From friend Steve, who is more knowledgeable than me:
Hi Graham -
I just read your latest substack post, including the comment about Nvidia. I thought I'd share a bit of insight in a way that I'm sure you uniquely will appreciate, if you don't already have this insight, benefiting in part from the fact that I'm currently working in AI.
The bottom line: Nvidia isn't just making boards for video games and also heavily invested in AI. Those video game cards have themselves have become the hottest item in Silicon Valley since the iPhone as the cards themselves ARE the heart and soul of virtually all AI work today. And the AI industry and most every corporate org on the planet cannot get enough of those cards.
The original video game boards were uniquely designed to be number-crunchers to drive the dynamic graphics generation that was, and is, behind video games. Games have had a voracious appetite for better and faster and more detailed and more realistic visual generations. Thus the investment into the video game cards which have always been number crunchers, nothing more.
Somewhere long the way someone realized those cards were also useful for the sort of unique number crunching behind most AI algorithms. AI algorithms uses a lot of "tensors". Tensors are just multi-dimensional arrays of numbers. Even text-based AI systems "tokenize" text - i.e., convert text into numerical representations - for most efforts.
Furthermore, AI algorithms start by performing a complex "training" process on data they ingest. The "training" is the real work in building AI systems, as it involves potentially complex data analysis. The output of such an effort is a knowledge base, a carefully filtered and structured set of data itself, resulting from the analysis of the ingested input data, and that stored knowledge base, combined with the algorithm, work together to perform the job of the trained agent. Without that knowledge base, the AI algorithm is nothing, just another trainer.
So that training process is the starting point, and a big deal. And it often requires enormous compute power. A typical AI training process can take hours, weeks, literally months, even years. And that's on something like an Nvidia card. On a CPU many of the processes would take so long as to be not worth the effort.
How long NVDA can retain such a unique advantage is an interesting question. Oracle, Cisco and Apple have all had unique advantages that lasted a decade or more and made investors a lot of money even after they started trading at seemingly insane multiples. I'm watching, but don't know enough to place a bet.
Both AMD and Intel see Nvidia and are creating competitive chip assemblies. They don't have as a compelling software stack, yet. Nvidia has a production bottleneck that AMD hopes to fill.
At the moment, NVDA, has had blowout profits justifying their P/E to many. But many designers have discovered that dependence on a single supplier can ruin your day. So the AI creators who build custom arrays for clients are concerned. The "cloud" may have rainy days.
My friend Chris has been a professional investor. He annotated this blog, correcting me on several points while generally supporting my thesis. I have uploaded his valuable comments here.
The most significant change in the market relates to the value of Intellectual Property. We once thought wealth came from exploiting/using natural resources via labor and capital to convert that resource into a useful widget. Now we have an idea that via human invention (software) we can create something of value which might not even be tangible.
Three guys had an idea and created software to produce a tool that people could use. It took them ~ year to build that software. Google bought their tool for a hefty price and the three became instant millionaires. While the labor wasn't trivial, they were able to capitalize on that idea. Hard to place a valuation on that kind of wealth creation.
Valuing an investment is now a lot harder. I'm betting on a few companies in the ag sector. They try to improve food production via technology insertion. Micro-this and that, much software decision assistance, sensors and analytics are their tools. A lot of small guys trying hard - a bet on the future. We can hope that yields on a piece of land can be improved while using fewer resources.
At the moment I can appreciate that cash is earning a decent return with no risk while the market seems quite risky. I have hopes for a few beat-up firms given a market dominated by just a few high flyers.
From friend Steve, who is more knowledgeable than me:
Hi Graham -
I just read your latest substack post, including the comment about Nvidia. I thought I'd share a bit of insight in a way that I'm sure you uniquely will appreciate, if you don't already have this insight, benefiting in part from the fact that I'm currently working in AI.
The bottom line: Nvidia isn't just making boards for video games and also heavily invested in AI. Those video game cards have themselves have become the hottest item in Silicon Valley since the iPhone as the cards themselves ARE the heart and soul of virtually all AI work today. And the AI industry and most every corporate org on the planet cannot get enough of those cards.
The original video game boards were uniquely designed to be number-crunchers to drive the dynamic graphics generation that was, and is, behind video games. Games have had a voracious appetite for better and faster and more detailed and more realistic visual generations. Thus the investment into the video game cards which have always been number crunchers, nothing more.
Somewhere long the way someone realized those cards were also useful for the sort of unique number crunching behind most AI algorithms. AI algorithms uses a lot of "tensors". Tensors are just multi-dimensional arrays of numbers. Even text-based AI systems "tokenize" text - i.e., convert text into numerical representations - for most efforts.
Furthermore, AI algorithms start by performing a complex "training" process on data they ingest. The "training" is the real work in building AI systems, as it involves potentially complex data analysis. The output of such an effort is a knowledge base, a carefully filtered and structured set of data itself, resulting from the analysis of the ingested input data, and that stored knowledge base, combined with the algorithm, work together to perform the job of the trained agent. Without that knowledge base, the AI algorithm is nothing, just another trainer.
So that training process is the starting point, and a big deal. And it often requires enormous compute power. A typical AI training process can take hours, weeks, literally months, even years. And that's on something like an Nvidia card. On a CPU many of the processes would take so long as to be not worth the effort.
Thus the crazy spike in demand for Nvidia cards.
Nvidia also provides a solid tool set to manage their chip assemblies. That tool set has kept competition at bay.
How long NVDA can retain such a unique advantage is an interesting question. Oracle, Cisco and Apple have all had unique advantages that lasted a decade or more and made investors a lot of money even after they started trading at seemingly insane multiples. I'm watching, but don't know enough to place a bet.
Both AMD and Intel see Nvidia and are creating competitive chip assemblies. They don't have as a compelling software stack, yet. Nvidia has a production bottleneck that AMD hopes to fill.
At the moment, NVDA, has had blowout profits justifying their P/E to many. But many designers have discovered that dependence on a single supplier can ruin your day. So the AI creators who build custom arrays for clients are concerned. The "cloud" may have rainy days.
My friend Chris has been a professional investor. He annotated this blog, correcting me on several points while generally supporting my thesis. I have uploaded his valuable comments here.
http://www.grahamseibert.com/blog/Chris%20comments%203-11-2024.pdf
The most significant change in the market relates to the value of Intellectual Property. We once thought wealth came from exploiting/using natural resources via labor and capital to convert that resource into a useful widget. Now we have an idea that via human invention (software) we can create something of value which might not even be tangible.
Three guys had an idea and created software to produce a tool that people could use. It took them ~ year to build that software. Google bought their tool for a hefty price and the three became instant millionaires. While the labor wasn't trivial, they were able to capitalize on that idea. Hard to place a valuation on that kind of wealth creation.
Valuing an investment is now a lot harder. I'm betting on a few companies in the ag sector. They try to improve food production via technology insertion. Micro-this and that, much software decision assistance, sensors and analytics are their tools. A lot of small guys trying hard - a bet on the future. We can hope that yields on a piece of land can be improved while using fewer resources.
At the moment I can appreciate that cash is earning a decent return with no risk while the market seems quite risky. I have hopes for a few beat-up firms given a market dominated by just a few high flyers.