Discussion Apple Silicon SoC thread

Page 391 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Eug

Lifer
Mar 11, 2000
23,992
1,610
126
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:


M1 Ultra discussion here:


M2 discussion here:


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:


M4 Family discussion here:

 
Last edited:

name99

Senior member
Sep 11, 2010
589
489
136
This is what Apple's GPU lacks.
David Huang wrote about this in his blog.
Lack is a strong word. Apple does matrix mul in the GPU by using a permute network to rearrange data as required. The patent is described in my GOU volume.
The cost is a permute network (which can be used for many other things) but you save area. You lose the capacity for weird formats like FP8 unless you want to add those to all datapath, but it’s unclear whether that’s an optimal solution as opposed to the path ANE is going down.
 

LightningZ71

Platinum Member
Mar 10, 2017
2,083
2,534
136
are these real world benefits? not just theory

the cost of 5G outweighs the benefits. think almost trillions implementing it globally. should have been skipped for 6G or 7G where mobile data is a real replacement for wired. 5g is not
Yes, they are real world benefits. The question is, did the cell networks do the proper background work to make sure that all of those benefits are experienced by the end customer?

I have a site that I support that is out in the sticks. There is a cell tower that serves the area, but it is a considerable distance away. It used to be a 4G tower that would allow me to send text messages, occasionally make phone calls provided atmospherics cooperated, and sometimes got me enough data throughput to move small files and send small, low res pictures. It got upgraded to a 5G tower a year ago or so. Now, I can get a couple of bars of signal and see the radio negotiate decent speeds, often in the 100s of Mbps. However, I can't ever achieve an actual data transfer rate higher than 1.5Mbps. My best guess is that the backhaul is probably the equivalent of a T-1 connection, maybe two, one for data and another for control.

Has that improved my experience? Yes, I get more consistent connections and higher data throughput than before. Is it drastic? That depends on your definition. Ideally, I should get much higher data throughput. In reality, it's still worlds better than before.

I have an office downtown. Before 5G, data throughput on wireless was highly susceptible to saturation and massive slowdowns. Looking at the raw side of the connections, I was seeing the radio having to deal with a lot of lots of sharing overhead with the tower. They upgraded it to 5G and now, the radio is doing much better with handling the saturation and overall throughput is much higher. Again, not night and day, but an overall improvement in usability.

It's all highly situational.
 

poke01

Diamond Member
Mar 8, 2022
3,391
4,632
106
Lack is a strong word. Apple does matrix mul in the GPU by using a permute network to rearrange data as required. The patent is described in my GOU volume.
The cost is a permute network (which can be used for many other things) but you save area. You lose the capacity for weird formats like FP8 unless you want to add those to all datapath, but it’s unclear whether that’s an optimal solution as opposed to the path ANE is going down.
Well the current GPU isn’t good enough
 

Doug S

Diamond Member
Feb 8, 2020
3,090
5,328
136
My best guess is that the backhaul is probably the equivalent of a T-1 connection, maybe two, one for data and another for control.

Running underground fiber out to a tower can potentially run into all sorts of roadblocks, while upgrading the tower's antennas only requires the antenna hardware and the crew to become available in that location so it is far less dependent on external factors. Heck they might be waiting on some street construction project that the city pushed back, that kind of stuff can fall years behind schedule (though spending big money can usually get around it, but that's not worth it for an individual tower unless that tower is in a very important spot)

So it is easy to imagine they simply had the antenna upgrade ready before the backhaul upgrade. It wouldn't make sense to upgrade LTE to 5G if they were going to leave it on T1.
 

Doug S

Diamond Member
Feb 8, 2020
3,090
5,328
136
Securing wifi is hard because there's no SIM equivalent, so you have this per-user IT overhead that is unsustainable. In the case of stadiums it was trying to displace wifi, in part because of the security arrangement, but in part because TMobile wasn't making money off of the wifi, but it is off the 5G. So some of it was just a way to capture someone else's revenue, rather than a benefit to users.

But in the larger space of not caring about T-Mobile's top line revenue goals, it wasn't solving a problem, because wifi was in theory there already solving it. In terms of 'what problems does this solve that other technologies don't solve' there were some ideas, but to my knowledge none ever got implemented.

Sure but then you need to have all the equipment include SIMs and until eSIM arrived on the scene (and no idea whether it has made it into the kind of gear you find on a factory floor) it was at least as big of a management headache as having to touch all the equipment to implement proper wifi security.
 

johnsonwax

Member
Jun 27, 2024
134
221
76
Interesting test. Battery life dosent look that much better when using one of the proper browser.

Also a interesting thing, looks like most reviewers are using two different browser to test battery life between MacOS and Windows without mentioning it properly.

Computerbase : https://www.computerbase.de/artikel/notebooks/apple-macbook-air-m4-test.91907/seite-3

View attachment 121336
Why would anyone expect much difference in browser when streaming Youtube? Apart from support of various codecs and maybe utilizing system APIs which could impact decode efficiency, there's not much 'browser' going on in such a test. Most of the efficiency in browsers is in their JS implementations and memory/process handling around multiple tabs. I'm kind of shocked it's even that different.
 

johnsonwax

Member
Jun 27, 2024
134
221
76
Sure but then you need to have all the equipment include SIMs and until eSIM arrived on the scene (and no idea whether it has made it into the kind of gear you find on a factory floor) it was at least as big of a management headache as having to touch all the equipment to implement proper wifi security.
No, I mean, if you have someone sharing kiddie porn over WiFi, that's a you problem. If you have someone sharing it over 5G, that's a them problem. That's what the SIM does for you legally. Everything derives from that. If you want to secure your wifi, you need to do some form of device traceback in place of the SIM (which the carriers can do, but the GB Packers cannot) and/or account management.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |