Intel's Upcoming Coffee Lake Processors Up to 30% Faster Than Kaby Lake Chips Coming to Mac Notebooks

[Ethosik]

i'm old

pro USED to mean "professional" level hardware that was intended for those users who relied on the performance of their hardware for work. People who, the faster the computer is, the faster they could complete their work.

this is in opposition to users who could do their work sufficiently on any computer that could run their software.

the Term pro USED to be used to refer to this distinction.

Now it's just a useless marketing that's being thrown around on any computer that's slightly upgraded from a lower version

IE: Surface Pro v Surface. Neither of these are truly power user performance level devices. despite the Pro badging.

So historically, the "PRO" devices weren't the ones where you compromised performance in return for "form". you complain about the battery life being too low if they put too much power in the evice, sure, but if the MacBook pro was the same thickness as the 2015 model, while offering the quad core performance, a class competing GPU, while having USB-c/thunderbolt and 10 hour battery life, I dont think anyone would complain

the problem with the 2016 MBp, is they sacrificed GPU and battery life for thinner. for those who fit into my description of what "pro" is, the new MacBook pro probably feels like a horizontal move at best from the 2015 model due to the decreased battery life and new thermal limits by smaller chassis

How does the Macbook Pro globally prevent any work done on it? Obviously you cannot work on 16K video or run 27 virtual machines from it. But even a "Professional" desktop has its limits too. Is the only product worthy of being classified as a professional system require 16 cores, 128GB of RAM, triple SLI, 100 TB of storage and more?

 

[LordVic]

How does the Macbook Pro globally prevent any work done on it? Obviously you cannot work on 16K video or run 27 virtual machines from it. But even a "Professional" desktop has its limits too. Is the only product worthy of being classified as a professional system require 16 cores, 128GB of RAM, triple SLI, 100 TB of storage and more?

there are always going to be limits on compute power dependant on the device design and chassis. and there is always going to be a required compromise.

the question becomes is where is the compromise too much towards form over function? (this won't necessarily be the same for everyone)

and your mistake is trying to assign a real objective measurement to what "pro" is. that's not at all what I've said.

"PRO" in my description meant tthat the more power you have available to you, the more work, or the faster your work gets completed. Therefore the ideal would always be "The fastest hardware possible"

A laptop, even a thicker bigger one, obviously is still going to have limits. Sure it can't fit the 16 core and 1080p within reasonable size. But a manufacturer should look at what it has to sacrifice to go smaller. And determine if that's acceptable for the target audience of the device.

So, 10 years ago, Apple's "Pro" meant what I mentioned. Sure, there were some sacrifices, but Apple didn't opt for a 20% slower part just to shave off 2mm. they realized there was a big distinction between their consumer lineup, who can get away with super light and less power because consumer use tends not to require maximization of performance.

But today, Apple's "pro" laptop very much trades maximization of performance for "form and design". When you look at the overall difference in chassis size and weight between the 2015 and 2016 models, for users who operate where "faster = more work", 2mm thickness and .3lbs, they have to ask themselves "is that worth a 20% slower GPU, and smaller battery".

For me, including the price increase, No, not at all. this device provides more consumer focused movement, at a price that isn't justifiable when compared to the competition who now is able to deliver that increased performance, at lower or the same prices

but who said anything about "prevent work done on it". never said that at all.

 

[leman]

Did we need an even thinner MBP? Cause frankly that is the cause of situation we are in. In their TDP bracket......and that is the issue.....

Again, not really sure how you arrive to this conclusion. Even when Apple laptops were 1cm thicker than they are now, they never used a GPU with TDP higher than 45W (for obvious reasons). And again, at their release data, the 460 Pro was the fastest card in the sub-50W TDP bracket, so it was a very reasonable choice. Right now this crown goes to 1050 GTX.
So as I've been repeating over and over again, we got the fastest available hardware in the same power bracket as Apple always used AND a thinner laptop. No performance was sacrificed for form factor here. If anything the new laptops have much better heat management.

 

[Stutterer]

So does this mean we could see quad-core 13" MacBook Pros and six or eight core iMac Pros? Because that would be incredible. I've been waiting to get a new iMac so I can get a serious processor upgrade.

