Kaby Lake-Y m3 vs i5 vs i7

[C64]

From https://liliputing.com/2016/10/intel-kaby-lake-y-benchmarks-4-5-watt-core-i5-7y54.html

For the past few chip generations, Intel has branded its Y series processors as “Core M,” but with the introduction of 7th-gen “Kaby Lake” chips, the company took a different approach. There’s a 1 GHz Core m3 chip, but other Y series processors include models with Core i5 and Core i7 branding.

http://ark.intel.com/products/95449/Intel-Core-m3-7Y30-Processor-4M-Cache-2_60-GHz-
http://ark.intel.com/products/95452/Intel-Core-i5-7Y54-Processor-4M-Cache-up-to-3_20-GHz
http://ark.intel.com/products/95441/Intel-Core-i7-7Y75-Processor-4M-Cache-up-to-3_60-GHz

If you've been reading along on this forum for a while you know there are many topics about the 2015 and 2016 MacBook and which CPU to get. I haven't been able to find a definitive answer, but something close to it. Overall it seems like there's a noticeable jump from the m3 to the m5, but the difference between the m5 and m7 is not worth it. But then there are plenty of people who say to have no problems at all using the m3 for all sorts of tasks.

Looking at these specs, the m3 starts at a base frequency of 1.00 GHz, the m5 is 1.2 and the m7 is 1.3. That's about it? Doesn't look like the "m" vs. "i" branding is for anything except marketing purposes, and it's the same "jump" as with the m3/m5 2016 models. I'm wondering how much difference there is between the 2017 m3 and the 2016 i5, and whether or not in practice the 2017 m3 might be fine for even more people this time around.

I'm pulling the trigger on Monday no matter what. Same design so I'm trusting that the whole manufacturing process is still the same and there are no "newness" problems there. And jumping in right after the latest update feels good.

Even though I don't need that much power for most of my tasks, I'm still leaning towards the i5. Planning to use this MacBook for quite a few years, and I'm worried the m3 might become sluggish just a little too fast. And if 16GB is an option with the i5/i7, that'd be a no-brainer.

What are your plans?

 

[DrDan99]

I alway felt the “next to fastest” processor was the sweet spot and over the years usually where I buy. Probably the i5, but I have a profitable business that is buying it for me so Uncle Sammy is paying part.

I will see the price difference and packages offered. Totally agree, the 16bgb of RAM is a no brainier, if offered, and if only with the i7, I’m there.

 

[casperes1996]

From https://liliputing.com/2016/10/intel-kaby-lake-y-benchmarks-4-5-watt-core-i5-7y54.html


http://ark.intel.com/products/95449/Intel-Core-m3-7Y30-Processor-4M-Cache-2_60-GHz-
http://ark.intel.com/products/95452/Intel-Core-i5-7Y54-Processor-4M-Cache-up-to-3_20-GHz
http://ark.intel.com/products/95441/Intel-Core-i7-7Y75-Processor-4M-Cache-up-to-3_60-GHz

If you've been reading along on this forum for a while you know there are many topics about the 2015 and 2016 MacBook and which CPU to get. I haven't been able to find a definitive answer, but something close to it. Overall it seems like there's a noticeable jump from the m3 to the m5, but the difference between the m5 and m7 is not worth it. But then there are plenty of people who say to have no problems at all using the m3 for all sorts of tasks.

Looking at these specs, the m3 starts at a base frequency of 1.00 GHz, the m5 is 1.2 and the m7 is 1.3. That's about it? Doesn't look like the "m" vs. "i" branding is for anything except marketing purposes, and it's the same "jump" as with the m3/m5 2016 models. I'm wondering how much difference there is between the 2017 m3 and the 2016 i5, and whether or not in practice the 2017 m3 might be fine for even more people this time around.

I'm pulling the trigger on Monday no matter what. Same design so I'm trusting that the whole manufacturing process is still the same and there are no "newness" problems there. And jumping in right after the latest update feels good.

Even though I don't need that much power for most of my tasks, I'm still leaning towards the i5. Planning to use this MacBook for quite a few years, and I'm worried the m3 might become sluggish just a little too fast. And if 16GB is an option with the i5/i7, that'd be a no-brainer.

What are your plans?

