Apple expected to invest in Arm ahead of possible September IPO

2»

Comments

  • Reply 21 of 32
    gatorguygatorguy Posts: 24,096member
    gatorguy said:
    danox said:
    williamh said:
    avon b7 said:
    NYC362 said:
    lorca2770 said:

    Please, correct me.
    I have been following Apple since 1984 (Mac 128 ߘꩦlt;/p>

    Was it not S Job who at some point, during the tough times, sold ARM (since it was Apple’s company)? I think I understood it was Apple Reduced Memory. It was one of those developments that were undermined by Microsoft with the permanent: “stop the presses we are coming with something better”. So many Apple developments were drowned by, at the time, all powerful MS.
    Please advise.

    I can't find anything online about Steve Jobs/Apple having owned ARM in the past.  ARM stands for Advanced RISC Machine.  
    Apple owned shares in ARM. Prior to that it was involved in Acorn Computer. 

    It sold off a chunk of its ARM shares around 20 years ago. 
    NYC362 is correct and it was a large part of ARM.  Apple didn't sell because of any problem with ARM (they used the chips since the Newton, if not earlier) but rather because the company needed the cash.

    But they probably did not sell any of the patents, that they created when they owned part of the ARM triplets. I don’t think the processor that Apple used in the Newton was an off the shelf chip.

    In short, I don’t think Apple sold off the intellectual property (patents) that led to the creation of that chip that was used in the Newton or the Newton itself...
    What's the lifespan of patent protection? :) 
    Longer than the lifespan of the Newton. :wink:
    ROTFL!
    watto_cobra
  • Reply 22 of 32
    jrg_ukjrg_uk Posts: 64member
    tham said:


    ARM is/was a British company.  In fact, it started as a contest on show on BBC (or was it iTV) in the late 70s on how to make a low powered computer chip.

    Acorn Computer were involved in bidding for a microcomputer design to be used on a BBC educational tv show in the early 80s, and their design was eventually picked: that became the BBC Microcomputer, which later sold to lot of UK schools, and also led to Acorn’s designing a RISC chip (the Acorn RISC Machine, of which v2 went into the Acorn Archimedes models).

    Some other posts have correctly explained how ARM Ltd was later created, and how Apple’s investment in ARM came to be. Though Wikipedia is probably a better place to read the full story, assuming someone isn’t actively messing with it.
    watto_cobra
  • Reply 23 of 32
    blastdoorblastdoor Posts: 3,217member
    avon b7 said:
    tmay said:
    tham said:
    lorca2770 said:

    Please, correct me.
    I have been following Apple since 1984 (Mac 128 ߘꩦlt;/p>

    Was it not S Jobs who at some point, during the tough times, sold ARM (since it was Apple’s company)? I think I understood it was Apple Reduced Memory. It was one of those developments that were undermined by Microsoft with the permanent: “stop the presses we are coming with something better”. So many Apple developments were drowned by, at the time, all powerful MS.
    Please advise.

    Easy mistake.  The answer is no.  You are getting AIM mixed up with ARM.   AIM (Apple-IBM-Motorola) partnered to design the PowerPC chip.  

    ARM is/was a British company.  In fact, it started as a contest on show on BBC (or was it iTV) in the late 70s on how to make a low powered computer chip.

    Personally, I think Apple will eventually end up switching to RISC-V chip.  In fact, if they're not already developing a chip...they are...for sure.  10 years, they will come out with their own design.
    I don't see Apple doing more than dabbling in RISC-V, at least as long as they have a Fab in TSMC that is building its leading edge on ARM architecture, supported by the design tools that currently favor ARM, and ARM architecture that steadily evolves.

    There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple.
    I think you are underestimating how fast things are moving on the RISC-V front and the direction. 

    https://wccftech.com/jim-keller-tenstorrent-wants-to-compete-with-nvidia-ai-gpus-using-risc-v-based-ai-cpus/


    The thing that’s interesting about that link isn’t RISC-V, it’s Jim Keller thinking that a cpu can beat a GPU at AI workloads. But it’s a very interesting article— thanks for link!
    muthuk_vanalingamwatto_cobra
  • Reply 24 of 32
    tmaytmay Posts: 6,249member
    blastdoor said:
    avon b7 said:
    tmay said:
    tham said:
    lorca2770 said:

    Please, correct me.
    I have been following Apple since 1984 (Mac 128 ߘꩦlt;/p>

    Was it not S Jobs who at some point, during the tough times, sold ARM (since it was Apple’s company)? I think I understood it was Apple Reduced Memory. It was one of those developments that were undermined by Microsoft with the permanent: “stop the presses we are coming with something better”. So many Apple developments were drowned by, at the time, all powerful MS.
    Please advise.

    Easy mistake.  The answer is no.  You are getting AIM mixed up with ARM.   AIM (Apple-IBM-Motorola) partnered to design the PowerPC chip.  

    ARM is/was a British company.  In fact, it started as a contest on show on BBC (or was it iTV) in the late 70s on how to make a low powered computer chip.

    Personally, I think Apple will eventually end up switching to RISC-V chip.  In fact, if they're not already developing a chip...they are...for sure.  10 years, they will come out with their own design.
    I don't see Apple doing more than dabbling in RISC-V, at least as long as they have a Fab in TSMC that is building its leading edge on ARM architecture, supported by the design tools that currently favor ARM, and ARM architecture that steadily evolves.