Waiting for the next iMac too.

 

[MH01]

Again, not really sure how you arrive to this conclusion. Even when Apple laptops were 1cm thicker than they are now, they never used a GPU with TDP higher than 45W (for obvious reasons). And again, at their release data, the 460 Pro was the fastest card in the sub-50W TDP bracket, so it was a very reasonable choice. Right now this crown goes to 1050 GTX.
So as I've been repeating over and over again, we got the fastest available hardware in the same power bracket as Apple always used AND a thinner laptop. No performance was sacrificed for form factor here. If anything the new laptops have much better heat management.

And since you refuse to see my point, why do you think we are constraint to sub 50w TDP?

What the the main reason for that constraint?

 

[leman]

But today, Apple's "pro" laptop very much trades maximization of performance for "form and design". When you look at the overall difference in chassis size and weight between the 2015 and 2016 models, for users who operate where "faster = more work", 2mm thickness and .3lbs, they have to ask themselves "is that worth a 20% slower GPU, and smaller battery".

Sorry, but this is just factually wrong. Again, at the moment of their release the GPU was the fastest one around of all the available suitable GPUs. And it also was exceptionally energy efficient. I have no idea where you get this "20%" from. There was no sacrifice in performance.


P.S. Yes, they also theoretically had the option of upping the TDP of the GPU and thus giving it slightly better performance. However, I am more then sure that the clocks represent the sweet spot for Polaris 11 — go any higher and the power draw will increase quickly with only minor performance improvements. What they indeed sacrificed was VRAM speed — just as any other workstation manufacturer. This is called stability over gaming performance.

 

[LordVic]

Sorry, but this is just factually wrong. Again, at the moment of their release the GPU was the fastest one around of all the available suitable GPUs. And it also was exceptionally energy efficient. I have no idea where you get this "20%" from. There was no sacrifice in performance.


P.S. Yes, they also theoretically had the option of upping the TDP of the GPU and thus giving it slightly better performance. However, I am more then sure that the clocks represent the sweet spot for Polaris 11 — go any higher and the power draw will increase quickly with only minor performance improvements. What they indeed sacrificed was VRAM speed — just as any other workstation manufacturer. This is called stability over gaming performance.

wasn't even thinking about gaming performance myself, just overall "compute" power. Anyone who wants more GPU for gaming is looking at the wrong company for their devices.

I myself, would have been happier with keeping the thicker chassis, and opting for a higher powered GPU (actual 460, or maybe somehow fit in the 470).

it's just one of those things where Apple chose a compromise, and not everyone is going to be happy with it.

And lets even say that they opted for the 455 in a thicker chassis, they could have retained a larger battery for 15 hour battery life.

p.s. don't get me wrong, these are great laptops. Just IMHO these changes and the price point make it a very tough value proposition for this day of age.

 

[leman]

And since you refuse to see my point, why do you think we are constraint to sub 50w TDP?

What the the main reason for that constraint?

Im assuming it was because Apple engineers decided that they can't efficiently provide cooling to a bigger GPU given the design constraints. And by the way, this decision was made long before MacBook Pro was called MacBook Pro, so you can't claim that they offered more powerful GPU in the past.

Also, the reason why I refuse to see your point is simply because I can't follow your logic. You seem to say that because they made the laptop thinner, they have sacrificed performance. While in fact, they gave us the fastest GPU that was available at the moment within their usual design constraints. They actually doubled the GPU performance compared to previous generation. The 460 is still faster than all NVIDIA Quadros in comparable laptops.

You seem to suggest that Apple could have gotten more performance by staying with the old thicker design. How? Are you so sure that Polaris would perform that much better with increased TDP? Evidence I've seen makes this doubtful.
[doublepost=1496255838][/doublepost]

I myself, would have been happier with keeping the thicker chassis, and opting for a higher powered GPU (actual 460, or maybe somehow fit in the 470).

Well, 460 Pro is the actual 460 its a full Polaris 11 chip unlike the 460rx which is a cut-down chip that's clocked higher to provide 20% more performance at the cost of 2x more power draw. Doesn't really sound like a sound tradeoff.

