9 ms·

Most of the solutions in this thread are simply not viable. Stuff like the Pinebook and Raspberry Pi are several magnitudes away from Apple's M1 SoC. No amount

Most of the solutions in this thread are simply not viable. Stuff like the Pinebook and Raspberry Pi are several magnitudes away from Apple's M1 SoC. No amount of hackerish imitation will bridge that gap, the laws of physics are not kind to consumer product hardware hackers. It is very obvious that most of the people commenting here has never held a soldering iron, nor taken a single computer architecture class. The only true way to match Apple on both performance and efficiency is to fund custom silicon. Cerebras needed 100 million to get started. Apple has already laid the architectural groundwork, a chip with similar performance numbers can probably be commissioned with ~50 million. Another 10-20 million for the emulator/hypervisor crowd to get x86 virtualization up and running. This could probably be done by existing players so the cost can be discounted. A crowd investing approach is most likely the best bet, with the initial rounds being led by a VC experienced in hardware. This is the Sputnik moment for open source. Should ACPI (and its associated lassiez-faire mindset) become obsolete as a standard, then general purpose computing on consumer devices would become history. All the comments about how Apple's Macs are not worth it are entirely missing the point. There is nothing comparable to Apple's M1 right now. Your Purism/Lambda labs/Starlabs/System76/Pine64 usual Clevo shell machine with Linux pre-installed would never be able to compete with a M1 powered system. This isn't about 16:10/3:2 display ratios, machined aluminium, or HN's favorite complaint: "build quality". Even a well-funded premium laptop company like Razer would not be able produce something to compete with the MBP if they don't have a similarly efficient chip. Forget about the bigger players like Lenovo and Dell. Intel and AMD dropped the ball big time, and if the open source community doesn't step up, then the future will be locked-down ARM systems. RISC-V will not be competitive for another half a decade and there is no guarantee that it won't be locked down if the community does not push back against the embedded hardware culture (where VHDL/Verilog libraries are referred to as "intellectual property") which pervades ARM's heritage.

