Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Apple introduces M4 chip

apple.com

Post reply on HN
Print view View original

251–260 of 1001 posts

Re: Apple introduces M4 chip

#251

Post by paulpan »

dhx wrote: post #158

M2's Neural Engine had 15TOPS, M3's 18TOPS (+20%) vs. M4's 38TOPS (+111%). In transistor counts, M2 had 20BTr, M3 25BTr (+25%) and M4 has 28BTr (+12%). M2 used TSMC N5P (138MTr/mm2), M3 used TSMC N3 (197MTr/mm2, +43%) and M4 uses TSMC N3E (215MTr/mm2, +9%).[1][2] [1] https://en.wikipedia.org/wiki/5_nm_process#%225_nm%22_proces... [2] https://en.wikipedia.org/wiki/3_nm_process#%223_nm%22_proces...

The fact that TSMC publishes their own metrics and target goals for each node makes it straightforward to compare the transistor density, power efficiency, etc.

The most interesting aspect of the M4 is simply it's debuting on the iPad lineup, whereas historically it's always been on the iPhone (for A-series) and Macbook (for M-series). Makes sense given low expected yielded for the newest node for one of Apple's lower volume products.

For the curious, the original TSMC N3 node had a lot of issues plus was very costly so makes sense to move away from it: https://www.semianalysis.com/p/tsmcs-3nm-conundrum-does-it-e...

Re: Apple introduces M4 chip

#252

Post by kkielhofner »

ttul wrote: post #233
dhx wrote: post #158

M2's Neural Engine had 15TOPS, M3's 18TOPS (+20%) vs. M4's 38TOPS (+111%). In transistor counts, M2 had 20BTr, M3 25BTr (+25%) and M4 has 28BTr (+12%). M2 used TSMC N5P (138MTr/mm2), M3 used TSMC N3 (197MTr/mm2, +43%) and M4 uses TSMC N3E (215MTr/mm2, +9%).[1][2] [1] https://en.wikipedia.org/wiki/5_nm_process#%225_nm%22_proces... [2] https://en.wikipedia.org/wiki/3_nm_process#%223_nm%22_proces...

An NVIDIA RTX 4090 generates 73 TFLOPS. This iPad gives you nearly half that. The memory bandwidth of 120 GBps is roughly 1/10th of the NVIDIA hardware, but who’s counting!

TOPS != TFLOPS

RTX 4090 Tensor 1,321 TOPS according to spec sheet so roughly 35x.

RTX 4090 is 191 Tensor TFLOPS vs M2 5.6 TFLOPS (M3 is tough to find spec).

RTX 4090 is also 1.5 years old.

Re: Apple introduces M4 chip

#253

Post by makeitdouble »

bitwize wrote: post #192
praseodym wrote: post #10

"With these improvements to the CPU and GPU, M4 maintains Apple silicon’s industry-leading performance per watt. M4 can deliver the same performance as M2 using just half the power. And compared with the latest PC chip in a thin and light laptop, M4 can deliver the same performance using just a fourth of the power." That's an incredible improvement in just a few years. I wonder how much of that is Apple engineering a…

Ultimately, does it matter? Michelin-starred restaurants not only have top-tier chefs. They have buyers who negotiate with food suppliers to get the best ingredients they can at the lowest prices they can. Having a preferential relationship with a good supplier is as important to the food quality and the health of the business as having a good chef to prepare the dishes. Apple has top-tier engineering talent but they…

Qualcomm is also with TSMC and their newer 4nm processor is expected to stay competitive with the M series.

If the magic comes mostly from TSMC, there's a good chance for these claims to be true and to have a series of better chips coming on the other platforms as well.

