Discussion Apple Silicon SoC thread

[Eug]

M1
5 nm
Unified memory architecture - LP-DDR4
16 billion transistors

8-core CPU

4 high-performance cores
192 KB instruction cache
128 KB data cache
Shared 12 MB L2 cache

4 high-efficiency cores
128 KB instruction cache
64 KB data cache
Shared 4 MB L2 cache
(Apple claims the 4 high-effiency cores alone perform like a dual-core Intel MacBook Air)

8-core iGPU (but there is a 7-core variant, likely with one inactive core)
128 execution units
Up to 24576 concurrent threads
2.6 Teraflops
82 Gigatexels/s
41 gigapixels/s

16-core neural engine
Secure Enclave
USB 4

Products:
$999 ($899 edu) 13" MacBook Air (fanless) - 18 hour video playback battery life
$699 Mac mini (with fan)
$1299 ($1199 edu) 13" MacBook Pro (with fan) - 20 hour video playback battery life

Memory options 8 GB and 16 GB. No 32 GB option (unless you go Intel).

It should be noted that the M1 chip in these three Macs is the same (aside from GPU core number). Basically, Apple is taking the same approach which these chips as they do the iPhones and iPads. Just one SKU (excluding the X variants), which is the same across all iDevices (aside from maybe slight clock speed differences occasionally).

EDIT:



M1 Pro 8-core CPU (6+2), 14-core GPU
M1 Pro 10-core CPU (8+2), 14-core GPU
M1 Pro 10-core CPU (8+2), 16-core GPU
M1 Max 10-core CPU (8+2), 24-core GPU
M1 Max 10-core CPU (8+2), 32-core GPU

M1 Pro and M1 Max discussion here:

Page 78 - Discussion - Apple Silicon SoC thread

Page 78 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M1 Ultra discussion here:

Page 109 - Discussion - Apple Silicon SoC thread

Page 109 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M2 discussion here:

Page 127 - Discussion - Apple Silicon SoC thread

Page 127 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

Second Generation 5 nm
Unified memory architecture - LPDDR5, up to 24 GB and 100 GB/s
20 billion transistors

8-core CPU

4 high-performance cores
192 KB instruction cache
128 KB data cache
Shared 16 MB L2 cache

4 high-efficiency cores
128 KB instruction cache
64 KB data cache
Shared 4 MB L2 cache

10-core iGPU (but there is an 8-core variant)
3.6 Teraflops

16-core neural engine
Secure Enclave
USB 4

Hardware acceleration for 8K h.264, h.264, ProRes

M3 Family discussion here:

Page 215 - Discussion - Apple Silicon SoC thread

Page 215 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

M4 Family discussion here:

Page 263 - Discussion - Apple Silicon SoC thread

Page 263 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

forums.anandtech.com

 

[moinmoin]

Microtransactions are clearly out of control. I doubt Apple gets as much of that as you might think though.

Apple rules its walled garden with an iron fist so far.

 

[Doug S]

Apple rules its walled garden with an iron fist so far.

Even on Android most people will take the path of least resistance and buy the 'stuff' for their games in the app where Google collects the same 15/30% as Apple. I mean, if you're buying something for $10 and you know you can buy it for $7.50 (assuming they are giving you 100% of their savings which they probably won't) if you go to their web site are you going to bother? Between the extra hassle, the risk of providing credit card numbers to whoever they use for payment processing, and the "I want it NOW" factor most people will just click buy in the app and not worry about saving a few bucks. Then the next time you're spending $10 you're gonna do the same.

 

[Eug]

Who's rumor and what's the record of everyone here for their beliefs. I suspect if anyone cared to actually calculate it, a coin flip would be more accurate. We're probably not as bad as MLID, but that's hardly saying much.

The A17 switching from N3B to N3E rumour was from a Weibo user who claims to be a chip designer (dating back from the Intel Pentium days).

For mainstream Apple rumours, his/her track record may be 100%... because AFAIK s/he may have only 1 prior major Apple prediction that can be verified. (AppleTrack does not follow this Weibo user.) S/he stated before everyone else that A15 would be the chip for iPhone 14. After that statement, Mark Gurman started saying the same thing, as did others. So, does this count towards Mark Gurman's track record percentage if he started saying it only after someone else did?

This person also was the first to say that the USB-C port in iPhone 15 / 15 Pro and charging cables will include a Lightning-like authentication chip. This was later echoed by other "leakers". However, we can't verify this yet, since iPhone 15 isn't out yet.

 

[Doug S]

The A17 switching from N3B to N3E rumour was from a Weibo user who claims to be a chip designer (dating back from the Intel Pentium days).

