Details on Upcoming Amber Lake Chips Appropriate for 2018 MacBook Leak

[sublunar]

The Whiskey Lake 15W parts you reference are just 14nm++ parts that replace the i5-8250U, i5-8350U, i7-8550U and i7-8650U 14nm+ parts that were released last August. Quad with GT2 graphics. Apple wasn’t interested in them last year, but sure they could go into a machine meant to be a MBA replacement.

Regarding a possible new 13”, I don’t think the way to get a $999 MBA replacement is to introduce a larger version of the $1,299 12” MacBook. The $1,299 MB or $1,299 nTB MBP could be reduced to $999, or the MBA in its current form factor can be upgraded with a better processor and presumably a retina screen.

Or, a completely new 13” machine is introduced to replace both MBA and the nTB MBP. But they already have the current nTB, why a new machine? This is what doesn’t make sense about a new 13” machine. Apple always said the 13” nTB is the replacement for the Air.

How does a brand new machine become a better $999 replacement for the MBA than the current nTB—a platform that has easily sold 10-15 million units (nTB+TB) already.

I think they might have arrived a bit too late for the June 2017 MacBook release, and the MacBook may now be in a prime position to replace the MBA now that we've had 3 generations for Apple to recoup the development costs.

Currently the nTB is priced distinctly above the MBA range, but if Apple are to drop Thunderbolt 2 machines from the range something has to fill the $1000 Mac price range and I think it would be odd for Apple to simply refresh the MBA with Thunderbolt 3 ports and keep the existing large bezelled TN non retina display.

A price re-alignment may be due for the 12" model, especially if they introduce as 128Gb SKU which they would have to if there is to be a 13" model with 128Gb SKU.

Replacing the nTB with a model called the rMB 13" might be as simple as re-using the nTB case design with a 15w CPU with non-Iris GPU where Apple can save a few dollars. Apple could also shave a few more dollars off the cost if they were to use USB-C ports (at Gen 2 speeds - 10Gbps) which I believe would be possible on 300 series motherboards paired with Whiskey Lake CPUs.

It might even be feasible for Apple to spend engineering resources to allow 4 ports if they were all USB-C (Gen 2/10GBps)

 

[EugW]

That would require the use of a Thunderbolt controller chip, I guess.

The reason why Apple haven't fitted so far?
a. A cost reduction measure
b. It doesn't fit on the (tiny) motherboard
c. Heat and power consumption considerations are an issue
d. Marketing considerations

Since a 2018 MacBook refresh is probably going to include consideration for a 3rd generation butterfly keyboard (for extra quiet operation) I'd like to think that the MacBook would adopt a Thunderbolt port at the same time if the MBA was put out to pasture at the same time as the 2018 refresh.

I meant for 2019. If Ice Lake Y ships in 2019, then the MacBook could get Thunderbolt with no external controller chip needed.

BTW, at that time, the MacBooks would be solidly faster than the fastest MacBook Airs too, as the MacBooks would be quad-core.

 

[sir1963nz]

Just a note that MacBooks have two ports, one of them being an audio out.

I note with interest that a 13" MBP I have in front of me has the headphone socket on the RHS at the rear, which means the cable from my headphones now gets to dangle across the keyboard. VS my dead 2011 MBP which had it on the LHS near the front.

Apple used to get these details right, they have either lost the ability or no one seems to care that much anymore.

 

[Zdigital2015]

I note with interest that a 13" MBP I have in front of me has the headphone socket on the RHS at the rear, which means the cable from my headphones now gets to dangle across the keyboard. VS my dead 2011 MBP which had it on the LHS near the front.

Apple used to get these details right, they have either lost the ability or no one seems to care that much anymore.

Despite all that I might disagree with you regarding the MacBook, I absolutely despise this idiotic design decision. Headphone lines universally come out of the left side can, which somebody at Apple should have caught and made sure the jack was on the left hand side. I use wired headphones as much as wireless and it’s just a giant PITA.

 

[EugW]

Despite all that I might disagree with you regarding the MacBook, I absolutely despise this idiotic design decision. Headphone lines universally come out of the left side can, which somebody at Apple should have caught and made sure the jack was on the left hand side. I use wired headphones as much as wireless and it’s just a giant PITA.

Not all come out the left side, but yeah most do. But more importantly, Apple doesn't care because their main headphones are either earbuds where the wires go to both sides, or else Bluetooth. Yeah, they have Beats wired headphones, but most of those are legacy, and they are pushing wireless hard.