ironmagma 6y ago
> This is the Sputnik moment for open source Does it need to be? x86 had every opportunity to become a locked-down, exclusive standard too, and it grew the way it did, into a cross-vendor standard, for a reason. Why will a similar trajectory not apply to the M1?
- ddingus 6y ago
  Because M1 / Arm has a locked down base to draw from, unlike Intel / AMD. Way different roots in mobile when compared to PC
- fieldcny 6y ago
  The history of the x86 instruction set is very complicated and involves a number of lawsuits and settlements. What you see today is the fruit decades of fighting and wrangling.
- skohan 6y ago
  As far as I understand, x86 is a duopoly, and AMD is allowed to produce it because antitrust litigation resulted in Intel being required to allow a second party to implement their instruction set. Also Intel is a very different entity than Apple. They produce commodity hardware, where Apple creates consumer experiences. Apple's incentive is to keep M1 closed to give their software ecosystem an edge vs. competitors.
- yuhong 6y ago
  I'd probably trust MITS to not attack Altair clones (aka S-100) the way IBM attacked PC clones (in the 1980s).
frankchn 6y ago
> Apple has already laid the architectural groundwork, a chip with similar performance numbers can probably be commissioned with ~50 million. I think you might be two orders of magnitude off here. A class leading general purpose compute chip on a cutting edge process? I don’t think $5 billion to produce working chips is unreasonable. If it were only $50M to produce A14/M1 level chips, Qualcomm would already have done so. Cerebras is cool, but they are not on TSMC 5nm and don’t have to design general purpose silicon, where you have pesky things like huge reorder buffers or 8-wide decode blocks, which are not seen in any other production CPU right now.
- ampdepolymerase 6y ago
  You are right, I am mostly estimating based off upgrading existing architectures with insights obtained from M1. 50 million may be enough for the raw manufacturing cost for one batch but research and development from scratch would not be cheap. I am basing the cost off the assumption that there are already experienced chip designers at Intel/Qualcomm who are willing to take the opportunity to build a startup with a 50 mil budget.
- dev_tty01 6y ago
  Agreed. Apple has acquired multiple companies and been building custom CPUs for 10 years with a significant sized design team. They paid about 280M for P.A. Semi in 2008. They reportedly also hired a number of DEC StrongARM refugees as well as several other small design teams. Apple was also an original ARM investor and have been using ARM cores for various uses for decades. A competitor also needs 10 years experience designing higher and higher performance ARM cpus in tremendously heat/power constrained systems (phones). There are a lot of long term threads that have come together in the M1 chip.
- amluto 6y ago
  Building a useful 8-wide decode block on x86 might be extremely expensive in terms of area and power. For a long time, we’ve been told that the nasty x86 ISA is no big deal and that a good CPU can cache decoded instructions and play other tricks to make the ISA irrelevant. But x86 really does have fundamental problems. The length of an instruction can’t be determined until the instruction is almost fully decoded, which makes deciding in parallel quite nasty. And the x86 memory model is fundamentally more expensive than ARM’s. M1 is the first serious attempt to make a competitive high-performance non-x86 CPU. Perhaps the real lesson from M1 will be that x86 can’t compete.
krageon 6y ago
Your argument against an open platform is "more open hardware is slower", as if that matters somehow. 99% of what you use in a day doesn't depend on a fast platform, and the remaining 1% is stuff that you should migrate away from or modern videogames (which arguably you should also move away from).
- cbozeman 6y ago
  > and the remaining 1% is stuff that you should migrate away from or modern videogames (which arguably you should also move away from). Would you expound on this in greater detail? I am interested in your thoughts.
  - krageon 6y ago
    With some notable exceptions applications have become slower in a way that is just fundamentally unnecessary. Fifteen years ago you could do text editing, browsing the internet, chatting and possibly editing spreadsheets all in parallel on a machine that today would not be able to do anything with the same (but upgraded!) applications. This is not due to features that are "just more expensive" (though as said there are some notable exceptions, like probably decoding 4k video on youtube) but because the development process has migrated further and further away from any sort of performance concern as the platforms you could expect became faster. That is not to say the computing landscape was fundamentally better fifteen years ago. I think it is almost the same. What has really improved is the general usability and stability of drivers (no more rebooting, random freezes and if you run linux your X server not working every other update), which is definitely something to be proud of. What has not improved is basically everything else and I'm really not sure why. If I had to guess