    There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple.
    I think you are underestimating how fast things are moving on the RISC-V front and the direction. 

    https://wccftech.com/jim-keller-tenstorrent-wants-to-compete-with-nvidia-ai-gpus-using-risc-v-based-ai-cpus/


    The thing that’s interesting about that link isn’t RISC-V, it’s Jim Keller thinking that a cpu can beat a GPU at AI workloads. But it’s a very interesting article— thanks for link!
    Yeah. RISC-V, according to Jim, is new, so less cruft, and open source, so it can evolve "more efficiently", but it isn't really a necessity to his AI processor. In an interview, he noted that there are very few instructions, something on the order of six or eight, that are used 80% of the time.

    He had a great interview with EE/Times which is on YouTube, along with many others that can be found there.

    https://www.youtube.com/watch?v=_MrGNlXRi9M
    edited August 2023 watto_cobra
  • Reply 25 of 32
    tmaytmay Posts: 6,249member
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    tham said:
    lorca2770 said:

    Please, correct me.
    I have been following Apple since 1984 (Mac 128 ߘꩦlt;/p>

    Was it not S Jobs who at some point, during the tough times, sold ARM (since it was Apple’s company)? I think I understood it was Apple Reduced Memory. It was one of those developments that were undermined by Microsoft with the permanent: “stop the presses we are coming with something better”. So many Apple developments were drowned by, at the time, all powerful MS.
    Please advise.

    Easy mistake.  The answer is no.  You are getting AIM mixed up with ARM.   AIM (Apple-IBM-Motorola) partnered to design the PowerPC chip.  

    ARM is/was a British company.  In fact, it started as a contest on show on BBC (or was it iTV) in the late 70s on how to make a low powered computer chip.

    Personally, I think Apple will eventually end up switching to RISC-V chip.  In fact, if they're not already developing a chip...they are...for sure.  10 years, they will come out with their own design.
    I don't see Apple doing more than dabbling in RISC-V, at least as long as they have a Fab in TSMC that is building its leading edge on ARM architecture, supported by the design tools that currently favor ARM, and ARM architecture that steadily evolves.

    There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple.
    I think you are underestimating how fast things are moving on the RISC-V front and the direction. 

    https://wccftech.com/jim-keller-tenstorrent-wants-to-compete-with-nvidia-ai-gpus-using-risc-v-based-ai-cpus/


    "various arcane compute systems that aren't mainstream"

    If Jim Keller can create an AI CPU architecture with RISC-V, and take that mainstream, then kudos to him. So then the question becomes, would Nvidia be able develop a competitor AI CPU with ARM architecture, or would they too go down the RISC-V path.

    Remember, Nvidia is a $1.1 T company, so virtually infinite resources at their disposal, and the ARM ecosystem is very mature and robust compared to RISC-V.


    Everyone is making breakthroughs in efficiencies of one sort or another. 

    https://www.pcmag.com/news/nvidia-claims-breakthrough-in-chip-production-speed

    It will come down to many factors and cost is obviously one of them, be it direct or indirect. 

    It's not a case of this or that either. Systems can be mixed:

    https://semiengineering.com/risc-v-open-platform-for-next-gen-automotive-ecus-eth-zurich-huawei/

    Things have changed a lot for RISC-V in the last five years. It is still being used for IoT, microcontrollers etc but more and more designs and ideas are coming to market and I wouldn't bet anything against another big leap five years from now. 

    Here’s another AI design from a couple of years ago:

    https://spectrum.ieee.org/risc-v-ai

    The HPC goals of the EU are tied in with RISC-V too. 

    https://www.european-processor-initiative.eu/accelerator/
    I'm fine with you hawking RISC-V, but realistically, ARM isn't going to be displaced very rapidly, if at all, in the marketplace when companies that are perfectly fine with ARM are generating the bulk of revenue for TSMC. Both Intel and Samsung would have that opportunity as well with RISC-V, should they decide to compete in the consumer market, but I'm not seeing that either.

    Of course, Jim Keller might change that equation, but even then, there doesn't seem to be any technical reason that ARM with ongoing evolution, can't compete, so best watch Nvidia on that front.
    Well, I wasn't hawking on about anything. Simply pointing out a different reality to the one you painted. I didn't even go into your claim that open source is irrelevant to the players you mentioned.

    Cough, Darwin, cough!

    Nor the impact of weaponisation of IP by the US government. 

    RISC-V is growing at an unprecedented rate. 

    https://linuxfoundation.eu/newsroom/rise-project-launches-to-accelerate-development-of-risc-v

    Look at the names making up the consortium. 

    Given the "unprecedented" growth that ARM saw with the smartphone, holding 90% market share, and broad growth thereafter, I'm not sure how "unprecedented" the growth of RISC-V is in comparison, so I'd refrain from superlatives at this point in time. 

    Personally, I'd give credit to fabless foundries, primarily TSMC, for de-linking integrated ISA's and design on mainstream processors. That's the real power behind ARM and RISC-V.
    edited August 2023 williamlondonwatto_cobra
  • Reply 26 of 32
    avon b7avon b7 Posts: 7,520member
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    tham said:
    lorca2770 said:

    Please, correct me.
    I have been following Apple since 1984 (Mac 128 ߘꩦlt;/p>

    Was it not S Jobs who at some point, during the tough times, sold ARM (since it was Apple’s company)? I think I understood it was Apple Reduced Memory. It was one of those developments that were undermined by Microsoft with the permanent: “stop the presses we are coming with something better”. So many Apple developments were drowned by, at the time, all powerful MS.
    Please advise.

