of it.

So in short, it's the same thing, unless there is something else being upgraded with the processor (e.g. specs other than clock or improved thermal performance)

What you are missing is that lower-tier CPUs are usually parts that have been deemed inferior (in some way) during testing. They are often not able to hit same high clocks as their more "premium" siblings, and also tend to have worse performance/watt figures (as in needing more power to maintain same performance). Its all hit and miss, since its about luck of the draw, but on average, the 2.2Ghz CPU will end up slower than the 2.6Ghz CPU exactly for that reason.
[doublepost=1535574734][/doublepost]

Now it would be really nice to gather a large number of such measurements, to see if Intel is really binning those SKU's. My bet is they don't do it, there is so much spare room on 2.2 and 2.6 that there is no need for it.

It is of course possible that you are correct. Intel's marketing is totally messed up recently anyway...

 

[AppleHaterLover]

Sorry man, but it sounds like a misunderstanding on your side.

Does it?

 

[ashcairo]

Interestingly Intel price those two chips the same, the i9 is the only one which costs more.

Regarding performance figures you have to take into account the GPU, as the 2.6 has the 560X which can negatively affect the performance for certain workloads.

 

[airlied]

B


Honestly, benchmarks means nothing in real world performance and the i9 has its own benefits for those who need the extra horsepower. ;-)

To the one who mentioned the 2.2 runs cooler, i think there are plenty of videos on YouTube showing that they are same and in some cases the i9 will run cooler since it has higher clock speed and doesn't need to turbo as much as the 2.2 for example.
Check yourself the videos if you want, i only have the i9 so cannot test against another MBP.

Yes 2.9 has its own value to some people, but I believe 2.2 it pretty enough for the most. And the performance differences among them is less than the cost to get them.

And about which one runs cooler, yes I have both 2.2 and 2.6 so as far as I can tell 2.2 runs cooler when idle (of cource both run equally hot when doing some crazy)

 

[maflynn]

And about which one runs cooler, yes I have both 2.2 and 2.6 so as far as I can tell 2.2 runs cooler

the simple reason as to why the 2.2 runs cooler is because its a slower cpu, i.e., .4 GHz slower and the 550x vs 560x. The latter is going to produce more heat.

 

[airlied]

the simple reason as to why the 2.2 runs cooler is because its a slower cpu, i.e., .4 GHz slower and the 550x vs 560x. The latter is going to produce more heat.

yup, so my point is 2.2Ghz is slower on paper bot not that huge in real world. And BTW my 2.2 is 560X also.

 

[maflynn]

yup, so my point is 2.2Ghz is slower on paper bot not that huge in real world. And BTW my 2.2 is 560X also.

The thing with synthetic benchmarks, is they give an indication of how it will perform, but the proof will ultimately be in real world usage. I'm of the opinion that 2.6 will be marginally quicker (marginally because of the small increase in clock speed) in CPU activities, and much quicker when the GPU gets involved. Just my unscientific opinion