I would say it has something to do with managers the world over just not giving a shit about delivering something good, as long as it sells. As platforms have started having more headroom for bad software, "as long as it sells" has started to have a wider and wider performance profile. Landing us where we are today: with many of the same features, much less privacy and every other app packed with an entire browser "because it makes it easier to develop". Even as recently as last year I had a more relaxed stance about this but I guess it has just been bothering me now that I have more time to think about the state of my work (and everyone else's). Edit: I see now I responded to something slightly different than what was requested. Modern (AAA) videogames should not be bought or played because they are produced by a sick and exploitative industry that really needs to die. Instead, we can prefer the indie games that work well with what they have: Generally a lower graphical budget that runs significantly better on hardware that is "out of date". For office-type processing and messaging, there are generally application choices and tradeoffs that can be made for more performance instead of more eye candy. The average office worker (which I guess includes most people here) will not materially care that it doesn't look quite as good. Another upside is that as a company you don't need such an enormous budget for office hardware.
AtlasBarfed 6y ago
If the M1 is a close derivative architecture of high efficiency mobile scaled to desktop, would scaled up Android processors really be that far away from M1? I know Apple A-series had some edge over Snapdragons, but was the gap between Android processors that big? There are many hundred billion dollar companies motivated to close the gap. It may not be this year.
- ampdepolymerase 6y ago
  There is the Microsoft SQ2 and the Snapdragon 8cx Gen 2.
- frankchn 6y ago
  The A14 is significantly faster than the Kirin 9000 (Cortex A77 on TSMC 5nm) on SPECint and SPECfp: https://www.anandtech.com/show/16226/apple-silicon-m1-a14-deep-dive/3 https://www.anandtech.com/show/16226/apple-silicon-m1-a14-de...
- hachari 6y ago
  “I know Apple A-series had some edge over Snapdragons, but was the gap between Android processors that big?” Absolutely
  - AtlasBarfed 6y ago
    Remember there is the 5nm advantage.
    - sbuk 6y ago
      What about the other A-series processors? Excluding the A14, they all used the same process as other ARM based processors, but also have a big performance advantage.
pjmlp 6y ago
Well, PC clones only happened thanks to Compaq. Personally I would have been happier with Atari or Amiga instead, targeted mobile dev back in the J2ME/Symbian days, I did had access to console devkits for a while, so for what I care everything is good as it is.
samvher 6y ago
I share your view of the likely future, but unfortunately it seems very unlikely to me that the open source community is going to unite and prevent this in any significant way. What's happening might give a boost to open source hardware and that would be wonderful, but I don't think this hardware will have much of a chance of being economically competitive and capturing a market.
unionpivo 6y ago
Ehh I think people are overestimating M1 impact. Sure its the best laptop/ipad(other pad) chip out there. And it (and it's next versions) will likely remain best for quite a while. But it's not that the rest of the chips out there turned into garbage. They are still very usable. Personally I have gone away from laptops back to desktops. Having desktop and a phone/tablet combination for me is better than just laptop or laptop + phone/tablet. I still have my laptop, but I very rarely use it now. (mostly if I travel, which I don't because of covid :) ) But having latest gratest laptop is not anywhere near the top of my purchasing priority. And on desktops, latest AMD chips and lots of fast ram, are quite decent. Bottom line even if OS never has anything that matches M1, it will still be fine.
- iSnow 6y ago
  OSS for operating systems is unlikely to be fine: out of the four classes server, desktop, laptop, phone/tables, one is already fully locked down and with the M1, the laptop class also becomes less open and hackable. Most people - unlike you - don't care that much about desktops, they either have laptops or tablets.
AntiImperialist 6y ago
Apple has the same sort of ownership of M1 as it did with Thunderbolt, maybe a little bit more but not much more than you'd think. Let me explain. Apple has the performance edge now because it purchased ALL of the 5nm chip fabrication capacity of TSMC. This does not mean: - they will be able to do so forever - someone else won't be doing 5nm in the next couple of years So, the same sort of performance will come to other devices in the near future. The best Apple can hope to do at this time is to get enough attention to get some new users and hopefully lock them in into their cozy eco-system. Most importantly, most users don't care about an expensive powerful devices. Highest selling consumer devices are still budget devices and will continue to be so for the foreseeable future. Servers on the other hand will continue to demand powerful hardware, and Apple isn't doing anything there.