Generally speaking, I wouldn't say the performance difference between any of them is enough to justify any investment larger than, shall we say $70. it won't really matter in terms of future proofing it, cause the performance difference from m3 to i5 is presumably less than the difference between similar chip from Kaby Lake to Coffee Lake. I'm assuming the i7 till perform similarly to the m3 of Coffee Lake, and the growth from then on is exponential, so when we go beyond coffee lake, future proofing starts seeming less and less economically good, and a better strategy would be spending enough money to get performance that works for you here and now, and put the "future proofing" budget in a savings account to be used on future hardware instead. That goes for CPU and GPU anyway. Not RAM and storage. Well, storage to some extend. But the priority is not getting less than you need, but also not way more I'd say.

 

[Qu1ckset]

I want kaby lake for the upgraded gpu not the slightly boosted clocks of the CPU

 

[jerwin]

Looking at these specs, the m3 starts at a base frequency of 1.00 GHz, the m5 is 1.2 and the m7 is 1.3. That's about it?

Look at the Turbo frequencies.
Also, the i7 and the i5 have very slightly higher boost frequencies for their GPUs. (Every little bit helps when all you have is a HD 615, maybe?)

http://ark.intel.com/compare/95449,95441,95452

(You can select "highlight differences")

The i7 also supports vPro, TSX-NI, Intel Stable Image Platform Program, Trusted Execution Technology, which may not be entirely relevant if you're running a mac-- or if you're not managing a corporate IT department.

 

[Michael Scrip]

Look at the Turbo frequencies.
Also, the i7 and the i5 have very slightly higher boost frequencies for their GPUs. (Every little bit helps when all you have is a HD 615, maybe?)

http://ark.intel.com/compare/95449,95441,95452

(You can select "highlight differences")

The i7 also supports vPro, TSX-NI, Intel Stable Image Platform Program, Trusted Execution Technology, which may not be entirely relevant if you're running a mac-- or if you're not managing a corporate IT department.

I just checked PassMark benchmarks. These chips aren't all that different:

3620 - m3-7Y30
3425 - i5-7Y54
3879 - i7-7Y75

And the i5 actually has the lowest score(?)

Of course this is only a benchmark. Real-world test might give different results.

But I thought it was curious.

 

[jerwin]

Of course this is only a benchmark. Real-world test might give different results.

There were 12 samples.

I wish I had taken a statistics class, so that I could describe how bad the data seems to me.
[doublepost=1496550599][/doublepost]I was going to suggest using geekbench, but then I saw this.

https://browser.primatelabs.com/v4/cpu/2926079

umm, yeah.

 

[Michael Scrip]

View attachment 702030

There were 12 samples.

I wish I had taken a statistics class, so that I could describe how bad the data seems to me.

Whoa... good catch.

The vast difference in benchmarks of the same chip must be due to different clockspeeds.

The lowest benchmark was at 1.81 GHz and the highest benchmark was at 3.12 GHz.

Armed with this new knowledge... I just checked the top CPU scores with all three CPUs fully turbo'd

3,871 - m3
4,425 - i5
4,294 - i7

Now the i7 dropped to 2nd place!

It's all over the place.

 

[Appleaker]

I have believed since August that Apple will drop the m3, and offer just i5 and i7 processors (https://forums.macrumors.com/thread...se-i5-and-i7-processor.1988987/#post-23262160) for the sake of uniformity, power, and avoiding confusion. That would give the base model a noticeable boost.

 

[C64]

I have believed since August that Apple will drop the m3, and offer just i5 and i7 processors (https://forums.macrumors.com/thread...se-i5-and-i7-processor.1988987/#post-23262160) for the sake of uniformity, power, and avoiding confusion. That would give the base model a noticeable boost.

The problem is that Apple cares very little about those things. The current m3 is $1299, the m5 is $1599 and the m3 to the m7 is $1549. So any upgrade adds at least $250 to the price. Also, the actual cost of the m3 vs. i5 for Apple is, I believe, either the same or there's not much difference at all. That's too good a deal to not use and make money on, because you can add a difference in the retail price.

If Apple drops the m3, they can do a few things:

1. Lower the price of the i5 to $1299 and lose that $250-300 upsell

Why would they do this if they have a perfectly fine m3 available to them to sell at that price point? Even if people care at all and are confused about this new naming scheme, it's still the very successful "good-better-best" cost structure, which is easy to understand and explain.

