Just one catch with comparing NVIDIA to Rolls Royce of chips then pointing out not everyone has to buy a Rolls Royce.. unlike CPUs, there's not a whole lot of competition with GPUs or NPU/TPUs yet. Someone else is going to have to open that market up, and then AI developers or pulling a Dropbox becomes viable.
Intel's too far off with ARC to give it a go.. AMD could have production capability but need more polish. There's GraphCore, Cerebras Systems but they're still small. Watching that space certainly will give a better indication as to future supply & demand to feed the OpenAI & Anthropics of the world.
It’s a smart comment - it’s way easier from a timeline perspective to go and build software compared to a chip factory and supply chain. You need time to get more players in the game
Great overview CJ. I do think the gross margins will grow over time with optimizations but I think competition amongst these vendors will require them to make big investments outside of R&D to keep up. So as Moore’s law reduces the COGS there might be less differentiation requiring each to put more cash into S & M spend. I don’t know if that will be at the same scale as the R&D costs.
Yeah. I think me and Dartz are essentially saying the same thing. Right now the thing sells itself but this could become commodity in the next 5-10 years. Speculation for sure but possible.
Looks like we have two competing forces. Moore's law says the cost of computing will go down (# of transistors double, computing power doubles). That can also apply to power usage dropping per MIPS. On the other side we have demand increasing, being users demanding more complex requests and operations, which require more MIPS. The gross profit is where the race is measured.
The other race is how you charge a customer. What business model works and both takes market shared and improves gross profit?
Just one catch with comparing NVIDIA to Rolls Royce of chips then pointing out not everyone has to buy a Rolls Royce.. unlike CPUs, there's not a whole lot of competition with GPUs or NPU/TPUs yet. Someone else is going to have to open that market up, and then AI developers or pulling a Dropbox becomes viable.
Intel's too far off with ARC to give it a go.. AMD could have production capability but need more polish. There's GraphCore, Cerebras Systems but they're still small. Watching that space certainly will give a better indication as to future supply & demand to feed the OpenAI & Anthropics of the world.
It’s a smart comment - it’s way easier from a timeline perspective to go and build software compared to a chip factory and supply chain. You need time to get more players in the game
Great overview CJ. I do think the gross margins will grow over time with optimizations but I think competition amongst these vendors will require them to make big investments outside of R&D to keep up. So as Moore’s law reduces the COGS there might be less differentiation requiring each to put more cash into S & M spend. I don’t know if that will be at the same scale as the R&D costs.
Yes you make a great point around S&M - to my knowledge it’s been relatively light to date. The stuff quite literally just sells itself.
So maybe I underestimated the points they make up in higher cogs on lower S&M (at least currently)
Yeah. I think me and Dartz are essentially saying the same thing. Right now the thing sells itself but this could become commodity in the next 5-10 years. Speculation for sure but possible.
I think this is a classic case of jockeying for 1st mover advantage. The giants are willing to burn phenomenal amounts of money to lock in users now.
Looks like we have two competing forces. Moore's law says the cost of computing will go down (# of transistors double, computing power doubles). That can also apply to power usage dropping per MIPS. On the other side we have demand increasing, being users demanding more complex requests and operations, which require more MIPS. The gross profit is where the race is measured.
The other race is how you charge a customer. What business model works and both takes market shared and improves gross profit?
That’s a great line - gross profit is where the race is measured. I need to write something on that more broadly across tech
Snowflake chart isn’t 2009-2024z
It’s calendar not fiscal
It’s calendar year but title says 2009 - 2024, but legend at the bottom starts at 2021. Just trying to help would post a photo if it would let me.
appreciate it