Waiting for the 2018 MBP

[HerbertDerb]

New details emerge about the new Coffee-Lake HQ processors. Still this is an open question - will the new Macbook Pro 15 get hexacore Coffee-Lakes or quad-core Kaby-Lake G with Vega graphics?

https://videocardz.com/74983/gigaby...fee-lake-h-cpus-coming-late-march-early-april

My reaction if Apple decides to go with Kaby Lake-G option in the 15" lineup.

I want 6-core 15" MBP.

 

[Poki]

My reaction if Apple decides to go with Kaby Lake-G option in the 15" lineup.

I want 6-core 15" MBP.

Yeah, Kaby Lake G would be worst case. Let's just not think about that and read more benchmarks of Intel's six core chips.

 

[afir93]

Yeah, Kaby Lake G would be worst case. Let's just not think about that and read more benchmarks of Intel's six core chips.

For the uninitiated, what would Kaby Lake G mean in terms of performance increase? From my understanding, the jump to Intel's six core lineup would bring about 30-40% CPU performance increase, I'd assume the Kaby Lake G in comparison would only mean the usual 5-10% increase? If so then yeah that would be a big letdown.

Would there be any reason for Apple to choose the Kaby Lake G chips over the 8th gen six core ones? Like, do they have any advantages at all, like maybe more energy-efficiency or less internal space usage (which both might result in longer battery lifes)? Or do they have no advantages at all compared to the six core's? If so I can't really see Apple use an objectively worse option when the other chips are available to ship at around the same time.

 

[Loge]

Well if they go with the 6-core, they really do need to provide a 32GB option somehow, with hyperthreading a 16GB machine is only going to have just over 1GB per virtual core.

 

[Poki]

For the uninitiated, what would Kaby Lake G mean in terms of performance increase? From my understanding, the jump to Intel's six core lineup would bring about 30-40% CPU performance increase, I'd assume the Kaby Lake G in comparison would only mean the usual 5-10% increase? If so then yeah that would be a big letdown.

Would there be any reason for Apple to choose the Kaby Lake G chips over the 8th gen six core ones? Like, do they have any advantages at all, like maybe more energy-efficiency or less internal space usage (which both might result in longer battery lifes)? Or do they have no advantages at all compared to the six core's? If so I can't really see Apple use an objectively worse option when the other chips are available to ship at around the same time.

We're looking more at a 0 – 5 % CPU performance increase from Kaby Lake to Kaby Lake G. The GPU is a little faster than a GTX 1050, so that would be an upgrade – but not any kind of upgrade that's out of reach with a new Vega-based mobile GPU at the current 35W TDP that's reserved for the graphics.

From all reports surfaced so far, the Kaby Lake G platform is not nearly as efficient as Intel claimed, so it's mainly just a 7th gen Kaby Lake H CPU paired with a Vega mobile GPU. And it's not even cheap – Intel didn't release official prices, but their own NUCs featuring these chips are pretty expensive. So yeah, there are not a lot of reasons why Apple should use one of these chips. Pretty much the only valid one would be the lack of availability of Intel's 8th gen H-series chips and the lack of a dedicated Vega-based mobile GPU. But even then, going with a higher wattage GPU and a 15W U-series Kaby Lake R would result in better overall performance.

 

[afir93]

We're looking more at a 0 – 5 % CPU performance increase from Kaby Lake to Kaby Lake G. The GPU is a little faster than a GTX 1050, so that would be an upgrade – but not any kind of upgrade that's out of reach with a new Vega-based mobile GPU at the current 35W TDP that's reserved for the graphics.

From all reports surfaced so far, the Kaby Lake G platform is not nearly as efficient as Intel claimed, so it's mainly just a 7th gen Kaby Lake H CPU paired with a Vega mobile GPU. And it's not even cheap – Intel didn't release official prices, but their own NUCs featuring these chips are pretty expensive. So yeah, there are not a lot of reasons why Apple should use one of these chips. Pretty much the only valid one would be the lack of availability of Intel's 8th gen H-series chips and the lack of a dedicated Vega-based mobile GPU. But even then, going with a higher wattage GPU and a 15W U-series Kaby Lake R would result in better overall performance.