    Easy mistake.  The answer is no.  You are getting AIM mixed up with ARM.   AIM (Apple-IBM-Motorola) partnered to design the PowerPC chip.  

    ARM is/was a British company.  In fact, it started as a contest on show on BBC (or was it iTV) in the late 70s on how to make a low powered computer chip.

    Personally, I think Apple will eventually end up switching to RISC-V chip.  In fact, if they're not already developing a chip...they are...for sure.  10 years, they will come out with their own design.
    I don't see Apple doing more than dabbling in RISC-V, at least as long as they have a Fab in TSMC that is building its leading edge on ARM architecture, supported by the design tools that currently favor ARM, and ARM architecture that steadily evolves.

    There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple.
    I think you are underestimating how fast things are moving on the RISC-V front and the direction. 

    https://wccftech.com/jim-keller-tenstorrent-wants-to-compete-with-nvidia-ai-gpus-using-risc-v-based-ai-cpus/


    "various arcane compute systems that aren't mainstream"

    If Jim Keller can create an AI CPU architecture with RISC-V, and take that mainstream, then kudos to him. So then the question becomes, would Nvidia be able develop a competitor AI CPU with ARM architecture, or would they too go down the RISC-V path.

    Remember, Nvidia is a $1.1 T company, so virtually infinite resources at their disposal, and the ARM ecosystem is very mature and robust compared to RISC-V.


    Everyone is making breakthroughs in efficiencies of one sort or another. 

    https://www.pcmag.com/news/nvidia-claims-breakthrough-in-chip-production-speed

    It will come down to many factors and cost is obviously one of them, be it direct or indirect. 

    It's not a case of this or that either. Systems can be mixed:

    https://semiengineering.com/risc-v-open-platform-for-next-gen-automotive-ecus-eth-zurich-huawei/

    Things have changed a lot for RISC-V in the last five years. It is still being used for IoT, microcontrollers etc but more and more designs and ideas are coming to market and I wouldn't bet anything against another big leap five years from now. 

    Here’s another AI design from a couple of years ago:

    https://spectrum.ieee.org/risc-v-ai

    The HPC goals of the EU are tied in with RISC-V too. 

    https://www.european-processor-initiative.eu/accelerator/
    I'm fine with you hawking RISC-V, but realistically, ARM isn't going to be displaced very rapidly, if at all, in the marketplace when companies that are perfectly fine with ARM are generating the bulk of revenue for TSMC. Both Intel and Samsung would have that opportunity as well with RISC-V, should they decide to compete in the consumer market, but I'm not seeing that either.

    Of course, Jim Keller might change that equation, but even then, there doesn't seem to be any technical reason that ARM with ongoing evolution, can't compete, so best watch Nvidia on that front.
    Well, I wasn't hawking on about anything. Simply pointing out a different reality to the one you painted. I didn't even go into your claim that open source is irrelevant to the players you mentioned.

    Cough, Darwin, cough!

    Nor the impact of weaponisation of IP by the US government. 

    RISC-V is growing at an unprecedented rate. 

    https://linuxfoundation.eu/newsroom/rise-project-launches-to-accelerate-development-of-risc-v

    Look at the names making up the consortium. 

    Given the "unprecedented" growth that ARM saw with the smartphone, holding 90% market share, and broad growth thereafter, I'm not sure how "unprecedented" the growth of RISC-V is in comparison, so I'd refrain from superlatives at this point in time. 

    Personally, I'd give credit to fabless foundries, primarily TSMC, for de-linking integrated ISA's and design on mainstream processors. That's the real power behind ARM and RISC-V.
    Fabless foundries?

    I think you're crossing your cables. 

    The superlatives are fine if you look at RISC-V developments over the last few years, although that was not my point.

    All I pointed out is that RISC-V is far, far, far more than you painted it to be. 
  • Reply 27 of 32
    tmaytmay Posts: 6,249member
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    tham said:
    lorca2770 said:

    Please, correct me.
    I have been following Apple since 1984 (Mac 128 ߘꩦlt;/p>

    Was it not S Jobs who at some point, during the tough times, sold ARM (since it was Apple’s company)? I think I understood it was Apple Reduced Memory. It was one of those developments that were undermined by Microsoft with the permanent: “stop the presses we are coming with something better”. So many Apple developments were drowned by, at the time, all powerful MS.
    Please advise.

    Easy mistake.  The answer is no.  You are getting AIM mixed up with ARM.   AIM (Apple-IBM-Motorola) partnered to design the PowerPC chip.  

    ARM is/was a British company.  In fact, it started as a contest on show on BBC (or was it iTV) in the late 70s on how to make a low powered computer chip.

    Personally, I think Apple will eventually end up switching to RISC-V chip.  In fact, if they're not already developing a chip...they are...for sure.  10 years, they will come out with their own design.
    I don't see Apple doing more than dabbling in RISC-V, at least as long as they have a Fab in TSMC that is building its leading edge on ARM architecture, supported by the design tools that currently favor ARM, and ARM architecture that steadily evolves.

