6 ms·

How many years will it take before an Apple competitor can make an ARM ISA SoC that is anywhere near "x86-64 high end" performance? Is anyone trying currently?

by bni 6y ago

izacus 6y ago
Why does it have to be ARM? It seems way more likely that AMD will continue improving their Zen x86-64 cores to stay within the envelope. Their improvements have been very notable as well and the M1 doesn't humiliate Zen3s in the same way it does anything that Intel built.
- float4 6y ago
  And Zen3 doesn't even use TSMC's 5nm node whereas M1 does. Apple has payed a ridiculous amount of money to get all of the 5nm capacity for 2020, but it has definitely payed off when you look at all the hype.
  - deeringc 6y ago
    Yeah, this is what is jumping out at me. Not that the M1 isnt extremely impressive but a near future 5nm AMD will be a better "apples to apples" comparison. Intel is being left in the dust though.
- mlindner 6y ago
  Because the amount of legacy baggage on x86-64 (naming continuing to allow x86 and even 16-bit instructions to run) is ridiculous.
Veedrac 6y ago
Give it a few months, since the Cortex X1 should be competitive in laptops.
- ptomato 6y ago
  ARM claims X1 has about 30% better integer perf than the A77; given how the A77 performs in Snapdragon 865, that would mean the M1 is only about 45% faster single core than the X1.
boudin 6y ago
I don't see why it would be hard. Iterating on a architecture is harder than starting one from scratch. If the software market moves to ARM and there's bigger opportunities on the laptop/desktop side, it will attract other players in the high performance market. Until now there was very little incentive due to the lack of software on mainstream operating systems (microsoft learnt this the hard way). Linux market on laptop an desktop has been way too small to attract big investments in the area. After x86 is not dead, AMD is competing pretty well with their Zen architecture. It will be interesting to see what the next iteration will bring in term of power consumption and performance. I also wonder if being an SOC with integrated memory helps the M1 in term of performance (I guess it must improve data access latency). It might push competitors to build high performance SOC.
- spijdar 6y ago
  > "Iterating on a architecture is harder than starting one from scratch." Any reason to believe this is generally true? This seems like the same fallacy that leads to people believing that scrapping all the code and starting from scratch will make a project better, and typically seems to end in more-bugs and fewer-features with delays in release. It can be true when the old code/design is just incredibly bad and needs to be tossed, but I don't think it's true in general.
  - wmil 6y ago
    It's true if you read it as "breaking backwards compatibility is easier". Intel x86 chips would have had things like out of order execution many years earlier if they didn't need to be compatible with old software. Some people have said that a lot of the 737 Max problems are due to Boeing execs insisting on keeping the same airframe. I admit that isn't quite what the parent said, but realistically any architecture "from scratch" is going to reuse a lot of stuff from existing designs that works well.
  - slx26 6y ago
    When the iterations on a given project or architecture start hitting diminishing returns, trying new things from scratch is often a good strategy. I assume the commenter meant the quote for this context, but failed to make it explicit.
rsynnott 6y ago
AWS's one seems fairly impressive. I wouldn't expect much from Qualcomm, who seem kinda moribund.
- niek_pas 6y ago
  > Moribund: (of a thing) in terminal decline; lacking vitality or vigour.
- ptomato 6y ago
  Graviton 2 single core performance afaict is about the same as the best Qualcomm has to offer currently (though obviously AWS is putting a lot more cores in a chip), and the M1 is about 1.8x-2x as fast as both of those (based on SPEC2017)
amelius 6y ago
Google is trying to make the open-source hardware space more mature. Hopefully that will produce some serious competition. https://opensource.googleblog.com/2019/09/unleashing-open-source-silicon.html https://opensource.googleblog.com/2019/09/unleashing-open-so... https://news.ycombinator.com/item?id=21633406 https://news.ycombinator.com/item?id=21633406
akmittal 6y ago
There were few leaks suggesting Google might be working in their own ARM processor. Qualcomm is only other company capable to competing Apple SoC. They tried with 8cx(7 watt) but was not very successfull. With Custom X1 cores they may be able to achieve good performance.
- cbozeman 6y ago
  I don't think anyone will need to worry about Google. As soon as they realize its actually hard, and as soon as they hit the slightest roadblock, they'll give up, like they literally do on every fucking product that isn't a huge instant hit that everyone fawns over.
  - abakker 6y ago
    I don't know - imagine a chromebook, but with a custom SOC with compiler optimizations designed to improve battery life on auto-play video ad popups. Seems right up their alley /s
    - akmittal 6y ago
      I see Google moving from Advertisement model to Subscription based. They did same with YouTube.
  - akmittal 6y ago
    Google is not really good at hardware, they tried high end phones but were not very successful. They realised that and moved to mid tier phones. More than performance they are focusing on creating a Google ecosystem where all their hardware and services integrate seamlessly.
    - cbozeman 6y ago
      Yeah, as it turns out, making exceptional shit from the ground up is really really tough. Hence the reason Google Fiber got laughed out of existence. Google thought everyone would just bend over backwards and laud their arrival. I guess they were shocked that it would require money and actual effort, like most difficult undertakings.
ksec 6y ago
For many worksloads the AWS Graviton 2 is already competitive. ARM currently points to a roadmap of 30% Performance Improvement YoY. That is inclusive of Process Node Improvements. I would not be suspired if Amazon reveals Graviton 3 in re:Invent 2020, and Graviton 4 next year with 5nm. Many of these R&D will filter through to other ARM IPs. So it is perfectly reasonable to expect it would be available to Desktop as well. The problem is no one has the incentive to push ARM on Windows against Intel at its home turf.