2. Keep the price the same at $1599 and make the MBP way more attractive.

I don't see them ever doing this. The base MacBook Pro starts at $1499. They can't get closer to this than the current difference, let alone go over it.

3. Up the starting price to something like $1399-1499 and cut their upsell with $200 or $100.

Same as option 2, very unlikely. Harder to justify buying the MacBook over the MacBook Pro, and they're simply losing money without any good reason.

If the Air isn't upgraded and they're ready to push those out (I think they'll remain on the site as-is to order for another year, just like the old MacBook Pro with Super Drive did for ages) in favor of retina screens all across the line, it's way more likely that they'll drop the price by $100, but then they really need that m3 as the base model.

 

[Esquire1]

Apple dropped the iPad price.

 

[Appleaker]

The problem is that Apple cares very little about those things. The current m3 is $1299, the m5 is $1599 and the m3 to the m7 is $1549. So any upgrade adds at least $250 to the price. Also, the actual cost of the m3 vs. i5 for Apple is, I believe, either the same or there's not much difference at all. That's too good a deal to not use and make money on, because you can add a difference in the retail price.

If Apple drops the m3, they can do a few things:

1. Lower the price of the i5 to $1299 and lose that $250-300 upsell

Why would they do this if they have a perfectly fine m3 available to them to sell at that price point? Even if people care at all and are confused about this new naming scheme, it's still the very successful "good-better-best" cost structure, which is easy to understand and explain.

2. Keep the price the same at $1599 and make the MBP way more attractive.

I don't see them ever doing this. The base MacBook Pro starts at $1499. They can't get closer to this than the current difference, let alone go over it.

3. Up the starting price to something like $1399-1499 and cut their upsell with $200 or $100.

Same as option 2, very unlikely. Harder to justify buying the MacBook over the MacBook Pro, and they're simply losing money without any good reason.

If the Air isn't upgraded and they're ready to push those out (I think they'll remain on the site as-is to order for another year, just like the old MacBook Pro with Super Drive did for ages) in favor of retina screens all across the line, it's way more likely that they'll drop the price by $100, but then they really need that m3 as the base model.

Yeah there are many reasons why they wouldn't do it, the most important (for Apple) being that they could lose money.

Although, as you said the price is roughly the same (based on the deal they have with Intel) and Apple can still charge the premium for a higher clocked i5. People were paying for the higher model when they were all 'Core M' with a 100MHz difference, so I think they would pay for a higher clocked i5. Not to mention there is also 512GB storage which is also in the price. The i7 would still be a BTO option, and it has to be because it's a lot more expensive. So it could be 1.2GHz i5, 1.3GHz i5, and 1.4GHz i7. Or they could increase the clock speed even more.

As you can tell from the poll on my thread only 25% of those people think it will happen. It could go either way but I am hoping for the i5 because I think it's logical and better for consumers. The price drop is basically guaranteed for WWDC and not a Fall event, but this could still be possible with the i5 as the base processor.

 

[C64]

A bit more on the 2016 m5 vs. 2017 m3:

http://www.cpu-monkey.com/en/compare_cpu-intel_core_m3_7y30-657-vs-intel_core_m5_6y54-564

http://cpu.userbenchmark.com/Compare/Intel-Core-m5-6Y54-vs-Intel-Core-m3-7Y30/m40602vsm196956

http://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+m3-7Y30+@+1.00GHz
http://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+m5-6Y54+@+1.10GHz

There aren't that many benchmarks yet, but overall it seems the m3 is right up there or has a slight advantage. I think it's safe to say that if you'd buy the m5 today for the 2016 model, you'll be just as well off with the 2017 m3 model. So if you don't need the extra storage, it might be worth saving that extra money.

(All assuming the m3 will be the base model, and the config options remain the same)

 

[C64]

Alight, I changed my vote. Unless there's a 16GB RAM option with only the i5 or i7, I'm buying the m3.

The differences are just too small, and I don't need the extra storage. If, for some bizar reason, it turns out to be too slow I can always exchange it, but with my usage I highly doubt it. Just 5.5 hours to go!

Update: ordered the i5 anyway, now that you don't necessarily need to upgrade the SSD as well