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.
Apple's 'M2' Next-Gen Mac Chip Enters Mass Production, Expected to Debut in Redesigned MacBooks Later This Year
- Thread starter MacRumors
- Start date
- Sort by reaction score
The speeds these things swap to ssd is ridiculously fast, making the RAM limitations less of a concern unless you use a single application that needs that much RAM at once. Unfortunately upgrades other than adding external storage are basically a thing of the past. These days you upgrade by selling a used machine and buying a new one. My last few MBPs have depreciated about a $1 a day, which isn’t too bad.
Last edited:
Yeah, the value is going to tank! you should have waited for the m2 beast
You bet… I anticipate a streamlined M1 macbook… probably with colors like the iMac before long.
However, I expect a "tweaked" M1 that can support up to 32 GB of RAM. The M2 coming late this year will get 16, 32, and 64 GB RAM support.
Could be.
Apple already charges $200 for $25.60 worth of lpddr4x (spotmarket price, Apple pays less).
This highway robbery might have been acceptable when you could add your own RAM and SSD, because then this only affected the affluent-enough-not-to-care or organisations. But now everyone needs to factor in those costs in their purchases, and the configurations you actually want are way less competitive than the minimum config systems whose prices are typically bandied about.
I’ve used Macs since 1985, but it’s always been a preference, not a need. Nobody likes to feel that they are being manipulated and then squeezed like an orange for every last drop of profit. And when you know what RAM and flash memory component costs actually are…
Any predictions on what kind of binning we'll see with the M2? Right now they're only binning based on 7 vs. 8 functional GPU cores.
With more GPU cores, could we see more levels of binning based on number of cores? If they're pushing higher GPU clock speeds, might they also end up binning by that?
And with more CPU cores, and possibly higher clock speeds, could we see CPU binning by number of functional cores and/or clock speed?
No idea. I am surprised there hasn’t already been binning on clock speed (though perhaps there has been! Who says iPad Pro M1’s can hit the same max clock as iMacs?)
I was wondering if there might be an architectural reason they couldn't bin by number of CPU cores. I.e., could the M-series chips be integrated in such a way that they couldn't work with fewer than the designed number of CPU cores? [By contrast with, say, Intel i-series chips, where a 4-core i5 with one or two bad cores can be turned into a 2-core i3.]
That would seem very unlikely. Turning off cores is easy, and the circuitry required is pretty trivial.
I think thats a good guess, as too binning by CPU cores, the question is how many do fail to have the full number?
And with the M1 loosing a "big core" would mean quite a performance hit that would be felt much more than -1 GPU.
In the end it is also a question of logistics and keeping the store clean cos otherwise they end up with 1000s of possible configs confusing the costumer (which should be cheaper, broken CPU or GPU).
So it was most likely that there are enough broken GPUs to make that a separate product.
I wouldn't call it manipulation. For instance, here's two alternatives:
1) Right now, Apple charges $1300 for the base 13" M1, +$200 to upgrade RAM to 16 GB, and +$200 to upgrade storage to 512 GB.
2) Alternately, Apple charges $1600 for the base model, and +$50 for each of the RAM and storage upgrades.
Is Apple "better behaved" if it does #2 instead of #1? For me at least, the bottom-line issue is the overall expense of their products, which can sometimes be substantial, not the details of #1 vs. #2.
My guess is the reason that Apple chose #1 over #2 is they want to increase their customer base, and thus decided to make the base model more affordable. I.e., essentially the more expensive models are "subsidizing" the less expensive ones. How progressive!
Last edited:
I disagree. The A series will be the main platform Apple develops for. iPhones dwarf Laptops and Desktops now. An M1 is a bad fit for iPhones.
The M1 has drastically different IO requirements than A series do, the minor increase in power requirements alone would kill an iPhone. The A series will be the fastest cycle and get the newest hardware features because iOS is Apple's premier platform and the requirements of the A chips are much simpler. Once new designs shake out in A series, they'll be "scaled up/scaled out" a year later for the M series chips.
Axx == Mx with fewer cores and slight differences, so unless Apple plans to release a Mac with a TDP <5W the chip for the iPhone will always have to be a separate product and it will continue to be labeled "Axx".
Reasons why the IPP got the M1 and not an A14x are, a) marketing and b) too low expected sales numbers to do a specific chips with just minor differences to the one already used in Macs.
Reasons why the IPP got the M1 and not an A14x are, a) marketing and b) too low expected sales numbers to do a specific chips with just minor differences to the one already used in Macs.
Obviously it’s a marketing trade off. But as I wrote, 8GB lpddr4x is (less than) $25, and 256GB of TLC flash is (less than) $20. And I’m giving Apple the benefit of assuming that Apple uses the more expensive TLC variant of NAND.
