From ENIAC to IonQ: A Blueprint Moment for the Future of Fault Tolerant Quantum Computing

There are times where looking back, you can say that an industry was really born.

I think this MAY prove to be one of those moments. I gave a hint at this when I sent a tweet last week:

https://x.com/sakacc/status/2045114847440814106

Today I can reveal what I’m talking about, and know when I send these sort of things out there — it’s not coming from outer space. This is what was on the board:

It’s chicken scratch (like all authentic whiteboards) — but if you squint, you see an end-to-end architecture. On the left you have low level hardware (trapped ion implementation of cat states), then moving right you see a hierarchical, modular architecture that implements logical instructions with execution elements, memory elements, bus elements. It’s a model that is familiar to old-timers who have seen many compute architectures.

The Quantum Compute architecture team (all around the world, some here in my home office) have been working on something really big, and today we can talk about it.

BTW — when I made my IonQ trips around the world in November last year debating whether to join IonQ I hinted at “this company is going to do some big things”, it was because I was exposed to many of these ideas.

Huh? Let me tell you what I mean with a couple examples.

• The EDVAC paper. Published in 1945 (primary author was John von Neumann) — this 101 paper was central to the top secret ENIAC effort. It was the first time that all the pieces for what created the whole classical computer system architecture to come over the subsequent decades. Few understood it at the time — but think of the impact on the world we live in today (a great trip through history — click on the link to paper below).

https://en.wikipedia.org/wiki/First_Draft_of_a_Report_on_the_EDVAC

• A Protocol for Packet Network Intercommunication (the “TCP/IP paper”). Published in 1974 by the IEEE and authored by Vint Cerf and Bob Kahn, this was the huge milestone that ultimately led to the Internet. Few understood it at the time — but think of the impact on the world we live in today (a great trip through history — click on the link to the paper below)

https://www.cs.princeton.edu/courses/archive/fall06/cos561/papers/cerf74.pdf

• Attention Is All You Need paper. Published in 2017 by Google and authored by multiple acclaimed authors, this was the starting gun on the AI era that we are all navigating now. Few understood it at the time — but think of the impact on the world we live in today (a great trip through history — click on the link to the paper below)

https://arxiv.org/abs/1706.03762

What do these 3 have in common?

1. They were NOT the first things published in each of these domains. They built on lots of pieces of prior art and knowledge — decades of work. They stood on the shoulders of giants. For a variety of reasons, the prior art/knowledge moved forward incrementally. Importantly, but incrementally.
2. They each represented a seminal moment where many things came together — and were foundational for incredible innovations. For a variety of reasons, those pieces were not “incremental” steps — but were “giant leaps”.
3. At the time it wasn’t clear, but they marked milestones where history later proved that the world tilted on its axis so to speak. In each of their domains, there was a “before” and an “after”.

Time will tell, but I think the paper the IonQ Quantum Computing R&D team published today could eventually prove to be one of those history turning points. Pull up a chair, grab a pencil (real or digital) and take a slow careful read.

https://arxiv.org/pdf/2604.19481

• It is, to my knowledge, the first full system-level blueprint for a complete fault-tolerant system — something that will do REAL work with circuits that are millions, tens of millions, billions of gates deep and hundreds and thousands of fault tolerant qubits wide.
• It isn’t an “academic paper” — at least that’s not all it is. The team is actually building this system (full stack, from low level hardware to the software surfaces) as we speak. It won’t be perfect (far from it) — but it will be the first REAL, scaled fault tolerant system. We will get slings and arrows for sure — but the most meta response is: “ok, what have YOU got?” The industry has a lot of hype — heck, I’m prone to hyperbole. When I read a lot of analysis, there are a lot of dreams and aspirations in there. Architectures that show promise are exciting, but the steep physics hurdles are glossed over. The approach we’re taking is pragmatic, practical. It prioritizes simplicity, manufacturability. Everything in the paper doesn’t count on pixie dust or unknown leaps — but on technologies that we’ve already proven. This is being built as we speak. Every system we ship (including Tempo, and the upcoming 256 system), every job that runs in the cloud — those are living, breathing proofs (and are derisking our path to the 10K system and those that will follow).
• The details as you flow through it at the back for Shor’s and Hamiltonian Simulation are not resource estimations, they are a full accounting of complex circuits with millions of gates.

Oh, and we’re not stopping here.

Like the 3 other examples (EDVAC, TCP/IP, Attention is all you need) — this is what I think this paper represents:

1. It’s not the first deeply technical and thorough quantum computing paper (from us or the ecosystem as a whole). There have been many awesome papers from us and the ecosystem as a whole (on modalities, dives into one part of a hardware, compilers and other parts of system architecture, and even Fault Tolerance approaches and QEC). It builds on lots of pieces of prior art and knowledge — decades of work. For a variety of reasons, the prior art/knowledge moved forward incrementally (in important ways)
2. It represents a seminal moment where many things came together — a “giant leap” — and I believe it could prove to be equally foundational.
3. At this time it won’t be clear to many, but it may (history will be the judge) mark a milestone where history later proves that the world tilted on its axis so to speak. If I’m right, when it comes to quantum computing, there will be a “before” and an “after”.

We are at the “EDVAC” moment for Quantum computing — prior work is analogous to the pre-ENIAC phase of hand-wired computers.

I’ll do more posts in the future to double click on this (the “walking part” and the “cat part” are important, as are all the “factory” parts are all important, worthy of deep discussion).

Read, reflect, and input always welcome!!!