Breaking News
Donald Trump's Flagrant War Of Aggression Against Iran Is An Impeachable Offense
Iranian Cyberattack - Insiders Warn of "Cyber 9/11"
A new player in the technology world is making waves by offering computer memory...
AI Shock Troops at Trillion Dollar Scale
Top Tech News
US particle accelerators turn nuclear waste into electricity, cut radioactive life by 99.7%
Blast Them: A Rutgers Scientist Uses Lasers to Kill Weeds
H100 GPUs that cost $40,000 new are now selling for around $6,000 on eBay, an 85% drop.
We finally know exactly why spider silk is stronger than steel.
She ran out of options at 12. Then her own cells came back to save her.
A cardiovascular revolution is silently unfolding in cardiac intervention labs.
DARPA chooses two to develop insect-size robots for complex jobs like disaster relief...
Multimaterial 3D printer builds fully functional electric motor from scratch in hours
WindRunner: The largest cargo aircraft ever to be built, capable of carrying six Chinooks
RNA Crop Spray: Should We Be Worried?
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.