It looks like they've ditched the m5 and m7 entirely, since they didn't offer any real performance advantages over the m3. Now it seems we're getting a fanless m3 solely, with the next jump up being to the higher powered i5s and i7s.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Intel Announces First Low-Power 'Kaby Lake' Mobile Processors
- Thread starter MacRumors
- Start date
- Sort by reaction score
It looks like they've ditched the m5 and m7 entirely, since they didn't offer any real performance advantages over the m3. Now it seems we're getting a fanless m3 solely, with the next jump up being to the higher powered i5s and i7s.
Apple is always 2nd place when it comes to cutting edge tech, they wait till the cows come home to plug a cip in the mbp's.
But oh dear, when they do !
(next one is out)
Again, i feel Intel is just wasting our time with these x% designs when they should really go to "xxx" because they can and they have the tech.
But oh dear, when they do !
(next one is out)Again, i feel Intel is just wasting our time with these x% designs when they should really go to "xxx" because they can and they have the tech.
Right, we should embrace paying $2,000 for old hardware, got it.
First, Skylake was way late
Second, just the regular desktop chips were released an year ago.
Third, Skylake core m processors were released Q2/16.
Fourth, Iris Pro 580 based mobile chips were supposed to be launched at Q1/16, but first samples came out on May. And they're still low volume. Available only in Intel Nuc.
Fifth, Xeon EP Skylakes are coming next year.
So there. Skylake is a family of processors. What you said is a common mistake.
Last edited:
Since it's a same die sized tock upgrade to the Skylake architecture, and (I believe) it's using the same GPU series, I doubt there'll be any huge differences between the two. What you can expect is a slightly speedier, slightly more efficient version of what's already out.
Of course they offered performance advantages, that is very clear from benchmarks. Also, the i5 and i7 still carry a 4.5W TDP, which means they are not more high powered, nor destined to go with a fan.
The A Series is not going to all of a sudden stop improving. The A series already beats early x86 chips. The X86 kept improving, what makes you think the A series won't? Especially when intel is telegraphing that the x86 is not as important as it once was and further development is not a priority.
Apple has a lot invested in the A series and can't be bothered at this time to be at the whims of Intel when their entire product line depends on the A series.
It's a forgone conclusion, but people don't want to accept it. Fine. Burry your head in the sand.
Really? I remember reading that they had a higher max clockrate, but it only gave you slight advantages in burst speed, since m chips are designed so tightly around a specific thermal envelope. In practice, there wasn't much difference between the three.
Kaby Lake doesn't include HDMI 2.0, sorry. You're going to have to wait another year (Cannonlake) for that as well as for native Thunderbolt 3.
Go read the Macbook forum - It is quite clear from both benchmarks and real life experience, that the m5 and m7 had an advantage over the m3.
Imagine if Apple keep the MBA as a lame duck machine just to keep the entry level price a $999. Imagine the "new" MBA with the same chassis, screen, same 8GB/128GB memory/SSD just with the new Kaby Lake CPU (with weakened GT2 graphics) and with USB-C... I can't see Apple splashing the cash on developing a new MBA when the rMB is clearly destined to be the MBA replacement within a few years.
Will do!
...though I didn't even know the MacBook came with an m5 or an m7. Though it was, like just about everything else, m3 exclusive. Guess I gotta get my learn on.
No more waiting, they just need to release Skylake MBPs now, they have been delayed forever. Apple probably had Skylake processors in their labs in the beginning of 2015. If they want to release an updated MBP with Kaby Lake at WWDC 2017, that OK. I can live with a delay of 6-9 months after an Intel release, but not two years.
Last edited:
At being a SoC for a phone or small tablet. The A series is going to improve along the same lines it has improved.... as being a phone chip. Geekbench scores are not a solution. They are just a relatively narrow metric.
An x86 core all by itself doesn't make for a solution for classic PC form factor computers. Pragmatically, there is a single USB 2.0 port on an iOS device. That isn't going to fly on a laptop or desktop. Bandwidth for discrete graphics? No. More than one display? No. Thunderbolt? No. 3rd party USB 3.0 chips? No.
So if the A series is behind now and they both keep improving at what point is the A Series going to catch up?
Apple has about 1 ( maybe 1.5) world class design teams. Intel has 3-4x that number. Apple could spin up another one to create something that wasn't primarily a phone chip but won't. Three major reasons.
1. Apple is Scrooge McDuck. It is expensive and there is little to no reward here in terms of scale. Apple sells about 10x as many A series chips as they buy chips from Intel. It is cheaper for Apple just to buy them from Intel or AMD. Way cheaper.
