New Space Race A few weeks ago I posted about the five major Chinese projects to develop reusable rockets, as they never seem to get much attention over here. Since that post, China has now successfully demonstrated 2 reusable rocket projects: • Their own cable catch barge • Folding landing legs (falcon 9 style) There are 3 other major Chinese rocket programs, and there are a lot of smaller minor programs behind those. China’s efforts are serious and progressing fast. What should be very obvious to anyone following SpaceX is that space (or rather Earth orbit) is really a numbers game. A space data center is actually not a “center” at all, it is a satellite constellation, it is not a single large object assembled in space it is a constellation of satellites much like Starlink already is. The way to think about it is that each satellite is basically 1 server rack. A single satellite is optimally around 150-250kW, which is EXACTLY where current AI racks are in the 2020s. So the fit seems fairly attractive even if you have to swallow the launch cost, and the lack of synchronicity that comes from being spread out in a constellation. Space data is solar powered and China already makes 95% of the world’s solar, so going after Elon is a very natural strategy for China, they already totally dominate one of the main industrial substrates, the solar cells. But space is only 1 branch of this story. Nvidia’s next gen rack technology Kyber moves to 800VDC, and to liquid cooling, this pushes rack power density to 600kW and possibly 1-1.5MW, well beyond the solar satellite optimal package size. At those power densities you start to go beyond what is optimal for a satellite and so the space data center window is opening in terms of access, just as Nvidia is moving chips toward something much denser. Chip tech is about to fork by heat sink. • Radiative cooling into deep space • Conductive/convective cooling into water This is why SpaceX needs their own silicon, this is Terafab is being built. Space chips need to run at higher temperatures to dissipate heat radiatively and be more energy efficient, whereas conductive and convective cooled chips just want to cram as much power and compute as possible into the synchronous operation horizon (the distance that light travels during 1 clock tick). For example at 3GHz clock speed light can travel 10cm between each tick of the clock, this is why fast chips must be small, and why a satellite constellation cannot run synchronous operations. Radiative cooling in space means distributed asynchronous compute. Liquid cooling at home means densified synchronous compute. So this new Space Race sits entirely on one side of the fork. Either way, China seems to be doing their fast follower thing here, and is following Elon into building a big space constellation of hot asynchronous chips in space. China has pre-built the solar fabs to support scaling and has 5 launch programs pushing to give them access, 2 of which are now demonstrated. Space silicon with high chip temperatures and limited energy budget, is not Nvidia’s direction, and the two tech forks lead to very different application strengths and weaknesses and different applications in that layer will fall under one fork or the other.

The Electricity Grid I post a lot of stuff about the electricity grid, here the CEO of the largest grid in America (PJM) lays it out pretty clearly. • What worked for 2 decades… no longer works • This is structurally different from history • You are facing an era of scarcity • The situation is not tenable PJM are facing a demand explosion. Now a demand explosion in some industries is +500% demand. But this is infrastructure, this is tens of trillions of dollars of assets and it takes time to mobilise and deploy things at this scale. In infrastructure, when demand growth shifts from +1-2%/yr to +8%/yr, then you suddenly need to be building 4-8 times more assets per year, than you previously did. If you were deploying $1 trillion / yr to grow at 1%, you now need to deploy $8 trillion / yr to grow at 8%. Suddenly you need to deploy many trillions of dollars per year to meet this growth. If you cannot get it done, prices will rocket for everyone. Failure leads to inflation. This is not a PJM problem, this is not even a US problem, this is a global problem. PJM are formally validating what some people have been saying for a while now. This is not temporary, we cannot uninvent the technologies that have precipitated this change. The world has changed and we must adapt. Global retail electricity sales are about $3.6 trillion per year, of that, around $900 billion goes to transmission and about $2.7 trillion goes to wholesale generation. The transmission system many developed countries have is the wrong system going forward. Our transmission systems in the West are built for transporting power from big coal plants to power big towns. That’s not what we are doing now. We have replaced most of the coal plants with two largely decentralised but highly correlated fleets of intermittent generators (wind and solar), that are growing like fracking wells because they are also quick to deploy. Their quickness to deploy new generation projects is massively destabilising for the grid. The grid was designed for coal plants. The grid is a $50 trillion machine. It is by far the biggest asset in any country. It isn’t something you can toss away, it isn’t something you can swap out overnight. We also have new categories of industrial demand (hyperscalers) that will capture an increasing share of GDP. This new demand category is going to set the marginal price of electricity for everyone else, and these guys are not as price sensitive as your widowed grandmother. This is a difficult problem to address because of: i) the scale ii) the capital intensity It’s also a global problem, because it’s born out of a new technological paradigm. It will not spread around the world at equal pace, but everywhere is going to eventually face it down. Some people are fleeing to space for solutions to avoid this snafu, but that’s only a temporary fix. Once the hyperscalers have their demand satisfied, the next demand explosion immediately follows, and this second wave is 20x the scale of the current problem. The second wave is how do you power billions and billions of robots and billions of autonomous machines, doing work that currently can’t be done? This industrial revolution is very much a two stage revolution, first you power up the chips, then you power up the actuators. Chips scale down, actuators scale up. There’s no Moore’s Law for actuators, they obey Newton’s Laws of motion instead. This is the crux of the energy problem facing our civilisation. The energy system we have today is the one we wanted 20 years ago. The energy system we will have in 20 years from now, is the one we start building today. It’s time to build this solution.

Reverse Engineering One of the big challenges with all consumer products is that your competitors always get to do full tear downs of whatever you put on the market. It is extremely difficult to maintain an advantage with any mechanical components, unless you have robust IP. Probably what is going to happen with humanoid robots, is that countries are going to require products to be licensed. One of the biggest emerging concerns is national security, does a country want to import 5 million robots who live in private households and get their firmware updates from overseas? No. The security risk is real. The product is a potential trojan army. Countries are therefore likely to limit which models are allowed to enter their market. This will happen before sales hit the million unit mark. So maybe the first $ 50bn revenue will beat the restrictions, but after that the market will become much harder to enter. So once we reach a technological plateau, we might see regulatory drawbridges coming up quite quickly as a way to prevent markets being flooded with low end copycats that have security concerns. So whilst reverse engineering will be rapid and we will see lots of copying, it might not matter. What is probably more important is how the client nation trusts the OEM on a security.