:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/63699546/generalfusion1.0.1537364818.0.jpg)
Vancouver is a land of scenic harbors, tall mountains and startups trying to harness the limits of physics.
In town for the TED conference, I had the occasion to visit two such companies yesterday: D-Wave and General Fusion. D-Wave, a quantum computing company, is all about the very cold and the rather tiny. It has built enormous refrigerators that each house a single chip, laced with “qubits” that can be in the superposition of both 1 and 0 at the same time and can carry an electric current with no resistance at low temperatures.
Meanwhile, General Fusion is all about huge and hot. The company is putting together the pieces for an alpha version of a nuclear reactor plant that would use magnetized target fusion. That is, it slams together hydrogen atoms by shooting donut-shaped electrified plasma into a chamber where it’s squished by synchronized pistons from all angles. This happens at a temperature of 150 million degrees. The point: To create clean and cheap energy.
Both of the companies say their products are on the verge of a breakthrough. Over about a decade of research and development, they have each acquired their own posse of doubters, who say they are designing expensive, impractical systems that don’t really work yet.
That may well be true, but people at both D-Wave and General Fusion like to compare their cost of development — and the broader investment in their respective spaces — to the estimated trillion dollars of investment over the past decades that have fueled the rise of traditional computing and its generational leaps forward associated with Moore’s Law.
Fusion always seems like a far-away prospect, but it’s closer than you might think, said General Fusion founder Michel Laberge in a talk at TED. “Very soon, somebody will crack that nut,” he said.
In fact, plotted against the curve of Moore’s Law, progress in fusion performance looks pretty parallel, according to Laberge.
The difference is, fusion doesn’t really work at all until it crosses a threshold on that chart — one Laberge and General Fusion think they are very close to achieving, if they can only create a system that is 150 million degrees, dense and long-lasting.
Over at D-Wave, which is headquartered about a 20-minute drive from the Vancouver Convention Centre, a 512-qubit quantum computer is already in the hands of customers and research partners, who have demonstrated that these machines can match the state of the art in classical computing.
What that means is, for certain problems — generally where someone is trying to optimize something — the D-Wave machine can already compute something just as fast as a state-of-the art classical set-up by considering multiple options simultaneously.
But that’s not enough. At this point there’s no big advantage to quantum computing because it’s not cost effective. The big leap is when quantum computers can do things demonstrably better than classical computers so they justify the big fridge and the big price.
D-Wave hopes and believes that will happen later this year, upon the arrival of a 1024-qubit quantum computer that’s close to being ready.
The promise of quantum computing is the possibility of solving problems that would require massive quantities of computers — perhaps more than are available in the world.
Because the combination of computing capacity and big data has been so effective in machine learning, it’s quite possible that this work will aid leaps forward in artificial intelligence. In fact, Google has a D-Wave machine and is working on that now. So is D-Wave co-founder Geordie Rose, who emphasized in an interview yesterday, “Whether or not it’s quantum is not important. It’s what it does.”
Meanwhile, a 15-minute drive away, General Fusion, which shares investors with D-Wave, is about two to three years out from creating its own power plant. Today, the pistons work well, and the plasma is hot enough and dense enough. Within the last month, the gas donut has started lasting long enough for the system to work, so now the company is turning its focus to compression and timing, according to Michael Delage, VP of strategy and corporate development.
Similar to D-Wave, General Fusion is at a point where it needs to get its system working and cost effective. “We expect to be at break-even energy in a couple of years,” said CEO Nathan Gilliland.
When this is built out, General Fusion thinks it can provide power at a cost of seven cents per kilowatt, comparable to the cost of coal.
Why this is important? As Laberge said, “It could solve all our energy problems cleanly for the next billion years.”
So why are both of these companies in Vancouver? Part of it is the investment climate. D-Wave founder Rose called out Haig Farris of Ventures West and Mike Brown of Chrysalix as “gunslinger types” who invest for the long term. D-Wave has raised $130 million, while General Fusion has raised $50 million.
Or maybe it’s the mountains meeting the sea. “There’s something around competency in hardware and pushing boundaries,” said General Fusion’s Delage.
And especially during the week of TED, it’s a place of optimism. “With AI, or quantum computing, or fusion, there are these things the world should have — and the fact they don’t is a travesty,” said Rose. “We’ve created tremendous wealth on this planet, and we should be using it on our sense of wonder.”
This article originally appeared on Recode.net.