Re: Apple introduces M4 chip

#254

Post by lemcoe9 »

ttul wrote: post #233
dhx wrote: post #158

M2's Neural Engine had 15TOPS, M3's 18TOPS (+20%) vs. M4's 38TOPS (+111%). In transistor counts, M2 had 20BTr, M3 25BTr (+25%) and M4 has 28BTr (+12%). M2 used TSMC N5P (138MTr/mm2), M3 used TSMC N3 (197MTr/mm2, +43%) and M4 uses TSMC N3E (215MTr/mm2, +9%).[1][2] [1] https://en.wikipedia.org/wiki/5_nm_process#%225_nm%22_proces... [2] https://en.wikipedia.org/wiki/3_nm_process#%223_nm%22_proces...

An NVIDIA RTX 4090 generates 73 TFLOPS. This iPad gives you nearly half that. The memory bandwidth of 120 GBps is roughly 1/10th of the NVIDIA hardware, but who’s counting!

The 4090 costs ~$1800 and doesn't have dual OLED screens, doesn't have a battery, doesn't weigh less than a pound, and doesn't actually do anything unless it is plugged into a larger motherboard, either.

Re: Apple introduces M4 chip

#256

Post by bombcar »

haunter wrote: post #130

I love when Gruber is confidently wrong https://daringfireball.net/linked/2024/04/28/m4-ipad-pros-gu...

Wasn't it relatively well known that the M3 is on an expensive process and quickly getting to an M4 on a cheaper/higher yield process would be worth it?

Re: Apple introduces M4 chip

#257

Post by atommclain »

alberth wrote: post #154
haunter wrote: post #130

I love when Gruber is confidently wrong https://daringfireball.net/linked/2024/04/28/m4-ipad-pros-gu...

Especially about Gurman, who he loves to hate on.

Never understood the animosity, especially because it seems to only go one direction.

Re: Apple introduces M4 chip

#258

Post by Perceval »

lopkeny12ko wrote: post #54

I always wonder how constraining it is to design these chips subject to thermal and energy limitations. I paid a lot of money for my hardware and I want it to go as fast as possible. I don't want my fans to be quiet, and I don't want my battery life to be 30 minutes longer, if it means I get more raw performance in return. But instead, Apple's engineers have unilaterally decided to handicap their own processors for n…

Most of what Apple sells goes into mobile devices: phone, tablet, laptop. In their prior incarnation, they ran up real hard against the thermal limits of what they could put in their laptops with the IBM PowerPC G5 chip.

Pure compute power has never been Apple's center of gravity when selling products. The Mac Pro and the XServe are/were minuscule portions of Apple's sales, and the latter product was killed after a short while.

> Apple's engineers have unilaterally decided to handicap their own processors for no real good reason

This is a misunderstanding of what the limiting factor is of Apple products' capability. The mobile devices all have battery as the limfac. The processors being energy efficient in compute-per-watt isn't a handicap, it's an enabler. And it's a very good reason.

Re: Apple introduces M4 chip

#259

Post by dialup_sounds »

greggsy wrote: post #212
KeplerBoy wrote: post #186

Earlier quoted context omitted.

If these things will ever get MacOS support it will be useless with 8 GB of Ram. Such a waste of nice components.

This comes up frequently. 8GB is sufficient for most casual and light productivity use cases. Not everyone is a power user, in fact, most people aren’t.

Most people that consider themselves "power users" aren't even power users, either. Like how being into cars doesn't make you a race car driver.

Re: Apple introduces M4 chip

#260

Post by __turbobrew__ »

paulpan wrote: post #183
EduardoBautista wrote: post #13

The 256gb and 512gb models have 8gb of ram. The 1tb and 2tb models have 16gb. Not a fan of tying ram to storage. https://www.apple.com/ipad-pro/specs/

I don't think it's strictly for price gouging/segmentation purposes. On the Macbooks (running MacOS), RAM has been used as data cache to speed up data read/write performance until the actual SSD storage operation completes. It makes sense for Apple to account for with higher RAM spec for the 1TB/2TB configurations.

That is called a buffer/page cache and has existed in operating systems since the 1980s.
Post reply on HN