Why AI Keeps Wasting GPUs
NeuReality CEO Moshe Tanach builds the networking hardware that sits between AI chips, on the argument that the industry keeps answering its inference problem by buying more GPUs rather than using the ones it already has.
And then quantum computers walked into the room.
Quantum, accelerated computing, neutral atoms, GPUs, QPUs, hybrid workflows, and the changing physical stack beneath AI.
NeuReality CEO Moshe Tanach builds the networking hardware that sits between AI chips, on the argument that the industry keeps answering its inference problem by buying more GPUs rather than using the ones it already has.
Bob Sutor thinks too much of quantum computing is aimed at the history books, and that real progress comes down to basic engineering.
Rory Daniels works on emerging technology at techUK, the UK's technology trade association, and he thinks the UK's robotics strength lies in specialised machines rather than general-purpose humanoids.
Scott Crowder is vice president of IBM Quantum Adoption, and his message to anyone who doubts quantum computing is that quantum computers are real.
Nvidia's Sam Stanwyck and Infleqtion's Pranav Gokhale expect quantum computers to be built into AI supercomputers, expanding the range of problems a supercomputer can solve.
Shapol and his co-founder used to lead a reusable rocket programme. Now their company, Entangl, builds deterministic design software for the systems that keep AI data centres running.
Infleqtion CEO Matt Kinsella builds quantum computers and clocks from neutral atoms, cooled by lasers to colder than outer space inside a system that sits at room temperature.
Conductor Quantum founder Brandon Severin wants to turn the transistors the chip industry already makes by the billion into spin qubits, and uses AI to tune them.
Blake Johnson leads the quantum engine at IBM Quantum, the part of the software stack that runs a quantum computing query on IBM's machines.
Does probability mean anything without consciousness? In chapter twelve of Irreducible, Federico Faggin argues that probability has relevance only if consciousness exists.
GAIB lets someone with ten dollars buy a slice of the GPUs that power AI, instead of leaving that to venture funds. Kony explains how a graphics card becomes a financial asset.
Oliver Dial, chief technology officer of IBM Quantum, explains Starling, IBM's plan for a fault-tolerant, scalable quantum computer by 2029.
Katia Moskvitch was named European Science Journalist of the Year in 2019, while business editor at Wired. She has written a book on neutron stars and is now writing one on quantum computing, AI and the theory of everything.
Can quantum entanglement prove we have free will? Federico Faggin thinks so, and chapter eight of Irreducible is where he makes the case.
Coleman Collins is a director of product management at IonQ, which has built trapped-ion quantum computers since it spun out of university research at Maryland and Duke in 2017.
Kathrin Spendier is technical prize director of the XPRIZE Quantum Applications prize, a three-year competition co-sponsored by Google Quantum AI to find quantum algorithms for problems people actually care about.
Murray Thom joined D-Wave in 2002 and is now its vice president of quantum technology evangelism. He works on quantum annealing, which is already being applied to optimisation problems like scheduling.
Equal1 CEO Jason Lynch builds quantum processors on standard silicon, so the machine can eventually sit in a rack rather than fill a laboratory.