Unless you’re willing to deal with local sales and idiot OfferUp low-ballers offering you $500 on your M1Max MBP 16, you’re only leaving $100 or so on the table if you sell it on ebay. The average selling price on a non warranted 2021 MBP 16 M1Max 32/1TB is $1750 on ebay. After you factor in all the ebay and shipping junk fees, you’re getting $1450 which is just a whopping $100 more than what apple is offering for trade in. You also run the risk of a scam as the buyer is always right with ebay and can claim a broken laptop for a forced refund after he mails you back a rock. This has happened to multiple people.
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.
Trying to decide if now is the best time to trade in M1 Max for M3 Max unit
- Thread starter mdhaus72
- Start date
- Sort by reaction score
Yep. I looked at Ebay and a few other options locally and you're right - The difference is not that big and you have to jump through a ton more hoops on top of that. If I knew someone who was interested in a direct private-sale, I could maybe get a bit more...but then I also couldn't price it much higher than what they could also find on the market.
I have a M1 Max Studio. Personally, I'd rather wait for Thunderbolt 5 to arrive with whatever chip supports it at the time. With the cost of storage expansion from Apple we all basically try for the fastest external solution. Thunderbolt 5 will be a life saver for those PCI-E 4.0 drives that are being neglected by Thunderbolt 3 and 4. An upgrade should be determined by need and worthiness in my mind. If there is an M3 Max/Ultra that supports the new Thunderbolt that would make me upgrade. Otherwise, my M1 Max still rocks.
I really don't understand why you would want to trade in (and yes I read it all).
You are simply lusting a new toy and trying to justify if the lost of money is worth it. Its definitely not.
You've essentially lost over 1/2 value of your original purchase and you are hesitant to AGAIN invest more money (therefore throwing more money out the window) for something that will only marginally improve your workflow.
The answer is NO. Its not worth it at all. If you decide to go for it then its your choice but be aware that you are not gaining anything. You've essentially made a poor value choice.
You are simply lusting a new toy and trying to justify if the lost of money is worth it. Its definitely not.
You've essentially lost over 1/2 value of your original purchase and you are hesitant to AGAIN invest more money (therefore throwing more money out the window) for something that will only marginally improve your workflow.
The answer is NO. Its not worth it at all. If you decide to go for it then its your choice but be aware that you are not gaining anything. You've essentially made a poor value choice.
I can see that as viable. Jumping from 5nm to 3nm. it's a great upgrade. But, there needs to be more for me.
Echoing what magbarn said: I recently checked that with my own M1 Pro MPB. Apple offers $1070 for a 16" M1 MBP w/ 16 GB/1TB, and the average sellling price on eBay is $1,350, which works out to $1,170 after their 13.5% fee (not including what you'll have to pay for shipping, plus the hassle).
OTOH, the equation is entirely different for my 2019 iMac (partly because I've upgraded the SSD and RAM, and Apple's trade-in value doesn't include the upgrades I've done; indeed, I'd need to restore it to the original config before trading in to Apple, which would be a pain for the SSD). Selling price on eBay for my current config should be about $1,400 (and I could probably get more, b/c mine still has AC+), while Apple offers only $495.
I've notice some eBay sellers will only sell to buyers with at least a certain reputation. Does that provide protection against that particular scame?
Of course, you could create a fake reputation, but that's a lot of work for just one scam purchase, and I'd imagine most who engage in the behavior you describe would not do that.
Last edited:
If you don't want to deal with eBay or local selling, you should also see what Best Buy offers (Best Buy is an authorized Apple retailer). Sometimes it's more than Apple, sometimes it's less. The M1 Max isn't listed as a trade-in option on BB's site, so you'd need to bring yours into the store.
And with a high-ticket item like an M3 Max, it's probably worth paying the fee to join BB's preferred customer program to get an additional discount, if such is available.
And with a high-ticket item like an M3 Max, it's probably worth paying the fee to join BB's preferred customer program to get an additional discount, if such is available.
I’m a serial upgrader just because I can afford it and it’s sort of a hobby for me. In my experience Swappa is a good site for reselling. Fees have increased but are still better than eBay.