That's why I think Apple should just remove the headphone jack from the MacBook if they're set on having only two ports. Make the second port another USB-C port, and probably more people would be happy. They could sell a headphone jack dongle for $9.99.

I already own two sets of Bluetooth headphones with W1, and one I got from Apple's Back-To-School sale last year. ie. Apple is giving away wireless headphones these days, and they've already removed the headphone jack from the iPhone, so the next step is to remove it from the 12" MacBook.

 

[curmudgeonette]

Well, I hate to disagree with you, but the DisplayPort Alt-Mode on the USB-C connector is supposed to support...

I think that's what I said! Let me expand:

The presence of USB 3.1 Gen 2 (10Gbps) won't help provide USB 3 speeds when all high speed USB-C wires are being used for a "4 bit wide" DisplayPort 1.2 signal driving a 4Kx60 display. There are no wires left for USB 3.1 Gen 2 nor Gen 1.

What helps is a GPU that can output a "2 bit wide" DisplayPort 1.3/1.4 signal, thus leaving wires for USB 3. This requires DP 1.3/1.4 capability in the GPU, the USB-C mux, and the connected display.

This is similar to how a rMB12 can drive a 1440p display while also supporting USB3. (1440p requires the full cable width in DP 1.0/1.1, but only half at DP 1.2 speeds.)

Which also begs the question, since the 8th-Gen 28w U-Series CPUs now have x16 instead of x12 PCIe lanes, does the new 2018 13" MacBook Pro now support full speed across all four (4) Thunderbolt 3 ports? I haven't seen anything to tell me they changed that. Has anyone reported it?

I think someone has so reported!

 

[sublunar]

Well, I hate to disagree with you, but the DisplayPort Alt-Mode on the USB-C connector is supposed to support DisplayPort 1.4a, not 1.2- https://www.displayport.org/faq/#tab-displayport-1-4a-standard'

Will DisplayPort Alt Mode on the Type-C connector support DisplayPort 1.4a?
Yes. All of the features associated with DisplayPort 1.4a will be available for the DisplayPort Alt Mode.

According to this site - https://www.extron.com/product/videotools.aspx - 4K@60, 8-bit, 4:2:0 - would consume 8.91Gbps, so I assume that its possible over Type 2. Yes, we would be stuck at USB 2.0 for peripherals...hardly a satisfactory approach, although I saw the DisplayPort website say that USB-C over Alt-Mode supported 4K@60 AND USB 3.0 speeds. I swear. I will post the link when I find it.

At this point, two USB-C ports support Type 2(10Gbps) on a new 13" MacBook sure would be nice...I guess I can dream.
[doublepost=1531954843][/doublepost]
Thanks for the clarification there, that was very enlightening!

It does makes a bit more sense, and would explain why the 2016-2017 13" MacBook Pro half slower TB3 ports on the right hand side.

Which also begs the question, since the 8th-Gen 28w U-Series CPUs now have x16 instead of x12 PCIe lanes, does the new 2018 13" MacBook Pro now support full speed across all four (4) Thunderbolt 3 ports? I haven't seen anything to tell me they changed that. Has anyone reported it?

Thunderbolt 3 would be my preferred solution for the MacBook 2018 series, even if they stuck to 1 port for the 12" MacBook and 2 ports for a potential 13" one based on the existing nTB MacBook Pro. They are after all supposedly ultra books with the intent of using on the hoof - even with dongles.

I had assumed that the SSD, Wifi, and Bluetooth (4x PCIe plus 1x and 1x) came off the PCH in the case of the 13" 2017 models with Apple prioritising unfettered access to the CPU for the Thunderbolt ports.

I don't think it'll be long before we find out if the newly refreshed 2018 13" MacBook Pro has 4 full speed ports.
[doublepost=1531960575][/doublepost]

Not all come out the left side, but yeah most do. But more importantly, Apple doesn't care because their main headphones are either earbuds where the wires go to both sides, or else Bluetooth. Yeah, they have Beats wired headphones, but most of those are legacy, and they are pushing wireless hard.

That's why I think Apple should just remove the headphone jack from the MacBook if they're set on having only two ports. Make the second port another USB-C port, and probably more people would be happy. They could sell a headphone jack dongle for $9.99.

I already own two sets of Bluetooth headphones with W1, and one I got from Apple's Back-To-School sale last year. ie. Apple is giving away wireless headphones these days, and they've already removed the headphone jack from the iPhone, so the next step is to remove it from the 12" MacBook.