As to 470 or similar... again, Apple didn't even figure out how to put GPUs of that caliber in when their laptops were 1cm or more so thicker. So unless some principally new cooling system was invented, I don't see a chance for this happening.

What I am trying to say is that Apples goal is to create usable working laptops that have a sensible balance of features and not gimmick machines. Can a 100W GPU be put into a 2cm thin chassis? Yes, manufacturers like Razor prove that. But that means that the entire laptop must be dedicated to cooling down that behemoth, making it not a very practical computer for any purpose other than gaming.

 

[gate5blues]

Things that I'd like to see

touch bar: needs to be moved up a few millimetres to prevent accidental touch, that or they need to do a better job detecting accidental touch.

specs: Even an overpriced overweight battery heavy design to support 32 GB of Ram will appease a few including me. For a developer or anyone running VM's or docker the memory gets eaten quickly and 16GB is sooooo 2000's. I don't care that I have slightly more weight, I can deal with that until the 10nm process is figured out and power consumption is better.

Keyboard:
Well maybe not this release, but with all the MacBook Pro new keyboard issues i'd be hedging my bets on Apple eventually integrating haptic feedback into non-mechanical keys. Just do it already! we all know its coming.

 

[MH01]

Im assuming it was because Apple engineers decided that they can't efficiently provide cooling to a bigger GPU given the design constraints. And by the way, this decision was made long before MacBook Pro was called MacBook Pro, so you can't claim that they offered more powerful GPU in the past.

Also, the reason why I refuse to see your point is simply because I can't follow your logic. You seem to say that because they made the laptop thinner, they have sacrificed performance. While in fact, they gave us the fastest GPU that was available at the moment within their usual design constraints. They actually doubled the GPU performance compared to previous generation. The 460 is still faster than all NVIDIA Quadros in comparable laptops.

You seem to suggest that Apple could have gotten more performance by staying with the old thicker design. How? Are you so sure that Polaris would perform that much better with increased TDP? Evidence I've seen makes this doubtful.
[doublepost=1496255838][/doublepost]

Well, 460 Pro is the actual 460 its a full Polaris 11 chip unlike the 460rx which is a cut-down chip that's clocked higher to provide 20% more performance at the cost of 2x more power draw. Doesn't really sound like a sound tradeoff.

As to 470 or similar... again, Apple didn't even figure out how to put GPUs of that caliber in when their laptops were 1cm or more so thicker. So unless some principally new cooling system was invented, I don't see a chance for this happening.

What I am trying to say is that Apples goal is to create usable working laptops that have a sensible balance of features and not gimmick machines. Can a 100W GPU be put into a 2cm thin chassis? Yes, manufacturers like Razor prove that. But that means that the entire laptop must be dedicated to cooling down that behemoth, making it not a very practical computer for any purpose other than gaming.

Razer with thier limited resources seem to be able to get a 1060 in a 2Kg laptop. Though that is a gaming machine ....not bad

And frankly it's like a 2 hour battery difference ....

Does not matter, you are so bought into that the perfection of a laptop is what apple offers , I don't share that point , and I see the current laptop as form over function. So let's leave it at that . Especially for £4K ...joke performance . My Mac Pro cost less....

Apple frankly has little idea how to get performance from thier hardware, as they cripple their machines with unnecessary thinness . Classic example is the iMacs.....stupid machines throttle themselves , cause for some dumb reason a desktop has to be so damn thin that it cannot get rid of the heat...

Let's agree to disagree .

 

[sracer]

Great, just great. I've been trying to retrain my brain to correctly read "Kaby Lake" (instead of "Baby Cake")... Now "Coffee Lake" (which my brain reads as "Coffee Cake").

 

[LordVic]

Things that I'd like to see

touch bar: needs to be moved up a few millimetres to prevent accidental touch, that or they need to do a better job detecting accidental touch.

specs: Even an overpriced overweight battery heavy design to support 32 GB of Ram will appease a few including me. For a developer or anyone running VM's or docker the memory gets eaten quickly and 16GB is sooooo 2000's. I don't care that I have slightly more weight, I can deal with that until the 10nm process is figured out and power consumption is better.

