Difference between revisions of "Apple Silicon"
| Line 342: | Line 342: | ||
16-Core Neural Engine | 16-Core Neural Engine | ||
|8-10 Cores | |8-10 Cores | ||
|Dual channel LPDDR5-6400 | |||
|CPU: 15W | |CPU: 15W | ||
GPU: 15W | GPU: 15W | ||
| Line 356: | Line 356: | ||
16-Core Neural Engine | 16-Core Neural Engine | ||
|16-19 Cores | |16-19 Cores | ||
| rowspan="2" |Quad channel LPDDR5-6400 | |||
|CPU: 30W | |CPU: 30W | ||
GPU: 30W | GPU: 30W | ||
| Line 369: | Line 370: | ||
16-Core Neural Engine | 16-Core Neural Engine | ||
|30-38 Cores | |30-38 Cores | ||
|CPU: 30W | |CPU: 30W | ||
GPU: 55W | GPU: 55W | ||
Latest revision as of 17:45, 31 January 2023
A Series
| Image | Name | Manufacturing Process | CPU | GPU | Memory | First Released |
|---|---|---|---|---|---|---|
|
APL 0098 | Samsung 90nm | 412MHz
1-Core ARM11 |
1-Core PowerVR MBX Lite | Single channel LPDDR-266 | June 2007 |
|
APL 0298 | Samsung 65nm | 600MHz
1-Core ARM Cortex-A8 |
1-Core PowerVR SGX535 | Single channel LPDDR-400 | June 2009 |
|
A4 | Samsung 45nm | Up to 1.0GHz
1-Core ARM Cortex-A8 |
1-Core PowerVR SGX535 | Dual channel LPDDR-400 | April 2010 |
|
A5 | Samsung 45nm→32nm | Up to 1.0GHz
2-Core ARM Cortex-A9 |
2-Core PowerVR SGX543MP2 | Dual channel LPDDR2-800 | March 2011 |
|
A5X | Samsung 45nm | 1.0GHz
2-Core ARM Cortex-A9 |
4-Core PowerVR SGX543MP4 | Quad channel LPDDR2-800 | March 2012 |
|
A6 | Samsung 32nm | 1.3GHz
2 performance cores (Swift) |
3-Core PowerVR SGX543MP3 | Dual channel LPDDR2-1066 | September 2012 |
|
A6X | 1.4GHz
2 performance cores (Swift) |
4-Core PowerVR SGX554MP4 | Quad channel LPDDR2-1066 | November 2012 | |
|
A7 | Samsung 28nm | Up to 1.4GHz
2 performance cores (Cyclone) |
4-Core PowerVR G6430 | Single channel LPDDR3-1600 | September 2013 |
|
A8 | TSMC 20nm | Up to 1.5GHz
2 performance cores (Typhoon) |
4-Core PowerVR GXA6450 | September 2014 | |
|
A8X | 1.5GHz
3 performance cores (Typhoon) |
8-Core PowerVR GXA6850 | Dual channel LPDDR3-1600 | October 2014 | |
|
A9 | Samsung 14nm
TSMC 16nm |
1.85GHz
2 performance cores (Twister) |
6-Core PowerVR GT7600 | Single channel LPDDR4-3200 | September 2015 |
|
A9X | TSMC 16nm | 2.25GHz
2 performance cores (Twister) |
12-Core PowerVR GTA7850 | Dual channel LPDDR4-3200 | November 2015 |
|
A10 | Up to 2.35GHz
2 performance cores (Hurricane) 2 efficiency cores (Zephyr) |
6-Core PowerVR GT7600 Plus | Single channel LPDDR4-3200 | September 2016 | |
|
A10X | TSMC 10nm | Up to 2.35GHz
3 performance cores (Hurricane) 3 efficiency cores (Zephyr) |
12-Core PowerVR GT7600 Plus | Dual channel LPDDR4-3200 | June 2017 |
|
A11 | Up to 2.4GHz
2 performance cores (Monsoon) 4 efficiency cores (Mistral) 2-Core Neural Engine |
3-Core Apple GPU | Single channel LPDDR4X-4266 | September 2017 | |
|
A12 | TSMC 7nm (N7) | Up to 2.5GHz
2 performance cores (Vortex) 4 efficiency cores (Tempest) 8-Core Neural Engine |