If you don’t mind being perpetually one generation behind the latest and greatest you should buy from me. 😄
My general advice for serial upgraders is to buy the lowest spec that meets your actual needs (in terms of RAM, SSD, GPU and CPU cores). They lose less of their value over the course of a year than the upgrades.
If you don’t mind being perpetually one generation behind the latest and greatest you should buy from me. 😄
My general advice for serial upgraders is to buy the lowest spec that meets your actual needs (in terms of RAM, SSD, GPU and CPU cores). They lose less of their value over the course of a year than the upgrades.
That day will come. Face ID that is. The iPhone is a bit thicker than the MacBook lids. They just have to make it fit. Hell, they probably have it and just won't introduce it until it's absolutely necessary instead of it just being wanted. I've read many times of their existing equipment and the show dates they assign them. Many times they have waited years to finally add something they're "testing". The iPhone 15 Pro's titanium design has existed for years. If you checked their patents you'd see my point. It's all a waiting game and Apple rolls the dice.
The iPhone FaceID components are twice as thick and the MBP lid which is why it’s not being used.
Face ID might land in future MacBook Pro — and maybe even the iMac
Face ID might land in future MacBook Pro -- and maybe even the iMac
Apple's introduction of the notch in the MacBook Pro seemed like FaceID was imminent — but it has yet to appear. It's clear, though, that Apple is still thinking about implementing it in the MacBook Pro, in the iMac, and possibly a display too.
appleinsider.com
Good question. I was reading in the forums: https://forums.macrumors.com/threads/m4-chip-generation-speculation-megathread-merged.2393843/ Scroll down to Longplays (suspended) and who knows if they are correct but here is their suggestion, so if that is even close to true, it appears you have close to 2 years so trading now might be a wait and see for another year without too much price degradation, My best guess...
Assuming a 19.5 month refresh cycle
- M1: Q4 2020 5nm
- M2: Q3 2022 5nm
- M3: Q1 2024 3nm (N3)
- M4: Q4 2025 2nm (N2)
- M5: Q2 2027 1.4nm (A14)
- M6: Q4 2028 1.4nm (A14)
- M7: Q3 2030 1nm (A10)
- M8: Q2 2032 0.7nm (A7)
- M9: Q4 2033 0.5nm (A5)
- M10: Q3 2035 0.3nm (A3)
This seems optimistic for a couple of reasons (I'd expect far more cases where two, and maybe even three generations are on a single process node). One is that, while this list is pessimistic about generation timing, I don't think most people expect M4 to be on a 2nm node - I certainly don't. The consensus of opinion I have heard is as early as late 2024, as late as spring 2025, based on the A18 iPhone chip and on the N3E 3nm node. Apple doesn't want to put the M series on an extended refresh, unless they go all the way to a two year refresh cycle, because they want it to follow the iPhone. The iPhone is always released in September, and the iPhone pro models always reveal a new set of cores (we may start seeing the non-pro phones using the same cores, but fewer of them).
The M will probably come out between one and six months AFTER the iPhone, inheriting the cores. That means either a one-year release cycle (using each set of iPhone cores) OR a two-year release cycle, skipping every other generation of iPhone cores. If Apple were to not release M4 until Q4 2025 (due to Apple's fiscal year, that's actually June to September 2025), and it were to be 2nm they would either have to release a new generation of Macs AT the iPhone event (no Mac has ever shared the stage with the iPhone) or use the new 2nm cores in a Mac first. Far more likely is that the first 2nm Mac chip is M5, and we see it sometime between late 2025 and mid-2026, AFTER the A19 round of iPhone chips. It's also possible that the only Mac chip we see between now and late 2025 is M3 Ultra (and possibly a quad M3 Max that people have been calling M3 Extreme), that there is NO Mac chip based off the A18 iPhone cores, and that we see an A19-based (maybe 2nm) M4 shortly after the iPhone 17.
The second problem is that some of the generations farther out are "assuming that such a thing is possible" - and it may well not be. The atomic radius of silicon is between 0.1 and 0.21 nm, depending how you measure it (other likely materials are in the same ballpark). Making features on a chip that are one or two atoms wide involves a LOT of quantum effects, and may or may not violate the Standard Model of physics. In any case, there's not only new engineering, but interesting physics (where "interesting" is defined as "a good way to win a Nobel Prize") involved here. Where is the quantum limit on chip feature size? I'm not confident it allows for some of those generations...