Ok thanks for the breakdown. Let's hope we'll see some six-cores then indeed

 

[Queen6]

I strongly suspect Apple has more than enough metrics on it's MBP usage. Ultimately if software is not capable of taking advantage of more cores, the net effect may result in slower performance depending on the max frequency of the cores.

Time will tell, equally Apple has never followed the numbers game...

Q-6

 

[netmaskit]

I strongly suspect Apple has more than enough metrics on it's MBP usage. Ultimately if software is not capable of taking advantage of more cores, the net effect may result in slower performance depending on the max frequency of the cores.

Time will tell, equally Apple has never followed the numbers game...

Q-6

It was right in the past.
Today, with virtual machines and Docker containers running everywhere, and people using dozens and even hundreds of browser windows and tabs simultaneously, multi-cores are being used to the optimum, and even an OCTA (or 16!) will be welcome, if Intel releases something like this sometimes in the future ;-)
The real change will come to the 13" users, who will see a double of their cores (from 2 to 4).
And yet, my real dream is of a 15" machine with "only" 4 cores, but with lower power CPU (e.g. i7-8650U), where the extra power that is saved, will be used to replace the economic LPDDR3 by DDR4 (for 32GB support!).

 

[argentum47]

And yet, my real dream is of a 15" machine with "only" 4 cores, but with lower power CPU (e.g. i7-8650U), where the extra power that is saved, will be used to replace the economic LPDDR3 by DDR4 (for 32GB support!).

If there is extra power saved, it will be used to make the product thinner by shrinking the battery, not for increasing some numbers they see as secondary.

I should put a /s there, but that's exactly what Apple has been doing for the past few years so I really can't.

 

[netmaskit]

If there is extra power saved, it will be used to make the product thinner by shrinking the battery, not for increasing some numbers they see as secondary.

I should put a /s there, but that's exactly what Apple has been doing for the past few years so I really can't.

This is why I wrote it under the "dream" title.
But in any case, a refresh is urgent, at least for the 13" which has currently only 2 cores and priced higher than 4-core competitors.

 

[rebel_mac]

We're looking more at a 0 – 5 % CPU performance increase from Kaby Lake to Kaby Lake G. The GPU is a little faster than a GTX 1050, so that would be an upgrade – but not any kind of upgrade that's out of reach with a new Vega-based mobile GPU at the current 35W TDP that's reserved for the graphics.

From all reports surfaced so far, the Kaby Lake G platform is not nearly as efficient as Intel claimed, so it's mainly just a 7th gen Kaby Lake H CPU paired with a Vega mobile GPU. And it's not even cheap – Intel didn't release official prices, but their own NUCs featuring these chips are pretty expensive. So yeah, there are not a lot of reasons why Apple should use one of these chips. Pretty much the only valid one would be the lack of availability of Intel's 8th gen H-series chips and the lack of a dedicated Vega-based mobile GPU. But even then, going with a higher wattage GPU and a 15W U-series Kaby Lake R would result in better overall performance.

How long do you think we will at least get a confirmation or leaked specs about the upcoming 15"?

I'm at a crossroads since I just picked up a 15" TouchBar 512GB for $200 off in Canada, but I also have accidental/extended warranty on it.

I figured the keyboard on the 2017 will likely fail in two years leading to a free upgrade, which happened with my last gaming laptop.

Should I hold out for a March announcement? I can get by with my iPad Pro and main rig at home for the time being.

Or just keep the 2017 and upgrade in 2019?

I guess I am curious how likely it will be for us to get a 6 Core in the 15" in 2018?

Or will Apple wait till 2019 to refresh the chassis, ram and processor?

 

[Poki]

How long do you think we will at least get a confirmation or leaked specs about the upcoming 15"?

I'm at a crossroads since I just picked up a 15" TouchBar 512GB for $200 off in Canada, but I also have accidental/extended warranty on it.

I figured the keyboard on the 2017 will likely fail in two years leading to a free upgrade, which happened with my last gaming laptop.