4-Core Apple GPU | September 2018 | |
|
A12X | Up to 2.5GHz
4 performance cores (Vortex) 4 efficiency cores (Tempest) 8-Core Neural Engine |
7-Core Apple GPU | Dual channel LPDDR4X-4266 | November 2018 | |
| A12Z | 8-Core Apple GPU | March 2020 | ||||
|
A13 | TSMC 7nm (N7P) | Up to 2.65GHz
2 performance cores (Lightning) 4 efficiency cores (Thunder) 8-Core Neural Engine |
4-Core Apple GPU | Single channel LPDDR4X-4266 | September 2019 |
|
A14 | TSMC 5nm (N5) | Up to 3.0GHz
2 performance cores (Firestorm) 4 efficiency cores (Icestorm) 16-Core Neural Engine |
October 2020 | ||
|
A15 | TSMC 5nm (N5P) | Up to 3.2GHz
2 performance cores (Avalanche) 4 efficiency cores (Blizzard) 16-Core Neural Engine |
4-5 Core Apple GPU | September 2021 | |
|
A16 | TSMC 4nm (N4) | Up to 3.5GHz
2 performance cores (Everest) 4 efficiency cores (Sawtooth) 16-Core Neural Engine |
5-Core Apple GPU | Single channel LPDDR5-6400 | September 2022 |
M Series
Mac Architectures:
| Architecture | Years In Use |
|---|---|
| Motorola 68000 | 1983-1996 |
| IBM PowerPC | 1994-2006 |
| Intel x86 | 2006-2023 |
| Apple ARM | 2020- |
| Image | Name | Manufacturing Process | CPU | GPU | Memory | TDP | First Released |
|---|---|---|---|---|---|---|---|
|
M1 | TSMC 5nm (N5) | Up to 3.2GHz 8-Core
4 performance cores (Firestorm) 4 efficiency cores (Icestorm) 16-Core Neural Engine |
7-8 Cores | Dual channel LPDDR4X-4266 | CPU: 15W
GPU: 12W |
November 2020 |
|
M1 Pro | Up to 3.2GHz 8-10 Core
6-8 performance cores (Firestorm) 2 efficiency cores (Icestorm) 16-Core Neural Engine |
14-16 Cores | Dual channel LPDDR5-6400 | CPU: 30W
GPU: 30W |
October 2021 | |
|
M1 Max | Up to 3.2GHz 10-Core
8 performance cores (Firestorm) 2 efficiency cores (Icestorm) 16-Core Neural Engine |
24-32 Cores | Quad channel LPDDR5-6400 | CPU: 30W
GPU: 55W | ||
|
M1 Ultra | Up to 3.2GHz 20-Core
16 performance cores (Firestorm) 4 efficiency cores (Icestorm) 32-Core Neural Engine |
48-64 Cores | 8-channel LPDDR5-6400 | CPU: 60W
GPU: 100W |
March 2022 | |
|
M2 | TSMC 5nm (N5P) | Up to 3.5GHz 8-Core
4 performance cores (Avalanche) 4 efficiency cores (Blizzard) 16-Core Neural Engine |
8-10 Cores | Dual channel LPDDR5-6400 | CPU: 15W
GPU: 15W |
June 2022 |
| M2 Pro | Up to 3.5GHz 10-12 Core
6-8 performance cores (Avalanche) 4 efficiency cores (Blizzard) 16-Core Neural Engine |
16-19 Cores | Quad channel LPDDR5-6400 | CPU: 30W
GPU: 30W |
January 2023 | ||
| M2 Max | Up to 3.7GHz 12-Core
8 performance cores (Avalanche) 4 efficiency cores (Blizzard) 16-Core Neural Engine |
30-38 Cores | CPU: 30W
GPU: 55W |
| M1 | M1 Pro | M1 Max | M1 Ultra | M2 | M2 Pro | M2 Max | |
|---|---|---|---|---|---|---|---|
| VP9 | Decode | ||||||
| AVC (H.264) | Encode/Decode | ||||||
| HEVC (H.265) | Encode/Decode | ||||||
| ProRes | No | Encode/Decode | |||||
| AV1 | No | ||||||