Breaking News
This is the future we shall bring into being
Elon Musk's Vision For The Future Looks Like A Movie
Phosphorus Munitions Fired by Israel Burn Groves, Orchards in Southern Lebanon
Iran's IRGC Warns It Will Expand 'Geography of the War' If the US Attacks Again
Top Tech News
BEYOND THE MOON: NASA plans a nuclear-powered fleet to push DEEPER into space
Big Oil Backs Mazama's $135 Million Bet On Superhot Geothermal
Quantum compass takes flight on Northrop's Lumberjack drone
Terrifying AI Behavior Even Its Creators Couldn't Explain
WW3 Near-Miss? AI Hallucinated Nuclear Weapons Components Aboard Chinese Vessel Bound For Iran
Stripe Is Building the Toll Road for the Machine Economy
Toyota Dakar hopeful swaps twin-turbo V6 for hydrogen fuel cells
4-min turbo-inflatable cubic camp castle is now available in the USA
Researchers Used Claude To Hack OpenAI Employee Accounts
Chinese Scientists Develop New Oil Refining Method That Cuts Energy Use by 90%
Tesla Dojo, Supercomputers and AI
There does not seem to be a limit for neural nets to utilize more resources to get better and faster results.
Tesla is motivated to develop bigger, faster computers that are precisely suited to their needs.
The Google TPU architecture has not evolved as much over the last 5 years. The Google TPU chip is designed for the problems that Google runs. They are not optimized for training AI.
Tesla has rethought the problem of AI training and designed the Dojo AI supercomputer to optimally solve their problems.
If Tesla commercializes the AI supercomputer that will help to get to lower costs and greater power with more economies of scale.
One of the reasons that TSMC overtook Intel was that TSMC was making most of the ARM chips for cellphones. TSMC having more volume let them learn faster and drive down costs and accelerate technology.
99% of what neural network nodes do are 8 by 8 matrix multiply and 1% that is more like a general computer. Tesla created a superscalar GPU to optimize for this compute load.