Should I hold out for a March announcement? I can get by with my iPad Pro and main rig at home for the time being.

Or just keep the 2017 and upgrade in 2019?

I guess I am curious how likely it will be for us to get a 6 Core in the 15" in 2018?

Or will Apple wait till 2019 to refresh the chassis, ram and processor?

The 2018 refresh will most probably be just a speed bump, i.e. new CPUs and GPUs. The RAM limit will stay at 16 GB due to the limitations of Intel's LPDDR3 support. I don't expect a lot of other changes.

As for when it will happen – only time will tell. Might be March, might be June. But this might as well be a silent update, i.e. no rumors, they might just drop it into their website at some random point in time.

Also, don't hope for a "free upgrade". If they need to replace your MacBook Pro in two or even three years time, you are going to get the same model as you have now. The occasions where Apple actually swaps devices for newer models are very rare.

 

[TH3D4RKKN1GH7]

Man, the keyboards are really THAT bad? Insane. Wonder if those will get upgraded for the 2018 model. Seems like a keyboard redesign would be the most important upgrade.

 

[rebel_mac]

The 2018 refresh will most probably be just a speed bump, i.e. new CPUs and GPUs. The RAM limit will stay at 16 GB due to the limitations of Intel's LPDDR3 support. I don't expect a lot of other changes.

As for when it will happen – only time will tell. Might be March, might be June. But this might as well be a silent update, i.e. no rumors, they might just drop it into their website at some random point in time.

Also, don't hope for a "free upgrade". If they need to replace your MacBook Pro in two or even three years time, you are going to get the same model as you have now. The occasions where Apple actually swaps devices for newer models are very rare.

So, do you expect a huge boost in speed it is a 6 core?

What's the guarantee of a 6 core happening?

The extended is through a third party with a lemon/accidental policy.

They just issue a gift card for the full amount of the MacBook you paid for and let you choose what you want + add the difference.

 

[Poki]

So, do you expect a huge boost in speed it is a 6 core?

What's the guarantee of a 6 core happening?

No guarantees, but with pretty much all 65W desktop chips moving to six cores, it's fair to assume that the more expensive 45W H-series mobile chips will move to six cores as well. Especially since AMD already teased six-core Ryzen Mobile chips.

Performance improvement should be somewhere between 30 – 50 % in multi threaded workloads.

 

[oldtime]

Well if they go with the 6-core, they really do need to provide a 32GB option somehow, with hyperthreading a 16GB machine is only going to have just over 1GB per virtual core.

32GB is not going to happen this year. I mean, it literally can't with the rumored Coffee Lake upgrade. Next year though, it's certainly possible.

 

[awesomedeluxe]

32GB is not going to happen this year. I mean, it literally can't with the rumored Coffee Lake upgrade. Next year though, it's certainly possible.

One of many reasons 2018 will likely disappoint enthusiasts:

1. Still on 16GB of RAM. I've had 16GB since 2012.

2. Still on Intel's tired 14nm process. Coffee Lake will bring more cores, but not much more performance per core (and the 13 inch will probably use Kaby-G)

3. Still using AMD. nVidia is worlds ahead of them in power/performance. That's fine for the 13", especially if you're leveraging Intel's manufacturing. But Max-Q notebooks will walk all over the Vega part in the 15".

I'd like to say this means 2019 will be huge. But holding out for Intel to finally move to 10nm is practically a fool's game at this point.

 

[Loge]

https://theblog.adobe.com/announcing-february-update-lightroom-classic/

New Lightroom, with benefits for computers with at least 12GB of RAM. Getting hard to justify the MBP for this, particularly if you want it to last you a number of years.

 

[Queen6]

It was right in the past.
Today, with virtual machines and Docker containers running everywhere, and people using dozens and even hundreds of browser windows and tabs simultaneously, multi-cores are being used to the optimum, and even an OCTA (or 16!) will be welcome, if Intel releases something like this sometimes in the future ;-)
The real change will come to the 13" users, who will see a double of their cores (from 2 to 4).
And yet, my real dream is of a 15" machine with "only" 4 cores, but with lower power CPU (e.g. i7-8650U), where the extra power that is saved, will be used to replace the economic LPDDR3 by DDR4 (for 32GB support!).