Phil Schiller has previously said they retain the headphone jack because of professionals who use studio monitor headphones (which aren't wireless). I'd go along with dropping the headphone socket from the MacBook range (not the Pro) if they added another USB-C socket (one on each side) which people could use to add a DAC amplifier for wired headphones for example.
[doublepost=1531960991][/doublepost]

I meant for 2019. If Ice Lake Y ships in 2019, then the MacBook could get Thunderbolt with no external controller chip needed.

BTW, at that time, the MacBooks would be solidly faster than the fastest MacBook Airs too, as the MacBooks would be quad-core.

Given the mounting evidence that Intel will be delaying 10nm CPUs, even though the 5w range looks most likely to get released, would Apple not be keeping an eye using a possible future ARM CPU in the MacBook?

It's the obvious first machine to migrate over and made easier by not going Thunderbolt 3. And if Intel continue to drag their feet for bigger ticket CPUs then it also makes sense for Apple to look to GPU to provide annual improvements for the purposes of refreshes? AMD Navi has to be on the radar if Apple continue to avoid AMD Ryzen CPUs.

 

[EugW]

This is similar to how a rMB12 can drive a 1440p display while also supporting USB3. (1440p requires the full cable width in DP 1.0/1.1, but only half at DP 1.2 speeds.)

That works out well for legacy Apple products since old Cinema Displays and iMacs are 1440p.

The main problem is finding a USB-C hub that actually supports both Mini-DisplayPort and USB 3. There are a few like the Satechi, but the ones I've seen on Amazon have mediocre ratings, yet are quite expensive.

98% of the combo video + USB hubs out there are HDMI. The good news I suppose is 1440p is cheap, but then again not Retina. 4K is highly desirable.
[doublepost=1531966685][/doublepost]

Phil Schiller has previously said they retain the headphone jack because of professionals who use studio monitor headphones (which aren't wireless). I'd go along with dropping the headphone socket from the MacBook range (not the Pro) if they added another USB-C socket (one on each side) which people could use to add a DAC amplifier for wired headphones for example.

And Apple could also sell a USB-C dongle for people who just need a headphone jack. Having that USB-C port adds a ton of flexibility.

Given the mounting evidence that Intel will be delaying 10nm CPUs, even though the 5w range looks most likely to get released, would Apple not be keeping an eye using a possible future ARM CPU in the MacBook?

It's the obvious first machine to migrate over and made easier by not going Thunderbolt 3. And if Intel continue to drag their feet for bigger ticket CPUs then it also makes sense for Apple to look to GPU to provide annual improvements for the purposes of refreshes? AMD Navi has to be on the radar if Apple continue to avoid AMD Ryzen CPUs.

I think there are two things here. One is the 13" MacBook. One is the ARM MacBook. I think both are in the pipeline, but it's a matter of when, and they probably won't land at the same time.

 

[falcn]

Do people buy MBs? Isn't the 13" Pro nTB a better deal?

I don't need a lot of computational power, so I sold my 15" rMBP and bought 2017 MB.
It's super-light, so I can carry it with me in situations where I would leave 13" MBP at home.

When I'm at home, it's connected to a 27" 4k monitor (3840x2160@60Hz), Ethernet, external audio, mouse, keyboard, and two external disk drives. And it's only 1 wire.

It's not super-fast, but it does the job, so I keep it as my only computer.
I wish I got 16Gb version (didn't noticed they added this option for 2017 model), and I wish it had Thunderbolt instead of plain USB. Also 720p camera is a joke.

If they add thunderbolt, I'll be very tempted to upgrade.

 

[Zdigital2015]

I think that's what I said! Let me expand:

The presence of USB 3.1 Gen 2 (10Gbps) won't help provide USB 3 speeds when all high speed USB-C wires are being used for a "4 bit wide" DisplayPort 1.2 signal driving a 4Kx60 display. There are no wires left for USB 3.1 Gen 2 nor Gen 1.

What helps is a GPU that can output a "2 bit wide" DisplayPort 1.3/1.4 signal, thus leaving wires for USB 3. This requires DP 1.3/1.4 capability in the GPU, the USB-C mux, and the connected display.

This is similar to how a rMB12 can drive a 1440p display while also supporting USB3. (1440p requires the full cable width in DP 1.0/1.1, but only half at DP 1.2 speeds.)



I think someone has so reported!