For mainstream Apple rumours, his/her track record may be 100%... because AFAIK s/he may have only 1 prior major Apple prediction that can be verified. (AppleTrack does not follow this Weibo user.) S/he stated before everyone else that A15 would be the chip for iPhone 14. After that statement, Mark Gurman started saying the same thing, as did others. So, does this count towards Mark Gurman's track record percentage if he started saying it only after someone else did?

This person also was the first to say that the USB-C port in iPhone 15 / 15 Pro and charging cables will include a Lightning-like authentication chip. This was later echoed by other "leakers". However, we can't verify this yet, since iPhone 15 isn't out yet.

Maynard suggested a few years ago that Apple would start segmenting Pro and non Pro iPhones by this year's and last year's chip. That wasn't a "leak", just something that made sense to him. I was skeptical when he said it because there seemed to be little value in such segmentation when the price difference was so small, but between the cost of LPDDR5 for iPhone 14 and the cost of N3 for iPhone 15 it makes perfect sense they did it when they did.

This Weibo user could be making similar educated guesses, as a suggestion that Apple would switch from N3B to N3E is hardly outrageous as Apple might save a lot of money doing so. It is rumored that TSMC's KGD pricing is a short term deal, and it makes a lot of sense TSMC would end that deal when N3E enters mass production.

I'd say the same "educated guess" about an authentication chip being involved with iPhone 15 if it has a USB-C port. That seems like something Apple might do, but man if they do that is it ever going to set off the Apple haters!

 

[MadRat]

It makes sense in a sick way for Apple to provide an official USB-C support via dongle that requires the dongle to be from them. Satisfy the EU with a middle finger.

 

[eek2121]

It makes sense in a sick way for Apple to provide an official USB-C support via dongle that requires the dongle to be from them. Satisfy the EU with a middle finger.

IIRC some EU members already warned Apple this would not be acceptable. Apple will likely add the chip, but only enable in places like the U.S.

I actually don’t think the chip is bad, but Apple using it for profit is bad. It is easy to make a USB cable (you can even make one at home!). It is hard to make a safe/reliable one. These chips stop any old shop from pumping out junk for a quick buck. They allow for at least a minimum standard of quality in theory.

Apple owning that chip means they make money off every sale and can also form a walled garden. That is the part I take issue with.

 

[smalM]

It is rumored that TSMC's KGD pricing is a short term deal, and it makes a lot of sense TSMC would end that deal when N3E enters mass production.

I would be very surprised if the deal wasn't directly tied to specific targets in N3 production quality.

 

[Doug S]

IIRC some EU members already warned Apple this would not be acceptable. Apple will likely add the chip, but only enable in places like the U.S.

I actually don’t think the chip is bad, but Apple using it for profit is bad. It is easy to make a USB cable (you can even make one at home!). It is hard to make a safe/reliable one. These chips stop any old shop from pumping out junk for a quick buck. They allow for at least a minimum standard of quality in theory.

Apple owning that chip means they make money off every sale and can also form a walled garden. That is the part I take issue with.

If it is like the current MFi program it won't be a profit center, as that never was. When I got the first iPhone that had the Lightning port (iPhone 5 IIRC?) in the first few months spare cables cost a ton, but by the next spring I was able to buy a pack of three legit MFi cables for I think $7 on eBay. At that price Apple can't be making more than a few pennies on MFi. It was two 3' and one 6', and I still use that same 6' cable for charging by my bed a decade later (yes it is USB-A, but fast charging is pointless while I'm sleeping)

There's no way they wouldn't support the use of generic USB-C cables, but it might pop up a warning or limit functionality in some way (i.e. maybe it only works for charging not for data connections, which the EU rules do not cover)

 

[eek2121]

If it is like the current MFi program it won't be a profit center, as that never was. When I got the first iPhone that had the Lightning port (iPhone 5 IIRC?) in the first few months spare cables cost a ton, but by the next spring I was able to buy a pack of three legit MFi cables for I think $7 on eBay. At that price Apple can't be making more than a few pennies on MFi. It was two 3' and one 6', and I still use that same 6' cable for charging by my bed a decade later (yes it is USB-A, but fast charging is pointless while I'm sleeping)

You would be surprised at how quickly a few cents per sale of basically every wired phone gadget adds up. Apple likely has made quite a bit of money from the program. Especially since charging cables don’t last forever.

 

[MadRat]

You would be surprised at how quickly a few cents per sale of basically every wired phone gadget adds up. Apple likely has made quite a bit of money from the program. Especially since charging cables don’t last forever.