Keyboard:
Well maybe not this release, but with all the MacBook Pro new keyboard issues i'd be hedging my bets on Apple eventually integrating haptic feedback into non-mechanical keys. Just do it already! we all know its coming.

Touchbar I think would be a great idea if it didn't replace a row of keys on the keyboard. Put it above the function row and offset some of the functionality up there. Currently, My work needs rely on the function keys too much, and I need the ESC key. But, that doesn't mean it should be one or the other. Shrink that tackpad down by a few mm to make room if need be.

Specs: if the device was 2mm thicker (2015 thickness), this wouldn't be even a discussion. The bigger battery allowance, in addition to the better thermals could have easily allows for 32gb RAM, replacable hard drive via M.2 slot, and perhaps better GPU (more thermal headroom, better performancing parts)

Keyboard: go back to the same mechanism they've used in the MBPro's / airs from previous gens. The new keyboards hurt my fingers. I like lots of movement. not these super non-moving keys. Replacing moving keys with a haptic engine would be the death of Apple as a computer manufacturer for me.

 

[ThaRuler]

I'm holding off for "Beer Lake"

 

[flat five]

and your mistake is trying to assign a real objective measurement to what "pro" is. that's not at all what I've said.

"PRO" in my description meant tthat the more power you have available to you, the more work, or the faster your work gets completed. Therefore the ideal would always be "The fastest hardware possible"

sorry LV, this algorithm just isn't adding up
only in the rarest of rare cases will this actually be true.. and in those cases, i don't really think the solution lies in personal computer world.

also, it's almost an insult to people who've worked very hard to master particularly difficult software and become efficient with it.. like-- "what a waste of time that is.. they just need a faster computer in order to get more work completed"

 

[MagnusVonMagnum]

"Up to" are the key words. What's the REAL speed (average)? 2% faster? 3% faster? What's for certain is my 2012 quad i7 was/is faster than a 2015 dual i5 for the Mac Mini. So these newer processors mean JACK SQUAT to me. What we need are BETTER Macs in general (fastest possible minimum 1TB SSDs with 32GB of ram and a killer gaming capable GPU. Give me that with a CPU from 2012 and I'd be 10x happier than with some half-baked paddy cake CPU attached to a POS computer like Apple has been putting out lately.

 

[gadgetfreaky]

Again, not really sure how you arrive to this conclusion. Even when Apple laptops were 1cm thicker than they are now, they never used a GPU with TDP higher than 45W (for obvious reasons). And again, at their release data, the 460 Pro was the fastest card in the sub-50W TDP bracket, so it was a very reasonable choice. Right now this crown goes to 1050 GTX.
So as I've been repeating over and over again, we got the fastest available hardware in the same power bracket as Apple always used AND a thinner laptop. No performance was sacrificed for form factor here. If anything the new laptops have much better heat management.

Isn't AMD 75 TDP?
same with the 1050

 

[leman]

Isn't AMD 75 TDP?
same with the 1050

The desktop versions, yes. The mobile ones are using less power due to chips with more optimal behaviour etc. AMD is particularly impressive here, as they can deliver 80% of the desktop card at 50% of power consumption.

 

[576316]

Here's a question: I currently own a Kaby Lake i5 (7400). If, say, in 3 years I upgraded to a then-current generation equivalent i5 processor, how much faster would it be than my current? What's the performance increase each year with a new Intel generation?

 

[leman]

Here's a question: I currently own a Kaby Lake i5 (7400). If, say, in 3 years I upgraded to a then-current generation equivalent i5 processor, how much faster would it be than my current? What's the performance increase each year with a new Intel generation?

It can't be answered definitively, because the things are not very predictable. For the last few years, the performance increases were rather modest. But just the other day Intel has released a press statement where they say that their upcoming CPUs will be 30% faster than Kaby Lake, and thats on a general-purpose benchmark.