I managed to find a thread on Reddit where a System Information screenshot showing Thunderbolt Bus 1 at 40Gbps per port was offered as proof that the right side TB3 ports on the 2018 13" MacBook Pro now matched the left side. Despite several attempts to Search out a more authoritative source for the info have proven fruitless (har har).

I apologize for my previous post as I wrote it in haste and it does not come across as intended. I was relying on the DisplayPort.org website for much of my information, but I cannot reconcile the numbers they throwing around for DisplayPort over USB-C given the bandwidth requirements of 4K/60, 5K/60 and 8K/60 that I am seeing using the "4K Data Rate Calculator" on the Extron website. The DisplayPort website has some nice marketing bullets, but seems a bit short on technical details - https://www.displayport.org/displayport-over-usb-c/ - among those are:

The most advanced A/V display connection technology now uses the most versatile connector.

• Full DisplayPort audio/video (A/V) performance (up to 8k at 60Hz)
• SuperSpeed USB (USB 3.1) data
• Up to 100 watts of power over a single cable
• Reversible plug orientation and cable direction
• Backward compatibility to VGA, DVI, and HDMI with plug adapters or adapter cables
• Adapters supports HDMI 2.0a and full 4K UHD resolution

I am currently using a BenQ SW271 (refurb) that includes a USB-C connector which allows me to have 4K/60 and USB 2.0 bandwidth for the onboard USB 3.0 ports and the SDXC card reader OR 4K/30 and USB 3.0 bandwidth for the onboard USB 3.0 ports and the SDXC card reader. Both of the data rates calculated come in very close to or exceed the 10Gbps of USB 3.1 Type 2, so I am having trouble reconciling how this is being done and providing the full 5.0Gbps bandwidth for USB 3.0 devices. As I read more about how USB-C signaling occurs and the DisplayPort's use of signal compression across USB-C, the picture has become a bit clearer. I tend to forget that just because a protocol has a set bandwidth (10Gbps, 40Gbps, et al.) does not necessarily mean that engineers cannot figure out clever ways of sending more data via compression or using different channels or sending data between intervals from other devices. Thanks for the information.

 

[EugW]

Yes, all four TB3 ports on the 2018 MacBook Pros are the same speed now.

 

[Zdigital2015]

Yes, all four TB3 ports on the 2018 MacBook Pros are the same speed now.

Whoo! Do I remember the fuss surrounding that one when Apple disclosed that the right side ports were slower on the 2016 13" MacBook Pro...

 

[EugW]

I thought the original brouhaha was blown out of proportion. Most people didn't have more than 2 TB peripherals anyway. But now that is solved too. Too bad the highest spec MBPs now throttle hard.

2018 13-Inch MacBook Pro With Touch Bar Has Four Full-Speed Thunderbolt 3 Ports

Apple has confirmed that the new 13-inch MacBook Pro with Touch Bar is equipped with four full-speed Thunderbolt 3 ports.

2018 models ship with Intel's eighth-generation Core i5 or Core i7 processors, which both support up to 16 PCI Express lanes, providing enough bandwidth for maximum data transfer speeds up to 40Gb/s on all four Thunderbolt 3 ports.

 

[curmudgeonette]

Both of the data rates calculated come in very close to or exceed the 10Gbps of USB 3.1 Type 2, so I am having trouble reconciling how this is being done and providing the full 5.0Gbps bandwidth for USB 3.0 devices.

The USB-C "alternate" modes carry the alternate signal in lieu of USB. (Whereas Thunderbolt carries the payload within its own protocol.) Therefore, when USB-C is used to carry DisplayPort, the signals follow DisplayPort rules. When it carries Thunderbolt, then Thunderbolt rules apply.

The USB-C connector has four differential wire pairs, only half of which are needed for USB3. All four used together are just perfect for Thunderbolt or full width DisplayPort. For a less demanding display, the DisplayPort portion can be limited to two pairs, leaving the other two pairs for USB3.

 

[Zdigital2015]

The USB-C "alternate" modes carry the alternate signal in lieu of USB. (Whereas Thunderbolt carries the payload within its own protocol.) Therefore, when USB-C is used to carry DisplayPort, the signals follow DisplayPort rules. When it carries Thunderbolt, then Thunderbolt rules apply.

The USB-C connector has four differential wire pairs, only half of which are needed for USB3. All four used together are just perfect for Thunderbolt or full width DisplayPort. For a less demanding display, the DisplayPort portion can be limited to two pairs, leaving the other two pairs for USB3.