    There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple.
    I think you are underestimating how fast things are moving on the RISC-V front and the direction. 

    https://wccftech.com/jim-keller-tenstorrent-wants-to-compete-with-nvidia-ai-gpus-using-risc-v-based-ai-cpus/


    "various arcane compute systems that aren't mainstream"

    If Jim Keller can create an AI CPU architecture with RISC-V, and take that mainstream, then kudos to him. So then the question becomes, would Nvidia be able develop a competitor AI CPU with ARM architecture, or would they too go down the RISC-V path.

    Remember, Nvidia is a $1.1 T company, so virtually infinite resources at their disposal, and the ARM ecosystem is very mature and robust compared to RISC-V.


    Everyone is making breakthroughs in efficiencies of one sort or another. 

    https://www.pcmag.com/news/nvidia-claims-breakthrough-in-chip-production-speed

    It will come down to many factors and cost is obviously one of them, be it direct or indirect. 

    It's not a case of this or that either. Systems can be mixed:

    https://semiengineering.com/risc-v-open-platform-for-next-gen-automotive-ecus-eth-zurich-huawei/

    Things have changed a lot for RISC-V in the last five years. It is still being used for IoT, microcontrollers etc but more and more designs and ideas are coming to market and I wouldn't bet anything against another big leap five years from now. 

    Here’s another AI design from a couple of years ago:

    https://spectrum.ieee.org/risc-v-ai

    The HPC goals of the EU are tied in with RISC-V too. 

    https://www.european-processor-initiative.eu/accelerator/
    I'm fine with you hawking RISC-V, but realistically, ARM isn't going to be displaced very rapidly, if at all, in the marketplace when companies that are perfectly fine with ARM are generating the bulk of revenue for TSMC. Both Intel and Samsung would have that opportunity as well with RISC-V, should they decide to compete in the consumer market, but I'm not seeing that either.

    Of course, Jim Keller might change that equation, but even then, there doesn't seem to be any technical reason that ARM with ongoing evolution, can't compete, so best watch Nvidia on that front.
    Well, I wasn't hawking on about anything. Simply pointing out a different reality to the one you painted. I didn't even go into your claim that open source is irrelevant to the players you mentioned.

    Cough, Darwin, cough!

    Nor the impact of weaponisation of IP by the US government. 

    RISC-V is growing at an unprecedented rate. 

    https://linuxfoundation.eu/newsroom/rise-project-launches-to-accelerate-development-of-risc-v

    Look at the names making up the consortium. 

    Given the "unprecedented" growth that ARM saw with the smartphone, holding 90% market share, and broad growth thereafter, I'm not sure how "unprecedented" the growth of RISC-V is in comparison, so I'd refrain from superlatives at this point in time. 

    Personally, I'd give credit to fabless foundries, primarily TSMC, for de-linking integrated ISA's and design on mainstream processors. That's the real power behind ARM and RISC-V.
    Fabless foundries?

    I think you're crossing your cables. 

    The superlatives are fine if you look at RISC-V developments over the last few years, although that was not -my point.

    All I pointed out is that RISC-V is far, far, far more than you painted it to be. 
    TSMC is a semiconductor foundry, the customers are "fabless". Would you prefer fabless semiconductor company?

    I haven't painted it to be anything but another RISC ISA, that is relatively new, so cruft free, and open source, for whatever the advantages that brings. It doesn't do anything that can't be done with a "licensed" ISA, and its "free".

    If has advantages, but at the same time, the instruction set is being broadened by the users, and ultimately, given time, will need to be replaced by yet another "new" RISC ISA .
    williamlondonwatto_cobra
  • Reply 28 of 32
    avon b7avon b7 Posts: 7,520member
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    tham said:
    lorca2770 said:

    Please, correct me.
    I have been following Apple since 1984 (Mac 128 ߘꩦlt;/p>

    Was it not S Jobs who at some point, during the tough times, sold ARM (since it was Apple’s company)? I think I understood it was Apple Reduced Memory. It was one of those developments that were undermined by Microsoft with the permanent: “stop the presses we are coming with something better”. So many Apple developments were drowned by, at the time, all powerful MS.
    Please advise.

    Easy mistake.  The answer is no.  You are getting AIM mixed up with ARM.   AIM (Apple-IBM-Motorola) partnered to design the PowerPC chip.  

    ARM is/was a British company.  In fact, it started as a contest on show on BBC (or was it iTV) in the late 70s on how to make a low powered computer chip.

    Personally, I think Apple will eventually end up switching to RISC-V chip.  In fact, if they're not already developing a chip...they are...for sure.  10 years, they will come out with their own design.
    I don't see Apple doing more than dabbling in RISC-V, at least as long as they have a Fab in TSMC that is building its leading edge on ARM architecture, supported by the design tools that currently favor ARM, and ARM architecture that steadily evolves.

    There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple.
    I think you are underestimating how fast things are moving on the RISC-V front and the direction. 

    https://wccftech.com/jim-keller-tenstorrent-wants-to-compete-with-nvidia-ai-gpus-using-risc-v-based-ai-cpus/


    "various arcane compute systems that aren't mainstream"

    If Jim Keller can create an AI CPU architecture with RISC-V, and take that mainstream, then kudos to him. So then the question becomes, would Nvidia be able develop a competitor AI CPU with ARM architecture, or would they too go down the RISC-V path.

    Remember, Nvidia is a $1.1 T company, so virtually infinite resources at their disposal, and the ARM ecosystem is very mature and robust compared to RISC-V.


