If I'm not mistaken, you should be even more mad at Intel. Skylake is 'delayed' not because it wasn't ready but because Intel has a warehouse full of the current chip. They delayed Skylake to sell off their over supply. At least I think that's how it went down.
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 Unveils Full Lineup of Skylake Processors for Notebooks and Desktops, Early 2016 Likely for Most Macs
- Thread starter MacRumors
- Start date
- Sort by reaction score
If I'm not mistaken, you should be even more mad at Intel. Skylake is 'delayed' not because it wasn't ready but because Intel has a warehouse full of the current chip. They delayed Skylake to sell off their over supply. At least I think that's how it went down.
Last year's iPhone is the 6 and 6 plus. It's the basis of this year's model. Regardless, they are an evolution of the 5/5s. There's a significant chance that the people who want a smaller form factor will be able to get it before too long. Its not as if there are separate teams working on them.
Forgive me, but unless you have some inside knowledge, I don't think your proclamation that any product exists because an assembly line exists means much. And your prediction that it will be made for more years but will be basically unchanged is pretty close to no chance. They'll update it or they'll kill it.
Most of the same development that exists in a Macbook is and will be used in other laptops and desktops for that matter.
High end laptops require a high performance processor. You're saying ARM doesn't make a high end chip because they're power hogs, and Intel doesn't make a low end chip because X86 can't match ARMs low power usage.
That's where you're wrong.
Take a look at Core M vs the A8x. Power usage wise, they're fairly equal. The Core M chip is rated at 4.5w TDP, the A8x isn't published, but from what I'm reading it is estimated at 4.5w TDP as well. Comparing geekbench scores, they're fairly close in multicore score and the Core M beats it in single threaded performance. For regular usage, the Core M is faster.
Like I said, it didn't make sense from the beginning.
I believe next Macbook Pro Retina is going to be released somewhere between november 2015 and march 2016. So sad apple couldnt upgrade the processors in this latest release mid 2015 given that intel only gave slightly faster processors...
Pity or not...Ive just purchased a 15.4 macbook pro 2.5ghz -mid 2015- 2 weeks ago since ive sold my 7 year old Mac Pro (Early 2008) and i didnt have a computer powerful enough that allowed me to edit 4k videos and make motion graphics and put more money in upgrades... Sold it before it was too late... Anyway for me this Macbook is going to be a big speed bump... As is a 2014-2015 computer
Lets see what happens in the following months. I believe too that apple is going to upgrade macbook pro design case maybe with colors and with the new keyboard that the new macbook 12 has...
Pity or not...Ive just purchased a 15.4 macbook pro 2.5ghz -mid 2015- 2 weeks ago since ive sold my 7 year old Mac Pro (Early 2008) and i didnt have a computer powerful enough that allowed me to edit 4k videos and make motion graphics and put more money in upgrades... Sold it before it was too late... Anyway for me this Macbook is going to be a big speed bump... As is a 2014-2015 computer
Lets see what happens in the following months. I believe too that apple is going to upgrade macbook pro design case maybe with colors and with the new keyboard that the new macbook 12 has...
Last edited:
If the Air is not updated with retina and slowly phased out, there could be room for a larger 13.3 or 14" retina MacBook with 15 W chips down he line. There really needs to be something between the ultraportable 12" 4.5 W MacBook and the Pros.
You are comparing benchmarks across architectures - which always has to be taken with a grain of salt, but you are also taking benchmarks across operating systems, across different power envelopes (operating system will not necessarily run a processor flat out), etc. etc. etc. There are a hell of a lot of caveats.... Even then the difference (which can be much greater than your the error that you introduced by doing all that) is only in the order of of 5% difference on single core. Then you have the fact that the operating system will typically do it's stuff on a different core than your application that is running flat out burning that 5% on one CPU. The one thing that is probably a big factor is even IF they are comparable - the Core M is almost $300 / chip tray price...... You could not put it in an iPad without raising it's price by a couple hundred dollars. So if you want to pay significantly more for 5% single core performance and lower multi-core performance - then Core M is a wise choice.
ARM is doing exactly that, but you can't see it because you think "complex" applications need single core performance but that's just being a victim of Intel's marketing and not know much about the technical details.
So would you call Google's search engine a "complex application"? It certainly needs performance, right? But it doesn't reply on single core performance, it uses thousands of cores.
How do you make your "complex" application run 5% faster? Get a Skylake processor! How do you make it run 1000 times faster? Oops! Intel can't help. But it can be done using many cores.