How many years will it take before an Apple competitor can make an ARM ISA SoC that is anywhere near "x86-64 high end" performance? Is anyone trying currently?

by bni 6y ago

How many years will it take before an Apple competitor can make an ARM ISA SoC that is anywhere near "x86-64 high end" performance? Is anyone trying currently?

izacus 6y ago

Why does it have to be ARM? It seems way more likely that AMD will continue improving their Zen x86-64 cores to stay within the envelope. Their improvements have been very notable as well and the M1 doesn't humiliate Zen3s in the same way it does anything that Intel built.

float4 6y ago
And Zen3 doesn't even use TSMC's 5nm node whereas M1 does. Apple has payed a ridiculous amount of money to get all of the 5nm capacity for 2020, but it has definitely payed off when you look at all the hype.
- deeringc 6y ago
  Yeah, this is what is jumping out at me. Not that the M1 isnt extremely impressive but a near future 5nm AMD will be a better "apples to apples" comparison. Intel is being left in the dust though.
mlindner 6y ago
Because the amount of legacy baggage on x86-64 (naming continuing to allow x86 and even 16-bit instructions to run) is ridiculous.

Veedrac 6y ago

Give it a few months, since the Cortex X1 should be competitive in laptops.

ptomato 6y ago
ARM claims X1 has about 30% better integer perf than the A77; given how the A77 performs in Snapdragon 865, that would mean the M1 is only about 45% faster single core than the X1.

boudin 6y ago

I don't see why it would be hard. Iterating on a architecture is harder than starting one from scratch. If the software market moves to ARM and there's bigger opportunities on the laptop/desktop side, it will attract other players in the high performance market. Until now there was very little incentive due to the lack of software on mainstream operating systems (microsoft learnt this the hard way). Linux market on laptop an desktop has been way too small to attract big investments in the area. After x86 is not dead, AMD is competing pretty well with their Zen architecture. It will be interesting to see what the next iteration will bring in term of power consumption and performance. I also wonder if being an SOC with integrated memory helps the M1 in term of performance (I guess it must improve data access latency). It might push competitors to build high performance SOC.

spijdar 6y ago
> "Iterating on a architecture is harder than starting one from scratch." Any reason to believe this is generally true? This seems like the same fallacy that leads to people believing that scrapping all the code and starting from scratch will make a project better, and typically seems to end in more-bugs and fewer-features with delays in release. It can be true when the old code/design is just incredibly bad and needs to be tossed, but I don't think it's true in general.
- wmil 6y ago
  It's true if you read it as "breaking backwards compatibility is easier". Intel x86 chips would have had things like out of order execution many years earlier if they didn't need to be compatible with old software. Some people have said that a lot of the 737 Max problems are due to Boeing execs insisting on keeping the same airframe. I admit that isn't quite what the parent said, but realistically any architecture "from scratch" is going to reuse a lot of stuff from existing designs that works well.
- slx26 6y ago
  When the iterations on a given project or architecture start hitting diminishing returns, trying new things from scratch is often a good strategy. I assume the commenter meant the quote for this context, but failed to make it explicit.

rsynnott 6y ago

AWS's one seems fairly impressive. I wouldn't expect much from Qualcomm, who seem kinda moribund.

niek_pas 6y ago
> Moribund: (of a thing) in terminal decline; lacking vitality or vigour.
ptomato 6y ago
Graviton 2 single core performance afaict is about the same as the best Qualcomm has to offer currently (though obviously AWS is putting a lot more cores in a chip), and the M1 is about 1.8x-2x as fast as both of those (based on SPEC2017)

amelius 6y ago

Google is trying to make the open-source hardware space more mature. Hopefully that will produce some serious competition. https://opensource.googleblog.com/2019/09/unleashing-open-source-silicon.html https://opensource.googleblog.com/2019/09/unleashing-open-so... https://news.ycombinator.com/item?id=21633406 https://news.ycombinator.com/item?id=21633406

akmittal 6y ago

There were few leaks suggesting Google might be working in their own ARM processor. Qualcomm is only other company capable to competing Apple SoC. They tried with 8cx(7 watt) but was not very successfull. With Custom X1 cores they may be able to achieve good performance.

cbozeman 6y ago
I don't think anyone will need to worry about Google. As soon as they realize its actually hard, and as soon as they hit the slightest roadblock, they'll give up, like they literally do on every fucking product that isn't a huge instant hit that everyone fawns over.
- abakker 6y ago
  I don't know - imagine a chromebook, but with a custom SOC with compiler optimizations designed to improve battery life on auto-play video ad popups. Seems right up their alley /s
  - akmittal 6y ago
    I see Google moving from Advertisement model to Subscription based. They did same with YouTube.
- akmittal 6y ago
  Google is not really good at hardware, they tried high end phones but were not very successful. They realised that and moved to mid tier phones. More than performance they are focusing on creating a Google ecosystem where all their hardware and services integrate seamlessly.
  - cbozeman 6y ago
    Yeah, as it turns out, making exceptional shit from the ground up is really really tough. Hence the reason Google Fiber got laughed out of existence. Google thought everyone would just bend over backwards and laud their arrival. I guess they were shocked that it would require money and actual effort, like most difficult undertakings.

ksec 6y ago

For many worksloads the AWS Graviton 2 is already competitive. ARM currently points to a roadmap of 30% Performance Improvement YoY. That is inclusive of Process Node Improvements. I would not be suspired if Amazon reveals Graviton 3 in re:Invent 2020, and Graviton 4 next year with 5nm. Many of these R&D will filter through to other ARM IPs. So it is perfectly reasonable to expect it would be available to Desktop as well. The problem is no one has the incentive to push ARM on Windows against Intel at its home turf.