IBM Quantum System Two is the building block of quantum-centric supercomputing. IBM Quantum System Two is the bedrock for scalable quantum computation. It is now operational at IBMs lab in Yorktown Heights, NY. It is 22 feet wide, 12 feet high, and today features three IBM Quantum Heron processors. It combines cryogenic infrastructure with third-generation control electronics and classical runtime servers. IBM Quantum System Two is the modular-architecture quantum computing platform that they will use to realize parallel circuit executions for quantum-centric supercomputing.
Earlier in 2023, IBM published research that demonstrated that quantum computers could run circuits beyond the reach of brute-force classical simulations. For the first time, they have hardware and software capable of executing quantum circuits with no known a priori answer at a scale of 100 qubits and 3,000 gates. Quantum is now a computational tool, and what makes me most excited is that we can start to advance science in fields beyond quantum computing, itself.
Breaking the 1,000-qubit barrier with Condor
IBM has introduced IBM Condor, a 1,121 superconducting qubit quantum processor based on our cross-resonance gate technology. Condor pushes the limits of scale and yield in chip design with a 50% increase in qubit density, advances in qubit fabrication and laminate size, and includes over a mile of high-density cryogenic flex IO wiring within a single dilution refigerator. With performance comparable to our previous 433-qubit Osprey, it serves as an innovation milestone, solving scale and informing future hardware design.
Access to the highest performing quantum processor: Heron
Building on four years of research, IBM introduced the first IBM Quantum Heron processor on the ibm_torino quantum system. Featuring 133 fixed-frequency qubits with tunable couplers, Heron yields a 3-5x improvement in device performance over IBMs previous flagship 127-qubit Eagle processors, and virtually eliminates cross-talk. With Heron, IBM has developed a qubit and the gate technology that will form the foundation of their hardware roadmap going forward.
IBM Quantum Heron features 133 fixed-frequency qubits with tunable couplers, yielding a 3-5x improvement in device performance over our previous flagship 127-qubit Eagle processors, and virtually eliminates cross-talk.
Qiskit 1.0 coming in February 2024
Quantum-centric supercomputing is not achieved by hardware alone. It requires performant software for generating and manipulating quantum circuits and middleware for executing hybrid quantum-classical workflows in a heterogeneous computing environment. Qiskit 1.0 marks the first stable release of Qiskit, the most popular quantum computing SDK. It delivers marked improvements in circuit construction, compilation times, and memory consumption compared to earlier releases.
In addition, Qiskit 1.0 outperforms competing compilation frameworks in both runtime and resultant two-qubit gate counts when mapping circuits to quantum hardware.
IBM Quantum Computing Roadmap to 2033
The new IBM Quantum roadmap highlights improvements in the number of gates that our processors and systems will be able to execute. Starting with a target of Heron reaching 5,000 gates in 2024, the roadmap lays out multiple generations of processors, each leveraging improvements in quality to achieve ever-larger gate counts.
Then, in 2029, IBM hits an inflection point: executing 100 million gates over 200 qubits with our Starling processor employing error correction based on the novel Gross Code. This is error correcting codes for near-term quantum computers.
This is followed byu Blue Jay, a system capable of executing 1 billion gates across 2,000 qubits by 2033. This represents a nine order-of-magnitude increase in performed gates since IBM put their first device on the cloud in 2016. The new innovation roadmap will demonstrates the technology needed to realize the Gross code through l-, m-, and c-couplers to be demonstrated by Flamingo, Crossbill, and Kookaburra, respectively.
Brian Wang is a Futurist Thought Leader and a popular Science blogger with 1 million readers per month. His blog Nextbigfuture.com is ranked #1 Science News Blog. It covers many disruptive technology and trends including Space, Robotics, Artificial Intelligence, Medicine, Anti-aging Biotechnology, and Nanotechnology.
Known for identifying cutting edge technologies, he is currently a Co-Founder of a startup and fundraiser for high potential early-stage companies. He is the Head of Research for Allocations for deep technology investments and an Angel Investor at Space Angels.
A frequent speaker at corporations, he has been a TEDx speaker, a Singularity University speaker and guest at numerous interviews for radio and podcasts. He is open to public speaking and advising engagements.
The Traveling Salesman Problem is an NP-complete problem. Quantum computers can’t solve NP-complete problems in polynomial time. So there’s no reason to think that they would help much with that problem.