My wife likes to buy Walmart options because she twists up her cables. The OEM cable lasted a very short time, even less then her Walmart junk. The OEM quality is something to be desired. But you get what you pay for. I bought a $40 fast charge cable for her and its so thick she hasn't twisted it yet. Her wireless iWatch charger cord has also lasted, so its her thats the problem. I believe Apple builds stuff to break, taking advantage of people like my wife that aren't super careful with her things. That few cents per new cable probably adds up fast.

 

[Eug]

The chip and licensing fees for Lightning were said to be $3.50 per item in 2012. If accurate, it would have been impossible to sell 3 Lightning cables for $7 unless it was a clearout item.

Lightning connector viewed as long-term investment, may not be replaced for 10 years | AppleInsider

Apple's new Lightning connector, introduced alongside the iPhone 5 last week, is thought to be a key long-term investment for the company, and will possibly have a lifetime of ten years.

appleinsider.com

 

[Doug S]

The chip and licensing fees for Lightning were said to be $3.50 per item in 2012. If accurate, it would have been impossible to sell 3 Lightning cables for $7 unless it was a clearout item.

Lightning connector viewed as long-term investment, may not be replaced for 10 years | AppleInsider

Apple's new Lightning connector, introduced alongside the iPhone 5 last week, is thought to be a key long-term investment for the company, and will possibly have a lifetime of ten years.

appleinsider.com

This was not a closeout, and they are even cheaper now. Just googled for the heck of it. There's one place advertising 3 for $0.99 with free shipping. Might be a scam (and they are only 1 foot long so not very useful) but the better deals up being about $2 per cable.

The $3.50 was probably for stuff like Bose speakers that would have a Lightning port to dock to, though not long after Lightning came out the whole "docking" market died in favor of Bluetooth.

 

[Eug]

This was not a closeout, and they are even cheaper now. Just googled for the heck of it. There's one place advertising 3 for $0.99 with free shipping. Might be a scam (and they are only 1 foot long so not very useful) but the better deals up being about $2 per cable.

The $3.50 was probably for stuff like Bose speakers that would have a Lightning port to dock to, though not long after Lightning came out the whole "docking" market died in favor of Bluetooth.

The only times in the last decade I have ever seen a legit MFi Lightning cable in Canada for less than CA$8.99 / US$6.80 was on clearance, or else surplus stock sold at Dollarama. (Dollarama gets a lot of old stock and sells it for cheap. I bought a well-built aluminum iPad stand there for $4, but the original price was $49.99 or something like that.) Even the $8.99 I mentioned above was a sale price, as the MSRP has always been well above CA$10.

 

[eek2121]

The only times in the last decade I have ever seen a legit MFi Lightning cable in Canada for less than CA$8.99 / US$6.80 was on clearance, or else surplus stock sold at Dollarama. (Dollarama gets a lot of old stock and sells it for cheap. I bought a well-built aluminum iPad stand there for $4, but the original price was $49.99 or something like that.) Even the $8.99 I mentioned above was a sale price, as the MSRP has always been well above CA$10.

Yep, they are cheap these days, but it is a shame lightning is being killed off. We really need cable recycling here in the US. All of those lightning cables will end up getting tossed in the garbage.

 

[eek2121]

The only times in the last decade I have ever seen a legit MFi Lightning cable in Canada for less than CA$8.99 / US$6.80 was on clearance, or else surplus stock sold at Dollarama. (Dollarama gets a lot of old stock and sells it for cheap. I bought a well-built aluminum iPad stand there for $4, but the original price was $49.99 or something like that.) Even the $8.99 I mentioned above was a sale price, as the MSRP has always been well above CA$10.

Amazon has them really cheap. While some of the brands are questionable, even the reputable ones sell for $6-$10/cable.

The garbage brands that sell you 6 for $10 work fine, however. I have tried them. My kids use them currently. Unsure if the cables from the questionable brands are really certified or not, though they are supposedly.

 

[Eug]

It has been claimed that the Lightning authentication chip was reverse engineered back in 2012.

If there really is an authentication chip in the new iPhone USB-C connector, that would be a perfect time for them to implement a completely new authentication system.

 

[Doug S]

The more I think about it the more skeptical I am that they would introduce authentication into a USB-C cable. If they were going to do that, they would have done so with the iPad which already uses USB-C.

I'm not sure what such a thing would be intended to accomplish. In what scenarios would an iPhone refuse to use a generic USB-C cable and insist on an authenticated one, and why? I know, I know, the Apple haters will claim Apple would do it simply to gouge people, but Apple has a ton of pricing power. All the rumors claimed Apple would raise prices last year because inflation would give them a perfect excuse, but iPhone prices remained the same. Again this year it is rumored they will raise them. If they keep predicting that, eventually they will be right.

