#AshesOfTime I thought I was the winner.Until one day I looked into the mirror and saw the face of a loser.I failed to have the person I loved most to be with me in my best years.๐ณ๏ธโ๐
NVIDIA JUST MADE QUANTUM COMPUTING RESEARCH 13,000ร FASTER!
A computational task that previously took 12 weeks can now be completed in just 45 minutes.
Here's what happened:
NVIDIA just introduced TCBP-JOBS, a new GPU-accelerated quantum error correction decoding algorithm that dramatically speeds up research into protecting quantum information from errors.
But that's not all.
NVIDIA is partnering with IBM and Fermilab to develop next-generation modular quantum computers, combining NVIDIA's DGX B200 supercomputers with IBM's quantum processors.
Other breakthroughs include:
โข IBM, RIKEN and NVIDIA demonstrated quantum-classical chemistry simulations with 79ร better accuracy than earlier runs of the same experiment.
โข Their research modelled protein-ligand complexes containing up to 12,635 atoms, more than 40ร the previous 303-atom quantum-classical benchmark.
โข A new integration with Slurm will make it easier to schedule quantum processors alongside conventional CPUs and GPUs in supercomputing environments.
Think about the possibilities.
AI, GPU supercomputers and quantum processors working together could eventually accelerate research in drug discovery, advanced materials, chemistry, energy and fundamental physics.
We're moving toward a future where different computing architectures complement one another to tackle problems that remain extraordinarily difficult today.
We're building an entirely new computing ecosystem.
And its potential is enormous.
AMAZON QUIETLY BUILT A QUANTUM CHIP.
Itโs called Ocelot, it runs on โcat qubitsโ named after Schrรถdingerโs cat, and Amazon says it could cut the cost of fixing quantum errors by up to 90%.
Qubits make errors constantly, and fixing them takes huge numbers of extra qubits. That overhead is the biggest reason useful quantum computers are still years away.
Amazonโs fix is to build the protection into the qubit itself. A cat qubit stores information in a pulse of microwave light that naturally resists one of the two main kinds of error. On Ocelot, those errors got so rare they took about a second to show up, an eternity at quantum speeds.
The chip:
14 core parts: 5 cat qubits, 5 circuits that keep them stable, and 4 qubits that catch the remaining errors
Its best error-correcting setup used 9 qubits, where the standard approach needs 49
AWSโs head of quantum hardware, Oskar Painter, says the design could pull Amazonโs timeline to a practical quantum computer forward by up to five years.
Ocelot is a prototype for error correction, not a computer you can run programs on, and the 90% figure is Amazonโs own projection for future, larger versions.
Google has Willow. Microsoft has Majorana. IBM has its roadmap. And now Amazon has a chip of its own.
At SC26 in Chicago, @IBM, @nvidia, and @Fermilab announced plans for a modular quantum computing co-design platform to advance the integration of quantum and classical computers as both paradigms scale towards the future of compute. ๐งต
๐จ: Nobel Prize Physicist James Peebles says we must admit a hard truth: Dark matter and dark energy are just placeholders for our ignorance.
95% of the universe remains completely unknown to us.
Fermilab, IBM, and NVIDIA are planning a modular quantum computing co-design platform aimed at accelerating the development of scalable, fault-tolerant quantum computers for scientific discovery.
https://t.co/CAXstAyQtq
@SQMSCenter
Weโre developing a quantum-classical codesign platform with @IBM and @Fermilab to advance practical quantum computing.
NVIDIA DGX B200 systems will connect with IBM quantum computers to accelerate error-correction and applications research. Separately, our new TCBP-JOBS decoder in CUDA-Q QEC delivers up to 13,000x faster decoding, advancing fault-tolerant quantum computation. #SC26 https://t.co/4ho6jyziuv
AI IS NOW HELPING BUILD QUANTUM COMPUTERS. HERE'S 8 TIMES IT HAPPENED THIS YEAR:
1. Proofs. An AI system running an Anthropic model proved an 18-year-old conjecture about quantum speedups in 3 days.
2. Algorithms. Googleโs AI agents cut the cost of quantum algorithms by up to 290x, each one checked by mathematical proof.
3. New gates. Infleqtion says OpenAIโs GPT helped it discover a new way to entangle logical qubits, halving the cost of that step in its 30-logical-qubit demo.
4. Error correction. Harvardโs AI decoder cut quantum errors by 17x or more, and could reduce the qubits needed by about 40%.
5. Code design. qBraid used Googleโs AlphaEvolve to design quantum codes that need 4 to 5x fewer qubits.
6. Calibration. NVIDIA released open AI models that cut quantum chip setup from days to hours.
7. Running the machine. In China, an AI agent tuned 108 of 112 qubits in 4.7 hours. An expert needs 18 to 24.
8. Chip design. IBMโs AI agents made a quantum decoder design up to 204x faster, for about $299 in computing costs.
A year ago, AI and quantum were separate races. Now AI is helping build the quantum computer.
$INFQ $IONQ $QBTS $RGTI
The Apple TV app on Vision Pro now lets you discover immersive films from apps like Fovii and Theater.
Find a film on the Vision tab, then tap โOpen Inโ to choose where to watch.
CLAUDE JUST SOLVED TWO OPEN PROBLEMS IN QUANTUM GRAVITY
Quantum gravity is the biggest unsolved problem in physics: making quantum mechanics work with Einsteinโs gravity. Black holes are where the two collide.
In 1983, Stephen Hawking and Don Page showed that as certain model universes heat up, black holes donโt form gradually. They appear suddenly, like water flashing into steam.
Their proof only worked for large universes. In 2003, physicists laid out the quantum calculation needed to check small ones. Nobody finished it for 23 years. The math was too brutal.
Anthropic says its newest Claude model, Fable 5.2, just did it, โworking largely autonomously.โ Black holes appear suddenly in small universes too.
It also pinned down the exact limit of a technique physicists have used for 15 years to rule out candidate theories of quantum gravity.
Why quantum people should care: this is the same physics that connects black holes to quantum error correction, the thing every quantum computer needs to scale. In 2022, Google even used its quantum processor to simulate a tiny wormhole model.
AI is now doing original work in quantum physics. The field that builds quantum computers is about to move faster.
$INFQ $IONQ $QNT
Longevity Grok, a.k.a. your personal longevity superintelligence (PLSI), is an open-source multi-agent workhorse designed to keep you alive & healthy for as long as possible. A collaborative effort from @InSilicoMeds & the Sinclair Lab @harvardmed. Visit https://t.co/Hn42ZnyK2z
IBM JUST UNVEILED ITS MOST ADVANCED QUANTUM COMPUTER.
IT RUNS THE MOST POWERFUL QUANTUM CHIP IBM HAS EVER BUILT.
IBM just unveiled a new IBM Quantum System Two running its Nighthawk processor. IBM calls it the most advanced quantum computer it offers.
Nighthawk is a 120 qubit chip with square-lattice connectivity, meaning each qubit links directly to more of its neighbors. That lets it run bigger, more complex circuits with fewer workarounds. Itโs the same chip IBM just paired with Oak Ridgeโs AMD powered supercomputer to work on error correction.
IBM CEO Arvind Krishna says the system can deliver scientifically useful results today, and itโs already being put to work. Deloitte, AbbVie and Northwestern are using it for drug discovery, EPRI and the University of Illinois for power grids, and Fermilab for physics.
The system is running at the University of Chicagoโs Hyde Park Labs and will anchor Illinoisโ new National Quantum Algorithm Center, backed by a 25 million dollar state grant.
Is Nighthawk the chip that gets IBM to fault tolerance?
$IBM $INFQ $IONQ