Tech Newsround – Quantum & Nvidia

IBM lays out its wares & Nvidia goes for Japan

IBM & Quantum computing – Still no Pong.

• Quantum computing is lining up to be the replacement for silicon when the rules of physics finally make it obsolete, but even with advances from IBM, there is still a very long road to travel.
• IBM will publish a series of papers that make use of a combination of both supercomputers and quantum computers to solve various scientific tasks.
• These tasks are highly technical in nature and remain within the realm of tasks that Quantum computers are capable of handling.
• These are tasks like solving certain chemical puzzles and simulating quantum physical environments.
• While this represents a step forward in terms of using Quantum computers in practical applications, it does not do very much for Quantum’s ability to replace silicon.
• The problem is that when one tries to use Quantum computing for regular tasks like playing Pong or training neural networks, its efficiency and speed are no better than the supercomputers that are already available today.
• This is why I think the real “Quantum moment” is when these machines can be used for general-purpose computing, but I suspect that I will be long retired by then.
• Consequently, silicon still has a long road ahead and just as there were many tricks and techniques used to extend the life of deep ultraviolet lithography beyond what anyone expected, I think the same will be true for extreme ultraviolet (EUV).
• IBM badly needs something to bring it back into the cutting edge of technology, but it will have to wait for a while if it is betting solely on Quantum computing.

Nvidia – Ecosystem game in Japan.

• Not to be left out of the current jockeying for position that is going on in the AI ecosystem race, Nvidia is moving on Japan and is looking to lock the Japanese market into its silicon and platforms.
• Nvidia aims to build a series of “AI Factories” (data centres in English) so that Japanese companies can create and train AIs specific to Japan as well as their own use cases.
• This is a good fit for Nvidia, as Japan has not been a leader or even a proper participant in the AI journey so far and it clearly wants to get a foot on the ladder.
• Consequently, Japan is very keen to welcome the current hardware leader to build domestic facilities and this gives Nvidia a chance to lock Japan into its AI ecosystem.
• All of the others like Meta, OpenAI, Google and so on are looking at creating their silicon to compete with Nvidia and should the control point in AI shift, then Nvidia could see its current grip on AI weaken (see here).
• By locking in Japan early on, Nvidia is moving to mitigate this threat in what could end up being quite a large market.
• Japan is very good at manufacturing both chips and the equipment to make them but has not made any impact in the market for AI chips meaning that this partnership does not threaten a domestic Japanese business or strategy.
• Nvidia remains priced for perfection but while it continues to blow expectations out of the water, its valuation is likely to hold.
• However, the risk of its position weakening is slowly growing which is what I think this move into Japan is all about.
• I continue to think that there are much better places to look in the technology sector.