Thank you for clarifying that for me...sometimes I tend to be a bit rigid in how I think these protocols work. It may come from all those years assigning, checking and double-checking SCSI IDs on device chains to make sure things ran smoothly in my previous life. I did break a few SCSI terminator rules in my time, though.

 

[DeepIn2U]

Waiting for Apple to put A12 into this... A11 is already more powerful than these chips 8(

Apples to Oranges. Too many variables to compare running various office, pro, and leisure apps above OSX and its shells and libraries to say the A12 is ‘powerful enough’!

What an iPhone or iPad can do vs a Mac is very very different.

 

[EugW]

Apples to Oranges. Too many variables to compare running various office, pro, and leisure apps above OSX and its shells and libraries to say the A12 is ‘powerful enough’!

What an iPhone or iPad can do vs a Mac is very very different.

Well, MS Office performance is just fine on A8X, and A12 will probably be around three times as fast as A8X multi-core.

The main problem with most business apps these days isn't the performance. It's the interface. Business computing and office productivity apps often suck on the iPad, purely because of the interface.

 

[DeepIn2U]

Well, MS Office performance is just fine on A8X, and A12 will probably be around three times as fast as A8X multi-core.

The main problem with most business apps these days isn't the performance. It's the interface. Business computing and office productivity apps often suck on the iPad, purely because of the interface.

You’re talking about MS Office created for a mobile solution and use case. Not the full blown Office sweet complete with command line switches, support for add-ins like Tableau or Bain has created for years or like McKinsey & Co’s Own add-ins. You’re not talking about OneNote that gets full API support on the desktop which it cannot on mobile. Please don’t ever mix the two up and say performance is great or feature set or code is the same. It is nowhere close!

The interface is a reflection of the platform designed for and based on what features can be done and laid out resonably. This is the difference between a mobile centric app vs a desktop. Also why for many older generation users that iOS cannot and will not replace a Mac.

When ALL features of the desktop can be implemented and support by bare metal or processor design of the A series ... then we’ll talk about why a UI for particular platforms are more robust and easier laid out or less robust and a magnitude times more configurable with more features and more useful/potential to a wider end user appeal.

Trust me I’m all for A series to replace Intels monopoly, Microsoft has been killing it here in work done vs Apple. But we want quality and lasting robust performance to rely on while having apps supported without being what Microsoft is calling “tuned” or working with intel like they did for Windows 10 S for their latest Surface Mini.

 

[EugW]

You’re talking about MS Office created for a mobile solution and use case. Not the full blown Office sweet complete with command line switches, support for add-ins like Tableau or Bain has created for years or like McKinsey & Co’s Own add-ins. You’re not talking about OneNote that gets full API support on the desktop which it cannot on mobile. Please don’t ever mix the two up and say performance is great or feature set or code is the same. It is nowhere close!

The interface is a reflection of the platform designed for and based on what features can be done and laid out resonably. This is the difference between a mobile centric app vs a desktop. Also why for many older generation users that iOS cannot and will not replace a Mac.

When ALL features of the desktop can be implemented and support by bare metal or processor design of the A series ... then we’ll talk about why a UI for particular platforms are more robust and easier laid out or less robust and a magnitude times more configurable with more features and more useful/potential to a wider end user appeal.

Trust me I’m all for A series to replace Intels monopoly, Microsoft has been killing it here in work done vs Apple. But we want quality and lasting robust performance to rely on while having apps supported without being what Microsoft is calling “tuned” or working with intel like they did for Windows 10 S for their latest Surface Mini.

It sounds like you’re suggesting you need to port over everything first including third party applications for full support before considering ARM on Mac. If that’s what you’re suggesting, I’d say you’re barking up the wrong tree.

What Apple needs to do this is OS support and limited support from key apps. Then they go from there.

 

[DeepIn2U]

It sounds like you’re suggesting you need to port over everything first including third party applications for full support before considering ARM on Mac. If that’s what you’re suggesting, I’d say you’re barking up the wrong tree.

What Apple needs to do this is OS support and limited support from key apps. Then they go from there.

We saw what happened to other smartphone platforms when they tried exactly that: PalmOS (Pre phone), BB10 ... Apple buried these with the 3G and App Store. The initial app offerings where very duplicated but the platform wasn’t he limiting factor.

 

[Falhófnir]

I meant for 2019. If Ice Lake Y ships in 2019, then the MacBook could get Thunderbolt with no external controller chip needed.