    Everyone is making breakthroughs in efficiencies of one sort or another. 

    https://www.pcmag.com/news/nvidia-claims-breakthrough-in-chip-production-speed

    It will come down to many factors and cost is obviously one of them, be it direct or indirect. 

    It's not a case of this or that either. Systems can be mixed:

    https://semiengineering.com/risc-v-open-platform-for-next-gen-automotive-ecus-eth-zurich-huawei/

    Things have changed a lot for RISC-V in the last five years. It is still being used for IoT, microcontrollers etc but more and more designs and ideas are coming to market and I wouldn't bet anything against another big leap five years from now. 

    Here’s another AI design from a couple of years ago:

    https://spectrum.ieee.org/risc-v-ai

    The HPC goals of the EU are tied in with RISC-V too. 

    https://www.european-processor-initiative.eu/accelerator/
    I'm fine with you hawking RISC-V, but realistically, ARM isn't going to be displaced very rapidly, if at all, in the marketplace when companies that are perfectly fine with ARM are generating the bulk of revenue for TSMC. Both Intel and Samsung would have that opportunity as well with RISC-V, should they decide to compete in the consumer market, but I'm not seeing that either.

    Of course, Jim Keller might change that equation, but even then, there doesn't seem to be any technical reason that ARM with ongoing evolution, can't compete, so best watch Nvidia on that front.
    Well, I wasn't hawking on about anything. Simply pointing out a different reality to the one you painted. I didn't even go into your claim that open source is irrelevant to the players you mentioned.

    Cough, Darwin, cough!

    Nor the impact of weaponisation of IP by the US government. 

    RISC-V is growing at an unprecedented rate. 

    https://linuxfoundation.eu/newsroom/rise-project-launches-to-accelerate-development-of-risc-v

    Look at the names making up the consortium. 

    Given the "unprecedented" growth that ARM saw with the smartphone, holding 90% market share, and broad growth thereafter, I'm not sure how "unprecedented" the growth of RISC-V is in comparison, so I'd refrain from superlatives at this point in time. 

    Personally, I'd give credit to fabless foundries, primarily TSMC, for de-linking integrated ISA's and design on mainstream processors. That's the real power behind ARM and RISC-V.
    Fabless foundries?

    I think you're crossing your cables. 

    The superlatives are fine if you look at RISC-V developments over the last few years, although that was not -my point.

    All I pointed out is that RISC-V is far, far, far more than you painted it to be. 
    TSMC is a semiconductor foundry, the customers are "fabless". Would you prefer fabless semiconductor company?

    I haven't painted it to be anything but another RISC ISA, that is relatively new, so cruft free, and open source, for whatever the advantages that brings. It doesn't do anything that can't be done with a "licensed" ISA, and its "free".

    If has advantages, but at the same time, the instruction set is being broadened by the users, and ultimately, given time, will need to be replaced by yet another "new" RISC ISA .
    A 'fabless foundry' cannot exist because a foundry is a fab, so anything would be better than what you said but, like I said, I can understand that your cables got crossed. We all do it sometimes.

    What you painted was this (literally):

    "There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple."

    That isn't a good representation of where RISC-V is today:

    " Our vision for 2022 was to bring RISC-V everywhere, and that has really come to fruition with RISC-V adoption and development in everything from automotive to aerospace to the data center, as well as consumer devices" 

    https://riscv.org/announcements/2022/12/risc-v-sees-significant-growth-and-technical-progress-in-2022-with-billions-of-risc-v-cores-in-market/#:~:text=The RISC-V community has,in the next 120 days.




    edited August 2023
  • Reply 29 of 32
    "Apple and other major chip firms may want a piece of the action."


    You just wait three or four years until the M5 processor comes out. That'll be terrifying. It'll lock you out and call you "nonessential personnel." Although it would be a lot more fun if Apple updates Siri then with a male voice that has clipped, staccato diction.
    williamlondonwatto_cobraking editor the grate
  • Reply 30 of 32
    tmaytmay Posts: 6,249member
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    tham said:
    lorca2770 said:

    Please, correct me.
    I have been following Apple since 1984 (Mac 128 ߘꩦlt;/p>

    Was it not S Jobs who at some point, during the tough times, sold ARM (since it was Apple’s company)? I think I understood it was Apple Reduced Memory. It was one of those developments that were undermined by Microsoft with the permanent: “stop the presses we are coming with something better”. So many Apple developments were drowned by, at the time, all powerful MS.
    Please advise.

    Easy mistake.  The answer is no.  You are getting AIM mixed up with ARM.   AIM (Apple-IBM-Motorola) partnered to design the PowerPC chip.  

    ARM is/was a British company.  In fact, it started as a contest on show on BBC (or was it iTV) in the late 70s on how to make a low powered computer chip.

    Personally, I think Apple will eventually end up switching to RISC-V chip.  In fact, if they're not already developing a chip...they are...for sure.  10 years, they will come out with their own design.
    I don't see Apple doing more than dabbling in RISC-V, at least as long as they have a Fab in TSMC that is building its leading edge on ARM architecture, supported by the design tools that currently favor ARM, and ARM architecture that steadily evolves.

