Anyone can design and sell little circuit blocks and on chip peripherals, even proprietary ones, for use on any chip.
What’s the likelihood of a dominant player emerging and implementing patented, proprietary RISC-V architecture changes which turn out to be necessary for high-performance? And if such a company gains sufficient market share, they could turn RISC-V into basically another x86-64 with many proprietary extensions. Sure, others could create their own RISC-V base processor - but if their performance is 500% lower than processors from the proprietary vendor who would purchase them?
Sure, but not legally so. Patents are the reason there are only really two x86-64 vendors and a company could similarly patent their own RISC-V modifications.
Not really. Reverse engineering has traditionally had limited value because of the exponential growth of silicon and the time it takes to develop anything on a node.
It is impossible to understate how insane CPU speeds are at the present. These are high speeds even for radio frequencies. Traces are no longer required to be connected at these speeds. Capacitive coupling becomes a major aspect of signal transmission; the entire RLC gamut is impactful. Anyone can, and likely every major player does, reverse engineer every major hardware design they can acquire. This will only allow one to understand what is now irrelevant in terms of design. The new chip in question will already be well into the process of amelioration of development costs.
You see, the reason the chip costs so much initially is because of the cost of the node and all of the hardware involved in creating the chip. The tooling costs are immense now. The reason the price of hardware drops over time is because actually producing the chips is dirt cheap by comparison. The initial cost is a projection about how much time it will take to pay back the initial investment assuming the units sell in low volume. As that volume increases and the loans are paid back, the price is dropped to access more segments of the market at lower price points. There is a market saturation balance that must be maintained to keep the next generation viable. Each new product takes 10 years to create. This means a company like Nvidia has that (IIRC (prob not)) has something like an 18 month new major product release schedule Will have the next 6 generations of products already in various phases of planning and design. If you reverse engineer a publicly available product, you are largely reverse engineering the fab’s capabilities while also looking at a product that is 6 generations behind the real cutting edge. Not to mention, you are also never going to be successful releasing your reverse engineered product because you cannot sell it at a competitive initial price required to pay for the required tooling when the original product has already done so and can be sold for cost of manufacture with a small markup.
The enormous difference between the cost of tooling and cost of actually making the chip is absolutely essential to understand in combination with exponential growth. This is what William Shockley realized in the 1950’s and convinced the US government to pursue. The entire modern world we all have known is due to this exponential growth. This growth is absolutely remarkable in human history as the only time a civilian business endeavor has out classed the economic growth created by the largest militaries of the world. This is what Shockley actually realized; that the military budget could not match the growth potential of silicon up to the plateau of potential scaling when physics prevents further scaled nodes. There are only a few new nodes left, and those are the next 10 years. This means we are already at the end of scaling because hardware designers are actually already there. The concept of Venture Capital is actually a pseudo extension of military budgeting constraints in a round about way. Indeed, Silicon Valley is where the real battle of the cold war happened. The Soviet Union and China failed to realize how silicon scaling would alter and become a decisive military operation. Once that ball started rolling, it is impossible to catch the front line so long as the design edge is kept a closely guarded secret and the extreme capital required is too high to be viable.
This is the real reason why your mobile devices are all running orphaned kernels on undocumented hardware. Your need to buy new devices when you are told is the primary factor driving these new nodes. You are less likely to realize that this is ultimately a tax to avoid large scale wars and major conflicts. This is why the changes happening right now in politics are not trivial. Moves are being made that appear similar to the era before venture capital and the Pax-Silicon™. We are already at the end of silicon exponential growth. There is no replacement for the exponential growth of silicon to outstrip military based spending and growth as has been the standard for all of the rest of human history. This will inevitably lead to hording scarcity and conflicts. The enormity of funding will cause the use of militaries to press advantages before they disappear. These are the factors that created world wars of the past and they will arise again. The next era of technology is going to be biology, but we are at least a couple of centuries away from biology as an engineering science where something like a synthetic brain is capable of producing a Turing complete deterministic computer on par with a CPU of the present. There is no clear path to exponential growth either like there was with scaling silicon. Perhaps the software organization, libraries, and database scaling will be an exponential growth factor, but I don’t know how that will have a barrier of entry on par with silicon in a way that is insurmountable by everyone including large governments and coalitions of governments.
So this is the world that is changing around the issue of RISC-V. It will come into its own in a post VC growth era. We are at the end of that growth already. Reverse engineering becomes relevant now and proprietary secretive strategy is no longer of the same magnitude of military significance. My narrative here will become more and more obvious with time. This has been a years long curiosity between many of my interests like why the world prior the the 1950’s was so different, why the USA won the cold war, and understanding the history of the microprocessor to wrap my head around all the peripherals present in an Arduino which was born out of a desire to learn Megasquirt when I was a hotrod car nut ages ago. I abstract across broad spaces well like this and like to simplify complexity because I’m dumb like that.
Things like patents are just weapons of the super rich. They have no real relevance here. The outcome of these cases has nothing to do with justice or right and wrong. These are battlefronts of militaries with convoluted rules of engagement. In the next 10 years this proxy conflict space will be abandoned and everything changes into an unknown state. Likely new silicon will become far too expensive to create and incremental nonsense will give way to more nuanced innovations. There will also be a lot of very expensive products for the super rich and only scraps for the plebs as there is no reason to scale pricing over time instead of branding perception based marketing of exclusivity for the elite.