The "complex" applications you're talking about are just poorly written or legacy applications. Modern software isn't written that way anymore.
Why would ARM try to follow the same losing strategy Intel is following? Right now it's a software problem, not a hardware problem. ARM is doing the right thing in pushing power usage down while increasing performance. It's not going to make the same mistake that Intel is making, clinging to the legacy desktop and server application markets which sell less and less each year.
Meanwhile the latest mobile tech is using 8 and 10 cores and soon 16 cores. And they run off tiny batteries. And Intel has 0% market share.
Oh dear.
So after going with all the wait, do you recommend to buy the MacBook Pro with Core i7 4980HQ?
I've been waiting for 6 months now...and we didn't get the 5th Gen and now Apple may skip it to the 6th Gen!
Your opinion?
I've been waiting for 6 months now...and we didn't get the 5th Gen and now Apple may skip it to the 6th Gen!
Your opinion?
Windows already runs on ARMs, but it's a question of how fast the market moves. Microsoft has already been caught with its pants down, more than once, so it may respond rapidly with strong ARM support if Apple makes its move.
The LLVM move sounds like an emerging strategy. We may see ARM laptops from Apple sooner than we think!
Last edited:
That's what Intel has fooled you into thinking, because they sell processors with single thread performance and Windows is laden with legacy software.
Compare the Android OS which mandates aggressive threading and already makes good use of 8 or more cores. That's where things are headed.
Also Intel's mobile processors are not faster than ARMs within the same power envelope, they're slower. You're being blinded by Intel's tricks (single core turbo, complex multi-issue pipelines). Apple's A9s are not regular ARMs and make design choices that trade off power usage for single core performance. Most high end ARMs are 8 core these days. A9s are dual core.
Why do they do that? Because they use the same basic OS architecture on Intel and ARM, so they can't break free of the single thread performance addiction like Android has done.
Skylake is late because Broadwell was late or more specifically because the transition to 14 nm was more difficult than planned. Their resources were focussed on bringing out Broadwell, at least until a certain point in time. Those people working on getting Broadwell out were not available to work on Skylake, pure and simple.
Of course, committing those resources (including production lines) to Broadwell (and withholding them from Skylake) at some point might have been motivated by the desire to sell (some) Broadwell chips to people who otherwise would have bought Skylake chips and to keep some of the good stuff (aka Skylake) for later to keep a constant stream of people upgrading over the next couple of years instead of ending up in a situation where they don't have anything new for two years or so.
But very little third-party software for it appeared so far (and given that Microsoft switched back to Intel with the non-pro Surface models isn't exactly accelerating this).
Windows crippled ran on ARMs and they more or less discontinued it. The problem is they have no strategy of implementing an App store when a developer submits an app, and they upload the bitcode and the customer does not have to worry if they have an ARM or Intel... they just install the application. Right now even if you have an ARM Windows - you might have no applications because of the lack of strategy (which then becomes chickened and egged to death).
I know, thus my comment about proper support. There is the compatibility issue, but Apple may have a way of dealing with that. It may be as a part of a transition to a new OS platform. Some people may not remember the transition from OS 9 to OS 10, which took a few years as OS X matured. It could be something similar where the benefits are obvious but people move over when the software they use is available.
I think if Apple can offer tangible benefits (great battery life) and a new interesting OS design, it could work, just like the move to OS X was very successful in the end.
Not at all. It's two different product lines. The Macbook while thin doesn't take on a lot of what the Air has, and vice versa. The number of available ports is 1 example.
The Macbook also has a more crammed-appearing keyboard layout compared to the other MacBook Air/Pro models.
Yeah Microsoft are not good at software strategy. They seem to screw up again and again. But if Apple pulls off a transition (see my post above for a theory) and then has an ARM platform and OS that delivers extended battery life and performance with an array of apps, then Microsoft will have to do something, and Intel would be in a panic too. Would be interesting to watch.
Well, that is a new criticism.... not functionality but now it is the crammed-appearing keyboard. I don't own one, but I have played with it and I actually liked the new feel of the mechanism.... of course the key fall takes a little getting use to.... but then so did all the keyboards that came after my IBM PC from 1981.... now that was a keyboard.... I could even kill someone with it quite easily by whacking a bad developer with it.... oh those were the days when you had real keyboards....