Important to remember that these TSMC (and other) process-node measurements are just names at this point, and they don't correspond to actual measurements. They serve to indicate die shrinks and/or increased density, so they are not completely meaningless relative to one another, but they don't correspond to reality in any meaningful way, as I understand it.
If there's SOMETHING in there that actually measures as small as they're claiming, Heisenberg's Uncertainty Principle will rear its head at some point. Will it be at 10 atomic radii? 5?3? I'm not enough of a physicist to know... I'm quite sure you can't measure accurately enough to get reliable results out of a CPU at 1 atomic radius due to the Uncertainty Principle. The claimed pitches right now are at 15 atomic radii or so, and I just don't know how much tighter we can go (I suspect chip fabs DO employ physicists - but they may not let them talk, for fear of spooking investors).
Richard Feynman, who had both a Nobel Prize in Physics and a great way of explaining it, once said something like "you can ignore quantum effects until things get really big, really small or really fast". A few nanometers is really small, and several gigahertz is really fast. Old Heisenberg is out there somewhere in the next several die shrinks, and he's holding a STOP sign. We just don't know where (or whether it'll be gradual or we'll just hit a single shrink that doesn't work at all)...
Richard Feynman, who had both a Nobel Prize in Physics and a great way of explaining it, once said something like "you can ignore quantum effects until things get really big, really small or really fast". A few nanometers is really small, and several gigahertz is really fast. Old Heisenberg is out there somewhere in the next several die shrinks, and he's holding a STOP sign. We just don't know where (or whether it'll be gradual or we'll just hit a single shrink that doesn't work at all)...
Quantum effects have already been causing problems for the past 15 years. That’s why we had to move from SiO2 to hi-k gate dielectrics at the 45 nm node - the SiO2 gate dielectrics became so thin electrons would quantum tunnel through, preventing the chip from fully turning the transistors off. This off stage leakage current led to huge idle power draws. Switching out the SiO2 for another Hi-K material allowed high switching speeds at an increased gate dielectric thickness (yes, smaller nodes do have a thicker gate dielectric), preventing quantum tunneling. Then to push this further we went to FinFETs, and now are moving to GAA nanosheet transistors and nanowire transistors.
Physics is pretty cool.
TSMC seems confident in their roadmap up to 1 nm and up to 100 billion transistors in a single, "monolithic" chip (and up to 500 billion in 3D "hetero" chiplets), but beyond that there is some hesitation, for "1 nm" and 200 billion/1 trillion. The most telling quote I've seen is the following from the January 2024 earnings call, from Mark Liu, TSMC chairman: "... I want to give the industry an optimistic note that even though 1 nanometer or sub 1 nanometer could be challenging, but we have a new technology capability using AI to accelerate the innovation in science. And that is the -- that is our part, and we have been working on that for many years already." See page 9 of the transcript.
My numbers above are drawn from a TSMC slide from a panel presentation in December 2023, here: IEDM Panel Presentation FINAL2
It's worth noting that this PDF and two of the three TSMC slides in it have been replaced recently. The host removed the references to A14 (1.4 nm) and A10 (1.0 nm) nodes. Unfortunately, I didn't download the original PDF, the new version only mentions N2 (2 nm) and is not specific beyond that... I had commented on this in the thread for the MR article about one of these slides, here:
Apple Chipmaker Discusses Highly Advanced 1.4nm Chips for First Time
TSMC has officially mentioned its work on 1.4nm fabrication technology that is likely destined to underpin future Apple silicon chips. In a slide (via SemiAnalysis's Dylan Patel) from its Future of Logic panel, TSMC disclosed the official name of its 1.4nm node for the first time, "A14." The...
forums.macrumors.com
The photo from Dylan Patel on X shows one of the original slides (compare to the current version, which has deleted the A14 reference), but it doesn't show the original "System Scaling Innovation" slide that also mentioned A10... Regardless, all the references to anything beyond N2 are gone now.
Last edited:
Most likely it’s just that those are not (yet) the official names for those nodes, so TSMC didn’t want that out there. They want to keep their naming options open.
But I think we can read something like that into Mark Liu’s comment that I quoted…