If they raised prices by $50 on each model that wouldn't even account for the total amount of inflation since the start of the pandemic but would be an order of magnitude more revenue than they could possibly gain from authenticated USB-C cables, and it wouldn't annoy customers or give ammo to haters.

 

[MadRat]

The connector isn't propriety. How it communicates is propriety, hence USB-C has like 50 versions.

 

[poke01]

Apple Testing M3 Max Chip With 16-Core CPU and 40-Core GPU

Apple is testing the next-generation M3 Max chip, according to Bloomberg's Mark Gurman. The Apple silicon chip will be a replacement for the M2...

www.macrumors.com

M3 Max - 3nm TSMC
12 Performance and 4 Efficiency cores
Up to 40 core GPU

looks like M3 generation is looking good.

A17 chip reveal next month should more about the core architecture improvements

 

[mikegg]

You're sounding more and more ridiculous.

1) $900 is not $700.
2) Apple computers go on sale EVERY FRICKIN' YEAR on Black Friday/Cyber Monday. This is not new.
3) The list price is still $1000. $1000 is 43% more than $700.

Come back when you actually have something real to argue with.


Whatever you're smokin', I want some!

Yes, this isn't the Macbook SE. But it basically is in spirit - old MBA body, older SoC. It has been $750 on Amazon for weeks now as well as Costco.

Your post was 3 years ago. Inflation has been 18% since 2020. $750 today is roughly $615 in 2020 dollars.

Still think an inexpensive future Macbook SE is ridiculous?

As I explained years ago, Apple is perfectly fine with dipping into the "low-midrange market". Thet've done it with the iPhone SE, iPad, and Watch already. Yes, they'll still sell a $5,000 maxed out MBP but don't be surprised if Apple officially releases a $750 - $799 Macbook SE in the near future using the old MBA body and maybe M2 SoC when M3 is out.

 

[mikegg]

M3 Max - 3nm TSMC
12 Performance and 4 Efficiency cores
Up to 40 core GPU

looks like M3 generation is looking good.

A17 chip reveal next month should more about the core architecture improvements

It's likely up to 48 GPU cores. They could have been testing a lower-binned version of M3 Max.

iPhone 15 Pro's A17 Chip Said to Feature 6 GPU Cores, Support for 6GB RAM

Apple's upcoming A17 chip that is set to be used in the iPhone 15 Pro and iPhone 15 Pro Max models includes a 6-core CPU and a 6-core GPU,...

www.macrumors.com

A17 is rumored to have 6 GPU cores - up from 5 in the A15. That would cascade into 48 cores for the M3 Max.

16/48 is a significant upgrade from 12/32 of the M1 Max plus whatever improvements they have in store for the Neural Engine. Pretty enticing upgrade for M1 Max owners.

 

[smalM]

That would cascade into 48 cores for the M3 Max.

A17 5+1
M3 10+2
M3 Pro 19+4
M3 Max 38+8
1,5~1,6 GHz

 

[richardskrad]

It's crazy how the M1 is just as fast as the day I bought it at the end of 2020. It's going to be hard for Apple to entice M1 owners to buy new generations. They totally knocked it out of the park.

 

[igor_kavinski]

It's crazy how the M1 is just as fast as the day I bought it at the end of 2020.

What do you think feels fast? I have M1 Air and it just feels normal. Not lightning fast like I hoped. I mean, the software installations went normal fast. They didn't feel quicker than the ones I saw on Alder Lake. The main thing that's impressive is the silent operation.

 

[Doug S]

A17 is rumored to have 6 GPU cores - up from 5 in the A15. That would cascade into 48 cores for the M3 Max.

16/48 is a significant upgrade from 12/32 of the M1 Max plus whatever improvements they have in store for the Neural Engine. Pretty enticing upgrade for M1 Max owners.

The main thing with A17's GPU is it that it is getting the big upgrade that A16 was supposed to get but they couldn't get it working in time to make the cut so they had to reuse the A15's GPU core. Supposedly was going to be 20-30% faster per core.

So even if the number of cores remains the same it could be a pretty big boost once you add the clock speed gain from N3, and this is likely a further refined version of the GPU core intended for A16.

Kind of surprised we haven't seen any GB6 for A17 yet. Usually we see something by now presumably leaked from someone working at the Foxconn plant where they are stocking up for next month's launch. I don't count the leak from last March - I'm always skeptical of anything coming that early. Though the single core result of 3000 he claimed is about what I would expect, certainly that's more believable than the ridiculous claim of 3900 from the only other early leak I saw lol