Don't disagree, really very much dependant on usage, workflow and software involved. See some pretty big numbers being presented, although without specific use case users may or may not advantage. The good is this will promote more and more developers to capitalise on additional cores.

Apple could have offered 32Gb, had they not been so obsessive with producing ever thinner notebooks, lopping 25% off the battery capacity was a "genius" move, net result no progression...

Q-6

 

[oldtime]

One of many reasons 2018 will likely disappoint enthusiasts:

1. Still on 16GB of RAM. I've had 16GB since 2012.

2. Still on Intel's tired 14nm process. Coffee Lake will bring more cores, but not much more performance per core (and the 13 inch will probably use Kaby-G)

3. Still using AMD. nVidia is worlds ahead of them in power/performance. That's fine for the 13", especially if you're leveraging Intel's manufacturing. But Max-Q notebooks will walk all over the Vega part in the 15".

I'd like to say this means 2019 will be huge. But holding out for Intel to finally move to 10nm is practically a fool's game at this point.

I think both 2018 and 2019 could be nice if people set their expectations correctly. Coffee Lake will bring a rather significant performance boost compared to your average year to year update. That alone is enticing if you're buying this year.

In 2019 I think we'll get to 10nm and 32GB, finally. WWDC 2019 could give us not only Cannonlake MBPs but maybe even a redesign, especially if the keyboard issues simply can't be ironed out any better than they already are. Yes, that's probably a year early for a redesign, but Apple can't be loving the lukewarm reception that the latest MBP has received.

 

[Poki]

One of many reasons 2018 will likely disappoint enthusiasts:

1. Still on 16GB of RAM. I've had 16GB since 2012.

2. Still on Intel's tired 14nm process. Coffee Lake will bring more cores, but not much more performance per core (and the 13 inch will probably use Kaby-G)

3. Still using AMD. nVidia is worlds ahead of them in power/performance. That's fine for the 13", especially if you're leveraging Intel's manufacturing. But Max-Q notebooks will walk all over the Vega part in the 15".

While a higher RAM limit would certainly be nice, I guess for the majority of MacBook Pro customers 16 GB will be fine.

I don't agree on your points 2 and 3 though.

2. Pretty much all power hungry applications can use multiple cores by now, so this performance boost is going to be the biggest one in years. Also, Kaby Lake G, at the lowest end, have a TDP of 65W, compared to 15W / 28W currently used in the 13" MBPs. Not gonna happen.

3. That's actually no longer true. Even a GTX 1050 needs more power than the Radeon Pro 560, let alone the GTX 1060 max-Q at almost twice the TDP. If you break it down to performance per watt, AMD is actually very close to NVidia - and that's for a chip that basically launched in October 2016. Vega is even more efficient, I could see AMD surpassing NVidia in performance / watt again.

 

[The Mercurian]

Leak for new HP hex-core laptop suggests May 2018 for release:

https://www.notebookcheck.net/Leak-...with-Coffee-Lake-Hexa-Core-CPUs.282882.0.html

 

[domsodidavid]

I think we'll get new macbooks in the first half of april.

 

[lpolarityl]

Well if they go with the 6-core, they really do need to provide a 32GB option somehow, with hyperthreading a 16GB machine is only going to have just over 1GB per virtual core.

That would be ideal, but the cost of memory right now is just... bad. Both SSD's and memory is a lot more than it was a year ago. I can only imagine what Apple will charge for a MBP 15" / 6 Core | 12 Thread / 32GB Memory / 512GB variant. I do agree though, the next MBP 15" should come with a 6 Core | 12 thread Coffe Lake variant and it should come with 32GB of memory, but we all know its going to come with 16GB by default and 32GB will be a incredibly expensive option. They'd have to use two low voltage 16GB DIMM's, so you know its going to be very expensive.

 

[leman]

I am getting rather pessimistic about the whole thing. The most recent leak mention Coffee Lake H series towards the end of Q2 2018 and no mention of U series at all. So components suitable for the MBP (especially the 13" one) can still be months and months off.