ironmagma 6y ago

> This is the Sputnik moment for open source Does it need to be? x86 had every opportunity to become a locked-down, exclusive standard too, and it grew the way it did, into a cross-vendor standard, for a reason. Why will a similar trajectory not apply to the M1?

ddingus 6y ago
Because M1 / Arm has a locked down base to draw from, unlike Intel / AMD. Way different roots in mobile when compared to PC
fieldcny 6y ago
The history of the x86 instruction set is very complicated and involves a number of lawsuits and settlements. What you see today is the fruit decades of fighting and wrangling.
skohan 6y ago
As far as I understand, x86 is a duopoly, and AMD is allowed to produce it because antitrust litigation resulted in Intel being required to allow a second party to implement their instruction set. Also Intel is a very different entity than Apple. They produce commodity hardware, where Apple creates consumer experiences. Apple's incentive is to keep M1 closed to give their software ecosystem an edge vs. competitors.
yuhong 6y ago
I'd probably trust MITS to not attack Altair clones (aka S-100) the way IBM attacked PC clones (in the 1980s).

frankchn 6y ago

> Apple has already laid the architectural groundwork, a chip with similar performance numbers can probably be commissioned with ~50 million. I think you might be two orders of magnitude off here. A class leading general purpose compute chip on a cutting edge process? I don’t think $5 billion to produce working chips is unreasonable. If it were only $50M to produce A14/M1 level chips, Qualcomm would already have done so. Cerebras is cool, but they are not on TSMC 5nm and don’t have to design general purpose silicon, where you have pesky things like huge reorder buffers or 8-wide decode blocks, which are not seen in any other production CPU right now.

ampdepolymerase 6y ago
You are right, I am mostly estimating based off upgrading existing architectures with insights obtained from M1. 50 million may be enough for the raw manufacturing cost for one batch but research and development from scratch would not be cheap. I am basing the cost off the assumption that there are already experienced chip designers at Intel/Qualcomm who are willing to take the opportunity to build a startup with a 50 mil budget.
dev_tty01 6y ago
Agreed. Apple has acquired multiple companies and been building custom CPUs for 10 years with a significant sized design team. They paid about 280M for P.A. Semi in 2008. They reportedly also hired a number of DEC StrongARM refugees as well as several other small design teams. Apple was also an original ARM investor and have been using ARM cores for various uses for decades. A competitor also needs 10 years experience designing higher and higher performance ARM cpus in tremendously heat/power constrained systems (phones). There are a lot of long term threads that have come together in the M1 chip.
amluto 6y ago
Building a useful 8-wide decode block on x86 might be extremely expensive in terms of area and power. For a long time, we’ve been told that the nasty x86 ISA is no big deal and that a good CPU can cache decoded instructions and play other tricks to make the ISA irrelevant. But x86 really does have fundamental problems. The length of an instruction can’t be determined until the instruction is almost fully decoded, which makes deciding in parallel quite nasty. And the x86 memory model is fundamentally more expensive than ARM’s. M1 is the first serious attempt to make a competitive high-performance non-x86 CPU. Perhaps the real lesson from M1 will be that x86 can’t compete.

krageon 6y ago

Your argument against an open platform is "more open hardware is slower", as if that matters somehow. 99% of what you use in a day doesn't depend on a fast platform, and the remaining 1% is stuff that you should migrate away from or modern videogames (which arguably you should also move away from).

cbozeman 6y ago
> and the remaining 1% is stuff that you should migrate away from or modern videogames (which arguably you should also move away from). Would you expound on this in greater detail? I am interested in your thoughts.
- krageon 6y ago
  With some notable exceptions applications have become slower in a way that is just fundamentally unnecessary. Fifteen years ago you could do text editing, browsing the internet, chatting and possibly editing spreadsheets all in parallel on a machine that today would not be able to do anything with the same (but upgraded!) applications. This is not due to features that are "just more expensive" (though as said there are some notable exceptions, like probably decoding 4k video on youtube) but because the development process has migrated further and further away from any sort of performance concern as the platforms you could expect became faster. That is not to say the computing landscape was fundamentally better fifteen years ago. I think it is almost the same. What has really improved is the general usability and stability of drivers (no more rebooting, random freezes and if you run linux your X server not working every other update), which is definitely something to be proud of. What has not improved is basically everything else and I'm really not sure why. If I had to guess I would say it has something to do with managers the world over just not giving a shit about delivering something good, as long as it sells. As platforms have started having more headroom for bad software, "as long as it sells" has started to have a wider and wider performance profile. Landing us where we are today: with many of the same features, much less privacy and every other app packed with an entire browser "because it makes it easier to develop". Even as recently as last year I had a more relaxed stance about this but I guess it has just been bothering me now that I have more time to think about the state of my work (and everyone else's). Edit: I see now I responded to something slightly different than what was requested. Modern (AAA) videogames should not be bought or played because they are produced by a sick and exploitative industry that really needs to die. Instead, we can prefer the indie games that work well with what they have: Generally a lower graphical budget that runs significantly better on hardware that is "out of date". For office-type processing and messaging, there are generally application choices and tradeoffs that can be made for more performance instead of more eye candy. The average office worker (which I guess includes most people here) will not materially care that it doesn't look quite as good. Another upside is that as a company you don't need such an enormous budget for office hardware.