95% of the people do not need any more than one port if they are using it as a laptop/portable. At home - yes people like plugging more stuff in, but docks fill that void..... If I am taking something between two locations (home/office) then I already usually just get a power adapter for both locations to lighten the regular load. If the laptop can keep a full days charge, it already lightens the load. As someone that came off the road after 3 years with one shoulder pressed down an inch or more from the weight of laptops when they were heavy.... I can greatly appreciate the lightness and compactness of the Macbook..... sometimes I even had to carry two around - one for work (consulting company) and one for the customer that I was dedicated to.
The macbook CPU is sufficient for coding, for document editing, for spreadsheets, powerpoint/keynote, for touching up photos - still image editing, playing movies, music, conferencing, file transfer, browsing the internet, irc, messaging, remote desktop applications, charting, sqldeveloper, etc.
What the macbook is not really designed for is video editing (other than a quick touchup), 4K video editing especially, running vmware with multiple virtual machines - running Linux and Oracle.... and hardcore gaming.... but then the Macbook Air is not really meant for those tasks.
I would however like them to change the USB-C port from just USB to USB/Thunderbolt in the skylake version.... then you can even hook up DAS disk arrays, and external graphics cards within docks etc.
I just see a very small niche between the Macbook pro line and the Macbook line - something that could easily be absorbed in one line or another.
Simple, higher performance alternatives, e.g., 15 W chips with faster GPU. By todays standards, the Air is a quite normal laptop with respect to portability.
There should be a retina alternative between 12" ultraportable and the Pros. I am thinking a larger Retina MacBook designed for 15 W chips. Similar to the current 11 and 13" Airs.
The MB against the MBA isn't that huge of a gap according to geekbench.
Single thread:
MB- 2633
MBA 11"-2859 (i5)
MBA 13"- 3210 (i7)
Multi thread:
MB- 5272
MBA 11"- 5724
MBA 13"- 6857 (This is the only big gap, and this is an i7)
http://www.notebookcheck.net/Intel-HD-Graphics-6000.125588.0.html
http://www.notebookcheck.net/Intel-HD-Graphics-5300.125576.0.html
In some benchmarks, the two GPU are even basically equal.
Where is this power gulf?
Last edited:
I would not go that far, the top end of the lines compared there is a 30% gap, the bottom line of each of the lines there is a 25% gap - so on average a 27.5% performance gap on the CPU.
Now when you mix in all the components you really end up with maybe 1/3 of that in real performance gains..... taking all other components into consideration.
Specs aside - since in the real world they don't matter - what matters is what functions the laptops are capable of providing - and I find very little the air can do that the macbook cannot....
I do expect once the supply constraints are ironed out that other models in the macbook line will be released (likely at time of refresh) so maybe a 12 and 14 inch model, one that has less storage (like the low end macbook air) bringing an entry-level pricing down to $100 difference (the cost of the retina upgrade). Maybe even a non-retina macbook at the price of the macbook air.... at that point the macbook air ceases as line... and you go from the macbook to the macbook pro -- if you need more performance.
2/3 of the power while using 1/3 of the power. I was taking that into account when I said it wasn't that huge. And, like you said, there is very little that the MBA can't do. By the way, the 12" MB is about the same size of the 11" MBA while being within 10% or so of the power. That's what I call awesome.
I think you misunderstood my 1/3. 1/3 refers to the fact that when you add a more powerful CPU all other components being equal you will only see a speed boost of 33% of the increase that the CPU benchmarking would indicate. Memory speed, memory size, SSD speed all play a factor. I would not be surprised if in real world performance it works out to somewhere between 80% to 90% of the equivalent Macbook Air. Most programs - you would not notice any difference. As I have said before people tend to go for "power" then have their CPU idling 90% of the time.
It is also the reason why CPU makers have been more focused on multi-core vs the Ghz game - because increasing the speed of the processor gives you less performance boost overall since it is already racing ahead of the RAM - then having to wait a lot to move RAM into cache.
Makes a lot of sense to me. Except the non-retina MacBook. By the time the current entry level MacBook Airs are phased out, retina is probably standard throughout the line. (Thinking of how long the non-retina MacBook Pro is tagging along).
Well for sure I wouldn't take a 30% CPU speed drop in multithreaded performance, considering that what I need is more speed, not less. That is a gulf btw. The real kicker is that the new rMB is around the same performance of the 2011 i7 MBA, so there's no reason to upgrade. Yes the screen is nice, but the keyboard is not and the lack of ports is a complete bust.