2. The Mac sales are showing very little group. So no only are Mac not at iOS device scale they are show no signs that they are going to be in the future either.
3. If macOS primary market competitor is on x86 ( Windows ) and Windows is largely x86 only, they are not loosing much competitive wise. Using the some foundation as Windows allows Apple to spread the R&D costs over the whole x86 market which is 9x bigger than the Mac market. ( go back to point number #1 of Apple being Scrooge McDuck. )
Priority and significant source of growth are two different things.
Because the iPhone market more than pays for that investment. It is billions of dollars bigger market than the Mac market.
ARM ( A Series) solutions don't scale across the same performance market as the Mac sit in. You might be able to get a A Series solution into a MacBook , MBA , hobbled iMac (with MBA processor), but are no ARM (or A Series) that preform in the iMac , Mac Pro zone.
Apple has shifted processor when they can do the whole Mac line up. That was true in 68000 to PowerPC switch. Was true in the one year transition of whole line up from PPC to x86. If going off the x86 going to need the same solution. There is no ARM solution for that now. And Apple can't pull one out of the hat on the cheap.[/quote]
Last edited:
Pent up demand for new rMBPs at 4 year universities will be off the charts. What will students do?
https://web.stanford.edu/group/bookstore/SUprices/macintoshall.html
https://web.stanford.edu/group/bookstore/SUprices/macintoshall.html
I still dont understand why non-15 inch Macbook pros and Airs get Dual core I5s and I7s...
Before I bought my 13 inch, I had no Idea that dual core I7s exist...
Before I bought my 13 inch, I had no Idea that dual core I7s exist...
Yep...this. We already pay a premium when buying a Mac. If I want old tech I can go get an HP and save some money.
It's pure fiction. It's right up there with Nvidia Pascal's 10x compute performance, which they have since retracted heavily. Yes, I'm aware of OpenCL compute and other GPGPU work. Intel's OpenCL stack has nothing to do with OS X. Their claims are pure fiction on performance claims.
They pick an algorithm that increased 10x over prior capacity [mainly due to gutting their software stack] and now using a vector operation inside Kaby Lake [that already exists in prior incarnations] proclaim 10x performance improvement. It's garbage.
These will make it into Apple's product line no later than Q2 2019.
Include HDMI 2.0 native? No. But it is relatively inexpensive to put a DP 1.2 to HDMI 2.0 converter. Especially at the prices Apple charges for their systems.
"...One of the disappointing aspects from Skylake that has still not been addressed in Kaby Lake-U/Y is the absence of a native HDMI 2.0 port with HDCP 2.2 support. Intel has been advocating the addition of an LSPCon (Level Shifter - Protocol Converter) in the DP 1.2 path. ... "
http://www.anandtech.com/show/10610...six-notebook-skus-desktop-coming-in-january/3
Native TB v3? It is doubtful that comes widespread with Cannonlake either. TB controllers have length restrictions on how far away they can be from the physical TB port. It is far more flexible to use a discrete chip. Going to need discrete chips in the peripherals anyway also ( so still need to make as discrete product).
Perhaps in a subset of the Y series that is designed for physically small devices. If the central SoC is only few inches away from the ports of the device than integrated may work better.
More likely is that those chipsets and/or CPU package PCI-e v3 bandwidth will be better at handling TB v3 and faster PCI-e SSD concurrently. That is not at stage where every chip sold has one. TB is no where near where it is a "must have" for every system.
[doublepost=1472585652][/doublepost]
fiction is what you are peddling.
Look at the coverage of the media block here.
http://www.anandtech.com/show/10610...six-notebook-skus-desktop-coming-in-january/3
There is no "vector operation" . And yes it is for specific algorithm. That is what Fixed Function circuits do! There are a set of transistors that only do these specific algorithms ( HEVC and VP9 ).
Last edited:
The immersive internet? Huh...
Sure intel make great chips but they need to employ some humans to sell the stuff.
Sure intel make great chips but they need to employ some humans to sell the stuff.
Kaby Lake has been round for a long time now. Intel started shipping production processors to manufactures in July. So at least for Mac Book Air and possibly Mac Book updates are possible in the near future. Updates to the MBP are another issue and frankly there may be a surprise there in November. If not MBPs get SkyLake. However I really don't think Apple wants to put SkyLake in a MBP, especially a new revision, these are apparently heavily re-engineered platforms pushing boundaries that will apparently require Kaby Lake features. I wouldn't be surprised to see new MBA in September.