BTW, at that time, the MacBooks would be solidly faster than the fastest MacBook Airs too, as the MacBooks would be quad-core.

I'm not sure about the quad core part, upping the core count on the U and H chips has destroyed the thermals which on a passively cooled chip is a really, really big deal. 10nm would have to produce significantly cooler running cores to allow two more without throttling down to such an extent you start going backwards on performance. I believe where the MacBook falls down compared to the air is the HD615 graphics which is still quite anaemic compared to the HD6000, I guess the more severe TDP constraint limits this.

 

[sublunar]

I'm not sure about the quad core part, upping the core count on the U and H chips has destroyed the thermals which on a passively cooled chip is a really, really big deal. 10nm would have to produce significantly cooler running cores to allow two more without throttling down to such an extent you start going backwards on performance. I believe where the MacBook falls down compared to the air is the HD615 graphics which is still quite anaemic compared to the HD6000, I guess the more severe TDP constraint limits this.

Apple ARM CPUs are known to have 6 cores already (in the case of the A11). Who knows what tricks the A12 has up its sleeve.

iPad Pros with A12X CPUs would be a very interesting comparison with MacBooks this year.

Back to Intel though. It would be logical for them to have 2 decent cores on the 5w CPUs and the recent leak from the Dell Chile site as reported on Tom's Hardware on this leak seems to corroborate the theory.

Lets see where the core vs heat profile issues goes.

 

[Falhófnir]

Apple ARM CPUs which are known to have 6 cores already (in the case of the A11). Who knows what tricks the A12 has up its sleeve.

iPad Pros with A12X CPUs would be a very interesting comparison with MacBooks this year.

Back to Intel though. It would be logical for them to have 2 decent cores on the 5w CPUs and the recent leak from the Dell Chile site as reported on Tom's Hardware on this leak seems to corroborate the theory.

Lets see where the core vs heat profile issues goes.

It will certainly be interesting to see, Apple will probably be able to tailor any Mac CPU they were to design to idle it's cores at a very low power state when not needed, and I believe ARM cores tend to be much better at this than intel's anyway? the big.LITTLE design also gives ARM an advantage over Intel's co-equal core designs in this regard.

I wasn't all that surprised Intel has run into issues just adding two cores without also shrinking to 10nm to reap the benefits of a smaller, cooler node, but I was surprised by how much of an issue it has been. I thought they might have downclocked the individual cores a bit more, or else limited the simultaneous turbo clock speeds of the cores more aggressively to compensate.

Considering the i9 can't even maintain it's base clock in the MacBook pro, it makes you wonder whether intel should really be offering this as a mobile 45W chip, it seems almost like you're fitting a 65W desktop chip into a laptop form factor, and when gaming laptops have done this, they have been, thick, heavy beasts to accommodate the desktop grade cooling required. But then I don't think its that much of an issue considering that even with throttling, it's still more performant than the i7, and ultimately that's what peopler paying for - it's no different than paying for a higher clocked chip on previous models really.

 

[EugW]

I'm not sure about the quad core part, upping the core count on the U and H chips has destroyed the thermals which on a passively cooled chip is a really, really big deal. 10nm would have to produce significantly cooler running cores to allow two more without throttling down to such an extent you start going backwards on performance. I believe where the MacBook falls down compared to the air is the HD615 graphics which is still quite anaemic compared to the HD6000, I guess the more severe TDP constraint limits this.

We already know from leaks that ICL-Y has 4 cores.

Ice Lake Y is still alive... Here is a reference to it right on Intel's website:

https://www.intel.ca/content/www/ca...one&query=BGA1377-ICL-Y&keyword=BGA1377-ICL-Y

This references the ICL-Y 5.2 W quad-core part that was leaked months ago.

Don't expect this until late 2019 at the earliest though. BTW, this could theoretically support a 32 GB 12" MacBook.

The Intel reference has since been removed though.

 

[Falhófnir]

We already know from leaks that ICL-Y has 4 cores.

The Intel reference has since been removed though.

Intel might plan for this to be the case, but I'd say everything is a bit up in the air at the moment, 10nm is still very broken by the sounds of things, calling whether it will be launching in earnest next year into serious question. 10nm would have to be something special indeed to make an extra two cores feasible in a passively cooled computer. There again, 14nm allowed Microsoft to fit a U series in a surface pro without a fan, and I highly doubt that would have been possible with a haswell chip.