IBM is betting on quantum computing, but it can’t win without the help of startups. And how long it will take for the bet to pay off is anybody’s guess.

That was the message of the Q Summit, a one-day meeting of quantum computing researchers, investors, and entrepreneurs hosted by IBM in Menlo Park, Calif., last week.

“We need startups in the quantum space,” said Joe Raffa, director of IBM Ventures. “There’s a huge amount of work to do, he said, to cross the “long revenue desert.”

The potential is big, but the risks are huge.

“There are really only 15 or 20 people in the world” who will make or break quantum computing, said Matt Johnson, CEO of QC Ware, a startup that is building a commercial software package for quantum computing.

Quantum computing was ten years away 20 years ago, said Vijay Pande, general partner at Andreessen Horowitz, and it still is. But eventually, he said, it will make a leap and take the computing world by surprise.

“For generation after generation, the quantum computer will be way slower than the classic machine,” he said. But its rate of evolution is “hyper exponential, so suddenly, in some ‘n’ years, it will jump over the classic machine. The real question is when will this transition happen.”

He had some predictions along that ten-year timeline. Within 10 years, for certain limited disciplines, he says, quantum computers will dominate. New algorithms will be developed to use them—algorithms that aren’t currently being developed for classic computers because it would be pointless to run them. Why? They would be just too slow.

But, he said, this all makes it very difficult to address quantum computing as a venture capitalist. “It’s hard to talk about the market if I don’t even know what the algorithms are.” 

Bill Coughran, a Sequoia Capital partner and former senior vice president of engineering at Google, said that, as an investor, “I’ve struggled with the question whether quantum computing is development or still research. VC firms think [along] a ten-year time horizon, not 20 or 30 years. Are we on the cusp of a breakthrough?”

The question, he says, “is still open.”

Startups have formed, however, with venture money behind them. Johnson’s software startup, QC Ware, is betting “that the people building the hardware will get it [to the point] that our software is useful.”

Christopher Savoie, CEO of startup Zapata Computing, is betting that chemistry will be the first field to benefit from quantum computing, and is developing algorithms for drug discovery and chemical design.

These two, and six more startups, are part of the IBM Q Network, an organization launched in late 2017 to accelerate the development of practical applications for quantum computing.

And Sequoia has made an investment in a quantum startup, Quantum Circuits, just in case the answer to Coughran’s question about whether quantum computing is on the cusp of a breakthrough is yes. But Coughran has still another concern.

“Most prominent teams today,” he said, “are built around people with strong academic track records, but those are not always the ideal teams with which to build real companies.”

IBM’s Anthony Annunziata, leader of the IBM Q Network, admitted that, to date, there are no commercial applications of quantum computing that beat conventional computing. “We are at the point of getting quantum ready. We like to think that we are on the cusp of emerging, of moving out of pure science,” he said. “The era of the quantum advantage is still a few years out; it’s hard to see how many.”

It’s not the right time “to go all in, to put tens of millions of dollars into this,” Annunziata continued. “But it’s exciting enough, things are progressing well enough, that [every tech organization] should have at least one person who is your quantum person, who can learn the basics and take it from there.”