 

[576316]

It can't be answered definitively, because the things are not very predictable. For the last few years, the performance increases were rather modest. But just the other day Intel has released a press statement where they say that their upcoming CPUs will be 30% faster than Kaby Lake, and thats on a general-purpose benchmark.

Thanks for your response. Just trying to work out how many years I might get out of this processor before it's too slow. But for my use case, I can't see it ever being too slow. Unless something dramatic happens with CPU requirements in the PC games industry within the next couple years.

 

[leman]

Oh, not arguing that Apple's display isn't bright. I was just amused that you think brightness is an actual argument when OLED is already out there.

Brightness was only one of the points I was mentioning (others were contrast, wide gamut, color accuracy and energy consumption). Please do not quote my posts out of context. We can all agree that OLED is (most likely) the future, but right the only shipping OLED displays I am aware of are outclassed by MBP's displays in all parameters except of course contrast and response times. Not to mention that there are still some power consumption issues: current OLED just isn't as power-efficient as high-end IPS displays — and by a large margin so. OLED is still very new in computer screen market and it will take some time until it will get mature enough to surpass the mature high-end LED IPS panel. Calling the MBP outdated because of that is nonsensical.

Now that's a straw man. Outdated means it's not up-to-date with current technology. Your straw man is "those other computers". So according to you, so long as MacBook Pro's specs are slightly better than the rest, it's NOT outdated. Your measure of being up-to-date is not vis-à-vis current technology, which is how it should be, but rather vis-à-vis the laptops that are currently on the market.

First, my measure is actual tech availability. The main reason why no laptop currently ships with BT 5.0 support is lack of appropriate hardware. The chips required to have BT 5.0 in laptops are only starting to pop up around now. Given the current computer market situation, BT 4.2 is still state of the art. I do hope that BT 5.0 will become widely available as soon as possible and I have no doubts that Apple will be one of first to adopt it (as they always had with BT).

Second, my issue with your rhetorics is that you use the term "obsolete" in a very one-sided manner. Sure, technically you are right, BT 4.2 in the MBP is obsolete because a new spec is available. But the fact is that Apple's BT support is literally best in the entire computer market. Your rhetorics deliberately ignores this. Same goes for the majority of your other points. Yes, Apple uses the "obsolete" IPS/LED tech instead of newer OLED — but their displays are still better than what OLED can currently offer. I find your rhetorics highly manipulative and not constructive. It would be one thing to say "Oh, I wish that Apple would also offer us the new BT 5.0" or "Apple should invest more in OLED panel research" (which they do btw and they will use it when they think its ready). But what you do is paint this really weird picture of MBPs as using this outdated, inferior tech, which is fairly far from the reality.

P.S. Sorry for the later answer, I didn't discover your post until now.

 

[cult hero]

Quad in a 13 would be so nice. I really have nothing else to add.

 

[lysingur]

Brightness was only one of the points I was mentioning (others were contrast, wide gamut, color accuracy and energy consumption). Please do not quote my posts out of context.

On all those other points, LED/LCD technology isn't a clear winner, nor is the display used in MacBook Pro 2016 compared to the best OLED display in 2017. Just because I didn't mention something doesn't mean I'm out to misrepresent your words. You're right that brightness wasn't the only point you mentioned but it is one of your points.

Not to mention that there are still some power consumption issues: current OLED just isn't as power-efficient as high-end IPS displays — and by a large margin so.

I would love to see where you get this, especially the part about "a large margin". It also should be noted that "high-end IPS displays" aren't that much cheaper than OLED displays. So it's pretty much a moot point to say the best of LED/LCD is still competitive against OLED because LED/LCD's margin of improvement as a technology is very limited.

OLED is still very new in computer screen market and it will take some time until it will get mature enough to surpass the mature high-end LED IPS panel. Calling the MBP outdated because of that is nonsensical.

Retina display was pretty much just as new as OLED if not more so as a technology when it was shipped with MBP in 2012. It was so new that few mobile graphics cards could drive it adequately at native resolution. Apple should do with OLED what it did with USB, that is by being the first to adopt it in their laptop. It's not really asking too much since it's the sort of thing they were known to do.

First, my measure is actual tech availability.