What’s the likelihood of a dominant player emerging and implementing patented, proprietary RISC-V architecture changes which turn out to be necessary for high-performance? And if such a company gains sufficient market share, they could turn RISC-V into basically another x86-64 with many proprietary extensions. Sure, others could create their own RISC-V base processor - but if their performance is 500% lower than processors from the proprietary vendor who would purchase them?
Zero. There are no secrets on silicon. Everything can be reverse engineered based on observation and lapping the die.
Sure, but not legally so. Patents are the reason there are only really two x86-64 vendors and a company could similarly patent their own RISC-V modifications.
Not really. Reverse engineering has traditionally had limited value because of the exponential growth of silicon and the time it takes to develop anything on a node.
It is impossible to understate how insane CPU speeds are at the present. These are high speeds even for radio frequencies. Traces are no longer required to be connected at these speeds. Capacitive coupling becomes a major aspect of signal transmission; the entire RLC gamut is impactful. Anyone can, and likely every major player does, reverse engineer every major hardware design they can acquire. This will only allow one to understand what is now irrelevant in terms of design. The new chip in question will already be well into the process of amelioration of development costs.
You see, the reason the chip costs so much initially is because of the cost of the node and all of the hardware involved in creating the chip. The tooling costs are immense now. The reason the price of hardware drops over time is because actually producing the chips is dirt cheap by comparison. The initial cost is a projection about how much time it will take to pay back the initial investment assuming the units sell in low volume. As that volume increases and the loans are paid back, the price is dropped to access more segments of the market at lower price points. There is a market saturation balance that must be maintained to keep the next generation viable. Each new product takes 10 years to create. This means a company like Nvidia has that (IIRC (prob not)) has something like an 18 month new major product release schedule Will have the next 6 generations of products already in various phases of planning and design. If you reverse engineer a publicly available product, you are largely reverse engineering the fab’s capabilities while also looking at a product that is 6 generations behind the real cutting edge. Not to mention, you are also never going to be successful releasing your reverse engineered product because you cannot sell it at a competitive initial price required to pay for the required tooling when the original product has already done so and can be sold for cost of manufacture with a small markup.
The enormous difference between the cost of tooling and cost of actually making the chip is absolutely essential to understand in combination with exponential growth. This is what William Shockley realized in the 1950’s and convinced the US government to pursue. The entire modern world we all have known is due to this exponential growth. This growth is absolutely remarkable in human history as the only time a civilian business endeavor has out classed the economic growth created by the largest militaries of the world. This is what Shockley actually realized; that the military budget could not match the growth potential of silicon up to the plateau of potential scaling when physics prevents further scaled nodes. There are only a few new nodes left, and those are the next 10 years. This means we are already at the end of scaling because hardware designers are actually already there. The concept of Venture Capital is actually a pseudo extension of military budgeting constraints in a round about way. Indeed, Silicon Valley is where the real battle of the cold war happened. The Soviet Union and China failed to realize how silicon scaling would alter and become a decisive military operation. Once that ball started rolling, it is impossible to catch the front line so long as the design edge is kept a closely guarded secret and the extreme capital required is too high to be viable.
This is the real reason why your mobile devices are all running orphaned kernels on undocumented hardware. Your need to buy new devices when you are told is the primary factor driving these new nodes. You are less likely to realize that this is ultimately a tax to avoid large scale wars and major conflicts. This is why the changes happening right now in politics are not trivial. Moves are being made that appear similar to the era before venture capital and the Pax-Silicon™. We are already at the end of silicon exponential growth. There is no replacement for the exponential growth of silicon to outstrip military based spending and growth as has been the standard for all of the rest of human history. This will inevitably lead to hording scarcity and conflicts. The enormity of funding will cause the use of militaries to press advantages before they disappear. These are the factors that created world wars of the past and they will arise again. The next era of technology is going to be biology, but we are at least a couple of centuries away from biology as an engineering science where something like a synthetic brain is capable of producing a Turing complete deterministic computer on par with a CPU of the present. There is no clear path to exponential growth either like there was with scaling silicon. Perhaps the software organization, libraries, and database scaling will be an exponential growth factor, but I don’t know how that will have a barrier of entry on par with silicon in a way that is insurmountable by everyone including large governments and coalitions of governments.
So this is the world that is changing around the issue of RISC-V. It will come into its own in a post VC growth era. We are at the end of that growth already. Reverse engineering becomes relevant now and proprietary secretive strategy is no longer of the same magnitude of military significance. My narrative here will become more and more obvious with time. This has been a years long curiosity between many of my interests like why the world prior the the 1950’s was so different, why the USA won the cold war, and understanding the history of the microprocessor to wrap my head around all the peripherals present in an Arduino which was born out of a desire to learn Megasquirt when I was a hotrod car nut ages ago. I abstract across broad spaces well like this and like to simplify complexity because I’m dumb like that.
Things like patents are just weapons of the super rich. They have no real relevance here. The outcome of these cases has nothing to do with justice or right and wrong. These are battlefronts of militaries with convoluted rules of engagement. In the next 10 years this proxy conflict space will be abandoned and everything changes into an unknown state. Likely new silicon will become far too expensive to create and incremental nonsense will give way to more nuanced innovations. There will also be a lot of very expensive products for the super rich and only scraps for the plebs as there is no reason to scale pricing over time instead of branding perception based marketing of exclusivity for the elite.