They can factor large integers, and do the equivalent for elliptic curves, so they can break certain kinds of cryptography. But we already know of other forms of cryptography that we think they can’t break. So the world will switch to those.
If you have a problem you want to break by brute-force trying every solution, such as breaking AES-128 by trying every possible key, then quantum computers can use Grover’s algorithm for a square root speedup. But that just means you need to use keys that are twice as long. So if AES with a 128-bit key is secure against normal computers, then AES with a 256-bit key will be secure against quantum computers. For hash algorithms, it’s a different algorithm, so if 256-bit hashes are secure against normal computers, then 384-bit hashes are secure against quantum computers.
If you tried to solve something like the Traveling Salesman Problem by brute force, a traditional computer can only do N cities, for a pretty small N. A quantum computer using Grover’s algorithm could only do 2*N cities. That’s not much better. So you’d use heuristics for approximate solutions, instead. And it currently looks like the best heuristics come from giant neural networks. Which work better on normal computers than on quantum computers. So there isn’t much obvious use for a quantum computer there.
It’s possible that quantum computers would help with quantum physics simulations. So maybe they would help to simulate things like protein folding. Or they might not. It might take millions of logical qubits to simulate a very large molecule. And it could be that deep neural networks on a normal computer could actually do a better job of predicting the behavior.
There have been proposed AI algorithms for quantum computers. But we have no idea if they will be any better than our existing algorithms. At the moment, it looks like AI benefits more from huge amounts of memory, which is better for normal computers than for quantum computers.
It’s possible that someday we will find something useful for quantum computers to do. But for the moment, their only effect is forcing us to switch to different cryptography algorithms. That’s inconvenient, but not terrible.
GoatGuy, the above comment was intended as a reply to your question about applications. It looks like it ended up as a separate post, rather than a reply.
Brian, let me address this one to you directly.
I carefully read the marketing report (above), and was left feeling encouraged, rather excited and even elated. If one can believe the unscaled comparisons, things are really looking up for the IBM version of quantum computing in the future. Hêll … their Quantum Two modular design might already be ready to demonstrate solving — by hybrid quantum-classical computation — some interesting real world problems, right? At least the ad copy seems to promise it.
Then, being GoatGuy, I stepped back and asked, “well, what actual complex real-world computation has it done?”
Try as I might, I cannot elucidate what that might be. It’d be nice if “the answer” were as straight forward as “solving the 1000 city Traveling Salesman Problem in less than classical algorithmic time”. That kind of simplicity — for a tough computational problem that EVERY University trained computer science practitioner has grappled with, as part of their breadth requirements — would be ideal, as so many millions of tuned-in people have worked with it.
OK, I’ll admit: such a problem might not be “mappable to quantum”. Could be, I’m not an expert. The types of problems that seem to pop their scaly heads out of the computational swamp are much more abstract sounding. Working around the NP completeness of some relatively simple (but hard) problem in P time, quantum-wise. Sure.
But to frame my question neutrally, I just have to ask, “What has this modular quantum computational core done to solve real world ‘messy’ problems outside the confines of the gold-plated lab?”
That is my question.
________________________________________
Assuming (“the goat’s view”) that there really hasn’t been a nice meaty computational problem that the IBM super-duper quantum modular thing has solved yet, the next question seems obvious: so when WILL a real meaty computational problem be solved in a non-classical way, in a resolution interval at least within an order of magnitude of the classical approach? I don’t even need a win! ¹⁄₁₀ (i.e. 10× longer) would be fine. I’m not asking for 100% (99.99999%) accuracy. 80% or so would be just fine.
So, perhaps Brian you can dig around — taking advantage of your outstanding ‘creds’ — and gracefully ask this of your IBM quantum contacts. The world (certainly not “me”) needs to have hard answers as to the viability of quantum computing in the real world, far, far removed from the exquisite conditions isolated in a quantum lab.
⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
⋅-=≡ GoatGuy ✓ ≡=-⋅
GoatGuy .. you are a breath of fresh air! I’m an old coot & worked inside of vacuum tube computers (Q7) and archaic Burroughs, IBM, RCA machines. To date (80 y.o), I haven’t read of ANY applications for the massive capabilities for the promises to be afforded by AI, quantum computing etc. Other than research, air/space defense applications … It seems like razzle-dazzle to entice naive investors.