    There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple.
    I think you are underestimating how fast things are moving on the RISC-V front and the direction. 

    https://wccftech.com/jim-keller-tenstorrent-wants-to-compete-with-nvidia-ai-gpus-using-risc-v-based-ai-cpus/


    "various arcane compute systems that aren't mainstream"

    If Jim Keller can create an AI CPU architecture with RISC-V, and take that mainstream, then kudos to him. So then the question becomes, would Nvidia be able develop a competitor AI CPU with ARM architecture, or would they too go down the RISC-V path.

    Remember, Nvidia is a $1.1 T company, so virtually infinite resources at their disposal, and the ARM ecosystem is very mature and robust compared to RISC-V.


    Everyone is making breakthroughs in efficiencies of one sort or another. 

    https://www.pcmag.com/news/nvidia-claims-breakthrough-in-chip-production-speed

    It will come down to many factors and cost is obviously one of them, be it direct or indirect. 

    It's not a case of this or that either. Systems can be mixed:

    https://semiengineering.com/risc-v-open-platform-for-next-gen-automotive-ecus-eth-zurich-huawei/

    Things have changed a lot for RISC-V in the last five years. It is still being used for IoT, microcontrollers etc but more and more designs and ideas are coming to market and I wouldn't bet anything against another big leap five years from now. 

    Here’s another AI design from a couple of years ago:

    https://spectrum.ieee.org/risc-v-ai

    The HPC goals of the EU are tied in with RISC-V too. 

    https://www.european-processor-initiative.eu/accelerator/
    I'm fine with you hawking RISC-V, but realistically, ARM isn't going to be displaced very rapidly, if at all, in the marketplace when companies that are perfectly fine with ARM are generating the bulk of revenue for TSMC. Both Intel and Samsung would have that opportunity as well with RISC-V, should they decide to compete in the consumer market, but I'm not seeing that either.

    Of course, Jim Keller might change that equation, but even then, there doesn't seem to be any technical reason that ARM with ongoing evolution, can't compete, so best watch Nvidia on that front.
    Well, I wasn't hawking on about anything. Simply pointing out a different reality to the one you painted. I didn't even go into your claim that open source is irrelevant to the players you mentioned.

    Cough, Darwin, cough!

    Nor the impact of weaponisation of IP by the US government. 

    RISC-V is growing at an unprecedented rate. 

    https://linuxfoundation.eu/newsroom/rise-project-launches-to-accelerate-development-of-risc-v

    Look at the names making up the consortium. 

    Given the "unprecedented" growth that ARM saw with the smartphone, holding 90% market share, and broad growth thereafter, I'm not sure how "unprecedented" the growth of RISC-V is in comparison, so I'd refrain from superlatives at this point in time. 

    Personally, I'd give credit to fabless foundries, primarily TSMC, for de-linking integrated ISA's and design on mainstream processors. That's the real power behind ARM and RISC-V.
    Fabless foundries?

    I think you're crossing your cables. 

    The superlatives are fine if you look at RISC-V developments over the last few years, although that was not -my point.

    All I pointed out is that RISC-V is far, far, far more than you painted it to be. 
    TSMC is a semiconductor foundry, the customers are "fabless". Would you prefer fabless semiconductor company?

    I haven't painted it to be anything but another RISC ISA, that is relatively new, so cruft free, and open source, for whatever the advantages that brings. It doesn't do anything that can't be done with a "licensed" ISA, and its "free".

    If has advantages, but at the same time, the instruction set is being broadened by the users, and ultimately, given time, will need to be replaced by yet another "new" RISC ISA .
    A 'fabless foundry' cannot exist because a foundry is a fab, so anything would be better than what you said but, like I said, I can understand that your cables got crossed. We all do it sometimes.

    What you painted was this (literally):

    "There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple."

    That isn't a good representation of where RISC-V is today:

    " Our vision for 2022 was to bring RISC-V everywhere, and that has really come to fruition with RISC-V adoption and development in everything from automotive to aerospace to the data center, as well as consumer devices" 

    https://riscv.org/announcements/2022/12/risc-v-sees-significant-growth-and-technical-progress-in-2022-with-billions-of-risc-v-cores-in-market/#:~:text=The RISC-V community has,in the next 120 days.




    It's wonderful that RISC-V is catching the attention of 
    williamlondonwatto_cobra
  • Reply 31 of 32
    avon b7avon b7 Posts: 7,520member
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    tham said:
    lorca2770 said:

    Please, correct me.
    I have been following Apple since 1984 (Mac 128 ߘꩦlt;/p>

    Was it not S Jobs who at some point, during the tough times, sold ARM (since it was Apple’s company)? I think I understood it was Apple Reduced Memory. It was one of those developments that were undermined by Microsoft with the permanent: “stop the presses we are coming with something better”. So many Apple developments were drowned by, at the time, all powerful MS.
    Please advise.

    Easy mistake.  The answer is no.  You are getting AIM mixed up with ARM.   AIM (Apple-IBM-Motorola) partnered to design the PowerPC chip.  

    ARM is/was a British company.  In fact, it started as a contest on show on BBC (or was it iTV) in the late 70s on how to make a low powered computer chip.

    Personally, I think Apple will eventually end up switching to RISC-V chip.  In fact, if they're not already developing a chip...they are...for sure.  10 years, they will come out with their own design.
    I don't see Apple doing more than dabbling in RISC-V, at least as long as they have a Fab in TSMC that is building its leading edge on ARM architecture, supported by the design tools that currently favor ARM, and ARM architecture that steadily evolves.

