tidus1979
Captain
- Registriert
- Aug. 2016
- Beiträge
- 3.463
Das möchte ich bezweifelnW0dan schrieb:
Simple Erklärungen gibt es hier nicht. Apple macht auch im Design Dinge, die andere nicht machen und die Fachleute ins Staunen bringen. Z.B. Ultra Wide + Deep. Unified Memory. Ultra Fusion. Diese Dinge werden wahrscheinlich auch andere Hersteller inspirieren, was bedeutet, dass Apple hier Vorreiter ist. Teils sieht man das auch schon in den neuesten Designs anderer Hersteller.
[A14] The secret sauce lies in Apple’s in-house CPU microarchitecture. Apple’s long journey into custom CPU microarchitectures started off with the release of the Apple A6 back in 2012 in the iPhone 5. Even back then with their first-generation “Swift” design, the company had marked some impressive performance figures compared to the mobile competition.
The real shocker that really made waves through the industry was however Apple’s subsequent release of the Cyclone CPU microarchitecture in 2013’s Apple A7 SoC and iPhone 5S. Apple’s early adoption of the 64-bit Armv8 ISA shocked everybody, as the company was the first in the industry to implement the new instruction set architecture, but they beat even Arm’s own CPU teams by more than a year, as the Cortex-A57 (Arm own 64-bit microarchitecture design) would not see light of day until late 2014.
https://www.anandtech.com/show/16226/apple-silicon-m1-a14-deep-dive/2
[M1] One aspect we’ve never really had the opportunity to test is exactly how good Apple’s cores are in terms of memory bandwidth. Inside of the M1, the results are ground-breaking: A single Firestorm achieves memory reads up to around 58GB/s, with memory writes coming in at 33-36GB/s. Most importantly, memory copies land in at 60 to 62GB/s depending if you’re using scalar or vector instructions. The fact that a single Firestorm core can almost saturate the memory controllers is astounding and something we’ve never seen in a design before.
https://www.anandtech.com/show/16252/mac-mini-apple-m1-tested
[M1 Max] From a single core perspective, meaning from a single software thread, things are quite impressive for the chip, as it’s able to stress the memory fabric to up to 102GB/s. This is extremely impressive and outperforms any other design in the industry by multiple factors, we had already noted that the M1 chip was able to fully saturate its memory bandwidth with a single core and that the bottleneck had been on the DRAM itself. On the M1 Max, it seems that we’re hitting the limit of what a core can do – or more precisely, a limit to what the CPU cluster can do.
https://www.anandtech.com/show/17024/apple-m1-max-performance-review
[M1 Ultra]
Unlike multi-die/multi-chip CPU configurations, which have been commonplace in workstations for decades, multi-die GPU configurations are a far different beast. The amount of internal bandwidth GPUs consume, which for high-end parts is well over 1TB/second, has always made linking them up technologically prohibitive. As a result, in a traditional multi-GPU system (such as the Mac Pro), each GPU is presented as a separate device to the system, and it’s up to software vendors to find innovative ways to use them together. In practice, this has meant having multiple GPUs work on different tasks, as the lack of bandwidth meant they can’t effectively work together on a single graphics task.
But, if you could somehow link up multiple GPUs with a ridiculous amount die-to-die bandwidth – enough to replicate their internal bandwidth – then you might just be able to use them together in a single task. This has made combining multiple GPUs in a transparent fashion something of a holy grail of multi-GPU design. It’s a problem that multiple companies have been working on for over a decade, and it would seem that Apple is charting new ground by being the first company to pull it off.
https://www.anandtech.com/show/1730...bining-two-m1-maxes-for-even-more-performance
The real shocker that really made waves through the industry was however Apple’s subsequent release of the Cyclone CPU microarchitecture in 2013’s Apple A7 SoC and iPhone 5S. Apple’s early adoption of the 64-bit Armv8 ISA shocked everybody, as the company was the first in the industry to implement the new instruction set architecture, but they beat even Arm’s own CPU teams by more than a year, as the Cortex-A57 (Arm own 64-bit microarchitecture design) would not see light of day until late 2014.
https://www.anandtech.com/show/16226/apple-silicon-m1-a14-deep-dive/2
[M1] One aspect we’ve never really had the opportunity to test is exactly how good Apple’s cores are in terms of memory bandwidth. Inside of the M1, the results are ground-breaking: A single Firestorm achieves memory reads up to around 58GB/s, with memory writes coming in at 33-36GB/s. Most importantly, memory copies land in at 60 to 62GB/s depending if you’re using scalar or vector instructions. The fact that a single Firestorm core can almost saturate the memory controllers is astounding and something we’ve never seen in a design before.
https://www.anandtech.com/show/16252/mac-mini-apple-m1-tested
[M1 Max] From a single core perspective, meaning from a single software thread, things are quite impressive for the chip, as it’s able to stress the memory fabric to up to 102GB/s. This is extremely impressive and outperforms any other design in the industry by multiple factors, we had already noted that the M1 chip was able to fully saturate its memory bandwidth with a single core and that the bottleneck had been on the DRAM itself. On the M1 Max, it seems that we’re hitting the limit of what a core can do – or more precisely, a limit to what the CPU cluster can do.
https://www.anandtech.com/show/17024/apple-m1-max-performance-review
[M1 Ultra]
Unlike multi-die/multi-chip CPU configurations, which have been commonplace in workstations for decades, multi-die GPU configurations are a far different beast. The amount of internal bandwidth GPUs consume, which for high-end parts is well over 1TB/second, has always made linking them up technologically prohibitive. As a result, in a traditional multi-GPU system (such as the Mac Pro), each GPU is presented as a separate device to the system, and it’s up to software vendors to find innovative ways to use them together. In practice, this has meant having multiple GPUs work on different tasks, as the lack of bandwidth meant they can’t effectively work together on a single graphics task.
But, if you could somehow link up multiple GPUs with a ridiculous amount die-to-die bandwidth – enough to replicate their internal bandwidth – then you might just be able to use them together in a single task. This has made combining multiple GPUs in a transparent fashion something of a holy grail of multi-GPU design. It’s a problem that multiple companies have been working on for over a decade, and it would seem that Apple is charting new ground by being the first company to pull it off.
https://www.anandtech.com/show/1730...bining-two-m1-maxes-for-even-more-performance