I would refer you to the above point on Retina display. Your definition of being "up-to-date" is something probably only you use. Whether a high-end laptop is up-to-date doesn't hinge on "tech availability". It actually puzzles me that you'd think that. It seems you're just trying to give "outdated"/"up-to-date" a new definition to suit your own argument.

But what you do is paint this really weird picture of MBPs as using this outdated, inferior tech, which is fairly far from the reality.

Outdated, obviously. Inferior? Probably not since whether MBP is inferior depends on similar laptops at the same price range that you can buy.

The chips required to have BT 5.0 in laptops are only starting to pop up around now. Given the current computer market situation, BT 4.2 is still state of the art.

Please refer to my above points about Retina display and definition issues regarding "up-to-datedness". No, BT 4.2 isn't state-of-the-art at all. It's a technology that's two years old and was only a minor update to existing Bluetooth technology when it was first release at the end of 2014. Your argument for BT 4.2 being state-of-the-art wore even thinner when you consider that there is already a phone I can have right now that uses BT 5.0 and MBP mid-2017 is still months away from shipping. If MBP mid-2017 doesn't ship with BT 5.0, it'd not only be a huge let-down but a slap in the face for Apple for proclaiming courageously that they would take the lead in ushering in the wireless future.

 

[gadgetfreaky]

The desktop versions, yes. The mobile ones are using less power due to chips with more optimal behaviour etc. AMD is particularly impressive here, as they can deliver 80% of the desktop card at 50% of power consumption.

if I look up the mobile card - TDP =75 not 45.

 

[leman]

if I look up the mobile card - TDP =75 not 45.

I am not sure what you are referring to. I was comparing the desktop RX 460 (Polaris11 with 14 cores @ 1200Mhz, 75W TDP) and the mobile Pro 460 (Polaris11 with 16 cores @ 900 Mhz, 35W TDP). Thats roughly 20% less total performance at 50% power consumption.
[doublepost=1496393581][/doublepost]

On all those other points, LED/LCD technology isn't a clear winner, nor is the display used in MacBook Pro 2016 compared to the best OLED display in 2017. Just because I didn't mention something doesn't mean I'm out to misrepresent your words. You're right that brightness wasn't the only point you mentioned but it is one of your points.

[...]

I would love to see where you get this, especially the part about "a large margin". It also should be noted that "high-end IPS displays" aren't that much cheaper than OLED displays. So it's pretty much a moot point to say the best of LED/LCD is still competitive against OLED because LED/LCD's margin of improvement as a technology is very limited.

[..]

Apple should do with OLED what it did with USB, that is by being the first to adopt it in their laptop.

I am getting this from actual hardware reviews. Independent measurements of these displays and their performance are available and you can easily find them on reviewers websites. Comparing displays currently on the market, the MBP 2016 display has objectively better brightness, better color accuracy and dramatically lower power consumption, which makes it a superior screen for use in a laptop as of now. OLED wins on response time and contrast. I am sure that within few years OLED will the lead, but as of now, its production quality is simply not the same, especially when we look at power consumption. I am unsure why you'd want Appel to use a display that would mot likely cut the battery time of their laptops in half without providing any decisive benefits in return.

And finally, I never talked about price. Why are you bringing up price? I'd guess that the IPS display Apple uses costs more than the OLED panel Dell/Lenovo use.

I would refer you to the above point on Retina display. Your definition of being "up-to-date" is something probably only you use. Whether a high-end laptop is up-to-date doesn't hinge on "tech availability". It actually puzzles me that you'd think that. It seems you're just trying to give "outdated"/"up-to-date" a new definition to suit your own argument.

Again, my definition is about practical availability. IPS being old tech doesn't change the fact that it produces better screens than OLED at this moment. So calling it outdated is nonsensical. Same with Bluetooth. The current version of standard is pretty much irrelevant of there are no actual chips that support it. Sure, Apple could try using the same BT/WiFi chip that Samsung uses... and suffer significant cut in WiFi performance.

What I am saying is that your talk of obsoleteness makes no sense if there is no shipping hardware that would take advantage of all the benefits of new standards/tech.