For those in technical fields some large number of decades ago, slide rules were THE means of calculation.
No calculators, no PCs, no smart phones, no AI.
Recall (maybe) that slide rules work on the basic principle that adding logarithms is the equivalent of multiplication:
log(A x B) = log A + log B
log (A/B) = log A – log B
log Ac = c log A
Calculators got rid of that principle; Boolean (binary) logic was more conducive to electronic implementation (aka “digital” stuff).
And it’s been that way since … since roughly 1970.
But here now in 2026, AI and quantum computing are the current rage. And Nvidia is the king of the heap in providing the chips that drive that effort. If you have any investments in a ETF or mutual fund, you likely own shares of Nvidia – among if not the largest company in the world with a capitalization of over $5Trillion (1000 times 1 billion) – larger than all but 3 or 4 countries. (Elon Musk himself is now worth about $1T)
This just came over my transom:
A new company is challenging Nvidia … by implementing the electronic slide-rule methodology of using logarithms; binary logic systems have gotten too unwieldy. This is big if they can pull it off.
Tensordyne claims its 72-chip system can run large LLMs four times as fast using 20% of the power of Tensordyne’s new chip, Napier,s fast using one-fifth the power compared to a 72-Nvidia GB300 system.
The “secret” behind the outsized efficiency of the new device is how it does matrix multiplication, the main math of AI. It takes advantage of the fact that the logarithm of A times B equals the logarithm of A plus the logarithm of B.
“We’ve turned multipliers into adders,” explains Gilles Backhus, a Tensordyne founder and vice president of AI. Adders are smaller and more energy-efficient logic circuits than those that do multiplication, he says. So Napier can pack more compute into a smaller area and still save on power.
Can’t help it – this is the kind of stuff that captures my interest.