In the context of this discussion about a higher end SoC it makes sense to assume that those who are interested in such a SoC correlates well with those who would like more of both RAM and storage. In which case Apples pricing of these resources becomes even more unfortunate.
All IMHO of course. I’m not a share holder.
I hope Apple fixes the bugs in the display output.
www.eizoglobal.com
Apple M series Chip Equipped Mac Computer and Monitor Compatibility | EIZO
EIZO is a visual technology company that manufactures high-end monitors and display solutions for business, graphics, home entertainment, healthcare, maritime, security/surveillance, ATC and defense/aerospace.
Let's consider the ~$20/256GB you mentioned for TLC flash. Is that an OEM part price? If so, there's a multiplier you need to apply in going from part prices to retail prices. E.g., I recall that for audio equipment the rule of thumb was 5x: every $1 increase in parts costs equated to a $5 increase in retail cost. That's not the manufacturers ripping consumers off, that's them accounting for all the other costs that go with bringing a more expensive product to the market.
I don't know what the multiplier is for laptops, but it seems like the industry standard might be in that range as well. E.g., Dell charges $100 to increase the SSD size on the XPS 13 from 256 GB to 512 GB which, using your $20/256GB figure, is a 5x multiplier.
Thus the more accurate way to represent things is that Apple is charging the consumer an extra $200 for something that should actually cost the consumer an extra $100. By contrast, you're presenting it as Apple charging the consumer an extra $200 for something that should only cost consumer an extra $20, which (at least based on my limited understanding of manufacturing/research/distribution etc. costs) I don't think is accurate.
Also, you wrote "And I’m giving Apple the benefit of assuming that Apple uses the more expensive TLC variant of NAND." But according to https://www.mydigitaldiscount.com/everything-you-need-to-know-about-slc-mlc-and-tlc-nand-flash.html "TLC flash is the cheapest form of flash to manufacture."
Further, you seem to be writing as if flash is simply a commodity. But, IIRC, Apple's laptop SSD's have been significantly faster than those of other laptop mfrs., so maybe their SSD components are higher-end and thus should cost more.
Last edited:
Flash IS a commodity! The performance aspect is (mostly) organisation/controller/SLC or DRAM cache.
Look, even in these days of inflated component prices you can buy 1TByte of m.2 (PCI-e attached) for $100 or so, in an individual device sold to end users. How much do you think Apple pays through their supply chain, and seing as they use their own controllers and don’t even put it on a small PCB with a connector?
That Apple has huge, roughly order of magnitude margins on their RAM and NAND upgrades isn’t really debateable. The question is what the consequences of that is for them, their brand image and their customers.
You could argue that this allows them to post quite reasonable prices for their bare bones models, and thus gains them low end (for Apple) market share. This seems quite reasonable.
However, this also means that a lot of customers stay at that minimum configuration, some of whom will soon fill up their secondary storage with pictures and videos from their phone, which is not a particularly premium experience. (Also, the limited RAM will cause increased swapping to that quickly shrinking free area of SSD.)
The high prices for non-user accessible upgrades is obvious to reviewers and end users alike and, like their $999 monitor stand, creates the impression that Apple simply fleeces their customers. And this of course, has some truth to it.
But the practise becomes even more problematic for users that do things on their computers that either simply benefit from or require more RAM, or that generates a bit more data or they may already have a good bit of data that is migrated from their previous system. When we are talking about more performant SoCs I think it is quite reasonable to assume that many of us who are interested in those fall into one or more of these categories. So how is Apple going to deal with users that would like say, 32GB of RAM or more and/or 2TB or more of secondary storage? If Apple charges the same upgrade prices, the corresponding cost will become exorbitant and their competitiveness will obviously drop correspondingly. Which is why it is an interesting question in this context.
And of course, if you are interested in the market and competitive angle, the answer will clearly have consequences.
Possible… Apple isn't always particularly logical in their naming.
But core count wouldn't be the only difference, I imagine. A Mac Pro's chip needs to address external memory in a way an Air's doesn't, because it's not really practical to put 1.5 TiB RAM on the package.
Question is wether it is "practical" to have 1.5TiB at all....
a) good chance the CheeseGrater_Mk2 will stay for a bit longer for those that really need something like that
b) Flash based storage is getting faster and faster and you con run them in "parallel" (if your chip has enough I/O) to the point that it might be just as fast as having multiple RAM chips/modules hanging on a bus several inches long.
The rest is just SW, as in better paging better APIs and better compilers making sure SW gets back to being smart about RAM use.