    There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple.
    I think you are underestimating how fast things are moving on the RISC-V front and the direction. 

    https://wccftech.com/jim-keller-tenstorrent-wants-to-compete-with-nvidia-ai-gpus-using-risc-v-based-ai-cpus/


    "various arcane compute systems that aren't mainstream"

    If Jim Keller can create an AI CPU architecture with RISC-V, and take that mainstream, then kudos to him. So then the question becomes, would Nvidia be able develop a competitor AI CPU with ARM architecture, or would they too go down the RISC-V path.

    Remember, Nvidia is a $1.1 T company, so virtually infinite resources at their disposal, and the ARM ecosystem is very mature and robust compared to RISC-V.


    Everyone is making breakthroughs in efficiencies of one sort or another. 

    https://www.pcmag.com/news/nvidia-claims-breakthrough-in-chip-production-speed

    It will come down to many factors and cost is obviously one of them, be it direct or indirect. 

    It's not a case of this or that either. Systems can be mixed:

    https://semiengineering.com/risc-v-open-platform-for-next-gen-automotive-ecus-eth-zurich-huawei/

    Things have changed a lot for RISC-V in the last five years. It is still being used for IoT, microcontrollers etc but more and more designs and ideas are coming to market and I wouldn't bet anything against another big leap five years from now. 

    Here’s another AI design from a couple of years ago:

    https://spectrum.ieee.org/risc-v-ai

    The HPC goals of the EU are tied in with RISC-V too. 

    https://www.european-processor-initiative.eu/accelerator/
    I'm fine with you hawking RISC-V, but realistically, ARM isn't going to be displaced very rapidly, if at all, in the marketplace when companies that are perfectly fine with ARM are generating the bulk of revenue for TSMC. Both Intel and Samsung would have that opportunity as well with RISC-V, should they decide to compete in the consumer market, but I'm not seeing that either.

    Of course, Jim Keller might change that equation, but even then, there doesn't seem to be any technical reason that ARM with ongoing evolution, can't compete, so best watch Nvidia on that front.
    Well, I wasn't hawking on about anything. Simply pointing out a different reality to the one you painted. I didn't even go into your claim that open source is irrelevant to the players you mentioned.

    Cough, Darwin, cough!

    Nor the impact of weaponisation of IP by the US government. 

    RISC-V is growing at an unprecedented rate. 

    https://linuxfoundation.eu/newsroom/rise-project-launches-to-accelerate-development-of-risc-v

    Look at the names making up the consortium. 

    Given the "unprecedented" growth that ARM saw with the smartphone, holding 90% market share, and broad growth thereafter, I'm not sure how "unprecedented" the growth of RISC-V is in comparison, so I'd refrain from superlatives at this point in time. 

    Personally, I'd give credit to fabless foundries, primarily TSMC, for de-linking integrated ISA's and design on mainstream processors. That's the real power behind ARM and RISC-V.
    Fabless foundries?

    I think you're crossing your cables. 

    The superlatives are fine if you look at RISC-V developments over the last few years, although that was not -my point.

    All I pointed out is that RISC-V is far, far, far more than you painted it to be. 
    TSMC is a semiconductor foundry, the customers are "fabless". Would you prefer fabless semiconductor company?

    I haven't painted it to be anything but another RISC ISA, that is relatively new, so cruft free, and open source, for whatever the advantages that brings. It doesn't do anything that can't be done with a "licensed" ISA, and its "free".

    If has advantages, but at the same time, the instruction set is being broadened by the users, and ultimately, given time, will need to be replaced by yet another "new" RISC ISA .
    A 'fabless foundry' cannot exist because a foundry is a fab, so anything would be better than what you said but, like I said, I can understand that your cables got crossed. We all do it sometimes.

    What you painted was this (literally):

    "There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple."

    That isn't a good representation of where RISC-V is today:

    " Our vision for 2022 was to bring RISC-V everywhere, and that has really come to fruition with RISC-V adoption and development in everything from automotive to aerospace to the data center, as well as consumer devices" 

    https://riscv.org/announcements/2022/12/risc-v-sees-significant-growth-and-technical-progress-in-2022-with-billions-of-risc-v-cores-in-market/#:~:text=The RISC-V community has,in the next 120 days.




    It's wonderful that RISC-V is catching the attention of 
    And now there are rumours of Qualcomm's Hamoa chipsets (for desktop and mobile) moving away from ARM.

    Only sketchy rumours at the moment so time will tell. 
  • Reply 32 of 32
    tmaytmay Posts: 6,249member
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    avon b7 said:
    tmay said:
    tham said:
    lorca2770 said:

    Please, correct me.
    I have been following Apple since 1984 (Mac 128 ߘꩦlt;/p>

    Was it not S Jobs who at some point, during the tough times, sold ARM (since it was Apple’s company)? I think I understood it was Apple Reduced Memory. It was one of those developments that were undermined by Microsoft with the permanent: “stop the presses we are coming with something better”. So many Apple developments were drowned by, at the time, all powerful MS.
    Please advise.

    Easy mistake.  The answer is no.  You are getting AIM mixed up with ARM.   AIM (Apple-IBM-Motorola) partnered to design the PowerPC chip.  

    ARM is/was a British company.  In fact, it started as a contest on show on BBC (or was it iTV) in the late 70s on how to make a low powered computer chip.