AtlasBarfed 6y ago

If the M1 is a close derivative architecture of high efficiency mobile scaled to desktop, would scaled up Android processors really be that far away from M1? I know Apple A-series had some edge over Snapdragons, but was the gap between Android processors that big? There are many hundred billion dollar companies motivated to close the gap. It may not be this year.

ampdepolymerase 6y ago
There is the Microsoft SQ2 and the Snapdragon 8cx Gen 2.
frankchn 6y ago
The A14 is significantly faster than the Kirin 9000 (Cortex A77 on TSMC 5nm) on SPECint and SPECfp: https://www.anandtech.com/show/16226/apple-silicon-m1-a14-deep-dive/3 https://www.anandtech.com/show/16226/apple-silicon-m1-a14-de...
hachari 6y ago
“I know Apple A-series had some edge over Snapdragons, but was the gap between Android processors that big?” Absolutely
- AtlasBarfed 6y ago
  Remember there is the 5nm advantage.
  - sbuk 6y ago
    What about the other A-series processors? Excluding the A14, they all used the same process as other ARM based processors, but also have a big performance advantage.

pjmlp 6y ago

Well, PC clones only happened thanks to Compaq. Personally I would have been happier with Atari or Amiga instead, targeted mobile dev back in the J2ME/Symbian days, I did had access to console devkits for a while, so for what I care everything is good as it is.

samvher 6y ago

I share your view of the likely future, but unfortunately it seems very unlikely to me that the open source community is going to unite and prevent this in any significant way. What's happening might give a boost to open source hardware and that would be wonderful, but I don't think this hardware will have much of a chance of being economically competitive and capturing a market.

unionpivo 6y ago

Ehh I think people are overestimating M1 impact. Sure its the best laptop/ipad(other pad) chip out there. And it (and it's next versions) will likely remain best for quite a while. But it's not that the rest of the chips out there turned into garbage. They are still very usable. Personally I have gone away from laptops back to desktops. Having desktop and a phone/tablet combination for me is better than just laptop or laptop + phone/tablet. I still have my laptop, but I very rarely use it now. (mostly if I travel, which I don't because of covid :) ) But having latest gratest laptop is not anywhere near the top of my purchasing priority. And on desktops, latest AMD chips and lots of fast ram, are quite decent. Bottom line even if OS never has anything that matches M1, it will still be fine.

iSnow 6y ago
OSS for operating systems is unlikely to be fine: out of the four classes server, desktop, laptop, phone/tables, one is already fully locked down and with the M1, the laptop class also becomes less open and hackable. Most people - unlike you - don't care that much about desktops, they either have laptops or tablets.

AntiImperialist 6y ago

Apple has the same sort of ownership of M1 as it did with Thunderbolt, maybe a little bit more but not much more than you'd think. Let me explain. Apple has the performance edge now because it purchased ALL of the 5nm chip fabrication capacity of TSMC. This does not mean: - they will be able to do so forever - someone else won't be doing 5nm in the next couple of years So, the same sort of performance will come to other devices in the near future. The best Apple can hope to do at this time is to get enough attention to get some new users and hopefully lock them in into their cozy eco-system. Most importantly, most users don't care about an expensive powerful devices. Highest selling consumer devices are still budget devices and will continue to be so for the foreseeable future. Servers on the other hand will continue to demand powerful hardware, and Apple isn't doing anything there.