The technology giant said Thursday that several companies and universities have joined its initiative to create practical uses for quantum computing, a nascent technology that researchers believe could eventually surpass conventional computers in speed and power.
Other organizations joining the project include Daimler AG, industrial materials company JSR Corporation, Hitachi Metals, Honda, chemical company Nagase, Keio University in Japan, Oak Ridge National Laboratory, Oxford University, and the University of Melbourne.
The field of quantum computing is still in its experimental phase, with IBM’s (ibm) head of quantum research and vice president of artificial intelligence Dario Gil comparing it to the state of conventional computing in the 1950s.
Regarding quantum computing, Gil says that “the rate of progress has been phenomenal” for quantum computing and that the technology will become standard more quickly than it took for computers to reach the mainstream.
Companies could presumably use a single quantum computer instead of thousands of computers to train software for deep learning tasks like recognizing cats in photos, for example. This would save time and money by reducing the need for operating so many computers, among other benefits.
Quantum computing relies on particles called quantum bits, or qubits, to process data, unlike modern computers that rely on transistors packed into conventional silicon chips. Whereas silicon chips encode data in either two states—on or off as represented by a “0” or “1”— qubits can store data in both states at the same time, potentially letting the computers crunch data more quickly.
Despite the potential for more efficient data crunching, quantum computers don’t easily work with today’s software. These are fundamentally different kinds of computers, and therefore require new software using different mathematical tricks to make them compatible with modern apps.
JPMorgan, for instance, wants to research how quantum computing can improve tasks like risk analysis and asset pricing, Gil said. The software that currently underlies these kind financial tasks must be “mapped” to work on IBM’s quantum machines, a laborious and experimental task.
Get Data Sheet, Fortune’s technology newsletter.
Engineers from JPMorgan, Samsung, and the other organizations will work side-by-side with IBM researchers on making future software work with quantum computers, Gil said. IBM does not prohibit these organizations from working with other companies like Google or quantum computing startup D-Wave on their rival quantum experiments, Gil explained.
Gil said he hopes that IBM and its partner companies will be able to use quantum power to beef up current business-related software tasks by 2020 or 2021. Such a feat would be noteworthy because it would take the field of quantum computing out of the research lab and into the real world.
Still, Gil concedes that it’s still “early days” and that there is still much to do before quantum computing becomes widely adopted.