    Personally, I think Apple will eventually end up switching to RISC-V chip.  In fact, if they're not already developing a chip...they are...for sure.  10 years, they will come out with their own design.
    I don't see Apple doing more than dabbling in RISC-V, at least as long as they have a Fab in TSMC that is building its leading edge on ARM architecture, supported by the design tools that currently favor ARM, and ARM architecture that steadily evolves.

    There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple.
    I think you are underestimating how fast things are moving on the RISC-V front and the direction. 

    https://wccftech.com/jim-keller-tenstorrent-wants-to-compete-with-nvidia-ai-gpus-using-risc-v-based-ai-cpus/


    "various arcane compute systems that aren't mainstream"

    If Jim Keller can create an AI CPU architecture with RISC-V, and take that mainstream, then kudos to him. So then the question becomes, would Nvidia be able develop a competitor AI CPU with ARM architecture, or would they too go down the RISC-V path.

    Remember, Nvidia is a $1.1 T company, so virtually infinite resources at their disposal, and the ARM ecosystem is very mature and robust compared to RISC-V.


    Everyone is making breakthroughs in efficiencies of one sort or another. 

    https://www.pcmag.com/news/nvidia-claims-breakthrough-in-chip-production-speed

    It will come down to many factors and cost is obviously one of them, be it direct or indirect. 

    It's not a case of this or that either. Systems can be mixed:

    https://semiengineering.com/risc-v-open-platform-for-next-gen-automotive-ecus-eth-zurich-huawei/

    Things have changed a lot for RISC-V in the last five years. It is still being used for IoT, microcontrollers etc but more and more designs and ideas are coming to market and I wouldn't bet anything against another big leap five years from now. 

    Here’s another AI design from a couple of years ago:

    https://spectrum.ieee.org/risc-v-ai

    The HPC goals of the EU are tied in with RISC-V too. 

    https://www.european-processor-initiative.eu/accelerator/
    I'm fine with you hawking RISC-V, but realistically, ARM isn't going to be displaced very rapidly, if at all, in the marketplace when companies that are perfectly fine with ARM are generating the bulk of revenue for TSMC. Both Intel and Samsung would have that opportunity as well with RISC-V, should they decide to compete in the consumer market, but I'm not seeing that either.

    Of course, Jim Keller might change that equation, but even then, there doesn't seem to be any technical reason that ARM with ongoing evolution, can't compete, so best watch Nvidia on that front.
    Well, I wasn't hawking on about anything. Simply pointing out a different reality to the one you painted. I didn't even go into your claim that open source is irrelevant to the players you mentioned.

    Cough, Darwin, cough!

    Nor the impact of weaponisation of IP by the US government. 

    RISC-V is growing at an unprecedented rate. 

    https://linuxfoundation.eu/newsroom/rise-project-launches-to-accelerate-development-of-risc-v

    Look at the names making up the consortium. 

    Given the "unprecedented" growth that ARM saw with the smartphone, holding 90% market share, and broad growth thereafter, I'm not sure how "unprecedented" the growth of RISC-V is in comparison, so I'd refrain from superlatives at this point in time. 

    Personally, I'd give credit to fabless foundries, primarily TSMC, for de-linking integrated ISA's and design on mainstream processors. That's the real power behind ARM and RISC-V.
    Fabless foundries?

    I think you're crossing your cables. 

    The superlatives are fine if you look at RISC-V developments over the last few years, although that was not -my point.

    All I pointed out is that RISC-V is far, far, far more than you painted it to be. 
    TSMC is a semiconductor foundry, the customers are "fabless". Would you prefer fabless semiconductor company?

    I haven't painted it to be anything but another RISC ISA, that is relatively new, so cruft free, and open source, for whatever the advantages that brings. It doesn't do anything that can't be done with a "licensed" ISA, and its "free".

    If has advantages, but at the same time, the instruction set is being broadened by the users, and ultimately, given time, will need to be replaced by yet another "new" RISC ISA .
    A 'fabless foundry' cannot exist because a foundry is a fab, so anything would be better than what you said but, like I said, I can understand that your cables got crossed. We all do it sometimes.

    What you painted was this (literally):

    "There is a movement to RISC-V, but it is primarily IoT, which sets a low bar of entry, and various arcane compute systems that aren't mainstream. Mostly though, it is popular with academics, or regions that are not allowed leading edge ARM licenses or design tools. There isn't any performance or technical advantage to RISC-V, and open source is irrelevant to players such as AMD, Nvidia, Qualcomm, and of course, Apple."

    That isn't a good representation of where RISC-V is today:

    " Our vision for 2022 was to bring RISC-V everywhere, and that has really come to fruition with RISC-V adoption and development in everything from automotive to aerospace to the data center, as well as consumer devices" 

    https://riscv.org/announcements/2022/12/risc-v-sees-significant-growth-and-technical-progress-in-2022-with-billions-of-risc-v-cores-in-market/#:~:text=The RISC-V community has,in the next 120 days.




    It's wonderful that RISC-V is catching the attention of 
    And now there are rumours of Qualcomm's Hamoa chipsets (for desktop and mobile) moving away from ARM.

    Only sketchy rumours at the moment so time will tell. 
    For current ARM licenses, especially the large companies, there isn't enough savings, or performance benefit, switching to RISC-V in the near term, albeit Appe would consider Air Tags as a first step in the future.
    watto_cobra
Sign In or Register to comment.