Last year, IBM opened a cloud-based quantum computing platform for the public and the IBM Quantum Experience. In March, IBM announced that it will launch its first commercially available universal quantum computing platform by the end of the year. & Mdash; Quantum computers may develop faster than you might think.
Here from DigitalTrend, Lei Feng net do not change the original intention of the compiler.
Last week, when IBM launched the news of the first commercial general-purpose quantum computer, many people thought it would mean that the project would be handed over from IBM Research & rdquo; IBM System & rdquo ;. But Scott Crowder, IBM's CTO and vice president of technology strategy and systems division, said the fact is not exactly that, and that's not true.
"This is not exactly a & lsquo; handover & rsquo; is actually a kind of cooperation. "This is definitely the turning point from pure science, pure research to engineering, development and commercialization," explains Crowder. & Rdquo;
Crowder believes that the current stage of quantum computers, the equivalent of the traditional computer just started the 1940s, this is not 60 or 70 years, but the 40's. We are still in the early stages, and are still driving many of the most basic work, and we are trying to make more people able to use the system. & Rdquo;
Everyone wants to know how quantum computing will affect their daily lives, and IBM wants to find the answer to this question by introducing a commercially available universal quantum computer.
Let the quantum computer out of the lab
In the press release, IBM referred to the potential applications of quantum computers, as well as some of IBM's internal staff and experts in the field, but about when the quantum computer was introduced, IBM only promised that the general-purpose quantum computer would be in the future A few years & rdquo; delivery, and did not give a specific timetable. In addition, IBM also mentioned in the announcement, in the next few years to build 50 quantum bit system.
Because IBM only gives a very vague schedule, may lead to a lot of people think that IBM Q just to occupy, and no substantive plan. But Crowder says there is a definite schedule within IBM.
"Yes, we have a very precise definition of & lsquo; the next few years & rsquo; We have announced that our goal is to provide systems to some industry partners this year. We have already said that in the next few years, we will upgrade to 50 quantum level. & Rdquo;
IBM Q is not a project for stroke. In fact, Crowder also specifically mentioned that IBM's goal is to make the quantum system come true, rather than just stay on the PPT.
While its specific business hours are still in a state of confidentiality, IBM has high expectations for the next phase of the development of quantum computers through IBM Q. Although IBM is not ready to publish a detailed road map to the outside world, but it describes their own quantum computer and other systems on the market What is the difference between the no reservations.
The debate about propaganda
Unless you have been keeping a close eye on the development of quantum computing, you may be confused when IBM announced that it will launch its first commercially available universal quantum computer last week because D-Wave began offering business users a few years ago Quantum computing hardware.
However, D-Wave's propaganda means a lot of criticism. Because D-Wave has been using Quantum bits to promote the product, according to its publicity, D-Wave's latest system can access 2000 qubits, which is far more than IBM in the next few years to develop 50 quantum position vision. However, the two are not directly comparable, because the D-Wave is to use quantum annealing to achieve quantum computing, can only solve some specific problems, and IBM's direction is a general-purpose quantum computer, as with traditional computers, it Can solve various types of problems.
"I think people who have a deep understanding of this area will understand the difference." Crowder said, "Even the general technical experts, in the face of this situation, will be confused." & Rdquo;
He added, "We did not provide a fixed system, but a common system." More importantly, the ability of quantum computing to really make everyone excited is that you can use it to explore & lsquo; Exponential Space (index space) & rsquo ;. & Rdquo;
Crowder mentioned here is the concept of "full quantum entanglement", which allows the qubits to have more states, rather than the open and closed states of the traditional binary. In this case, only about 40 to 50 quantum bits are needed, beyond the traditional computer simulation capabilities. At present, for the D-Wave hardware can provide these advantages still have a lot of controversy.
Even at the current level, IBM's quantum computing hardware is still a bit further than any of the previously released commercial products. But in order to achieve their ambition, IBM not only looking for customers, but also looking for partners.
Find a partner
IBM is expected to launch its first quantum system by the end of 2017 and allow some industry partners to take the lead. However, this is not a traditional transaction similar to the purchase of hardware from the manufacturer. When IBM calls an early user a partner, it's not just about it.
Crowder said, "We certainly want to have close partners, and are able to promote eco-system partners." & Rdquo;
Of course, as IBM's partners will also benefit, because the hardware itself will be IBM is responsible for maintenance and regular upgrades. Quantum computers have high requirements for space, temperature and other practicality. IBM is responsible for maintaining the system and will enable partners to focus on the potential of the system.
Thanks to IBM Experience, IBM has a good track record of providing quantum computing through the cloud. This is not like you in the laboratory to develop a 5 qubit computer, you can not mind when the spirit. "Crowder said," We have a quantum computer that has been running since May 2016, which proves our ability. " & Rdquo;
However, as IBM's partner is not without conditions, IBM hopes that the other party can tell the future plans for IBM Q.
It is clear that IBM Q is a long term project. Once you get a partner ticket, it means that you will explore the unknown area with IBM. Of course, this can help them to seize the opportunities before the advantages of the quantum system become clear. While IBM is thinking about the long term, it does not mean that it is ready to give up all the previous efforts in the traditional computer field.
The real leap
When asked whether IBM believes that quantum and traditional computers will coexist in the future, Crowder said, of course, longer than my life. & Rdquo;
He explains that traditional computers are doing better in storing and processing large data than they are today, and that this is the case in the foreseeable future. In theory, general-purpose quantum computers can do anything that traditional computers do, but in my opinion they are not good at dealing with a lot of data, and quantum computers are really good at exploring the huge problem space. & Rdquo;
IBM does not think that quantum computers are the next generation of traditional computers. Quantum computer A brand new hardware category, it has its own strengths and weaknesses, but also has great application potential.
Of course, the current quantum computing is not yet mature. Over the past few years, while global researchers have made great strides in developing large-scale general-purpose quantum computers, there has been little progress in developing potential applications, which, to some extent, has created a vicious circle.
If you are 20 to 25 qubit system, you can still simulate on your notebook. Crowder said, but when you increase the quantum bit to 40 to 50, even the world's largest supercomputer can not simulate all the possibilities, this discontinuity is very interesting. & Rdquo;
In fact, you need to use the quantum system to really understand how to create algorithms for the quantum system, and the development of application cases. These are not through the traditional computer system simulation obtained. & Rdquo;
Set the measure
At present, we are still in the early stages of building a general-purpose quantum computer, and even a set of standard methods for comparing different systems are not. IBM hopes to solve this problem by introducing the Quantum Volume, which includes the number of qubits, the quantum run quality, the quantum bit join, the coherence time, and the parallelism.
"These parameters can tell you how much space the quantum computer can explore, so it can also be intuitive to show how powerful this quantum computer in the end. "Crowder said," We welcome other people to come in and improve this standard. " Over time, we will need some similar standards to make it clear that the differences between different quantum computers, as well as different generations of quantum computers. & Rdquo;
But the fact is that there is no such measure to judge how much work is needed to turn quantum computers from theoretical work into viable commercial products.
As I said, the quantum computer is still in the 40s. "We do not have an industry standard to rate the performance of quantum computing systems, but this will improve rapidly over the next few years as IBM and other companies will continue to raise the thresholds of these systems," said Crowder. & Rdquo;
IBM is doing the basic work, and these work in the next few years will prove to be very valuable, because quantum computing has become an important part of IBM's business interests.
Crowder said, "We are convinced that we have now reached the critical point, in the near future, we will see a way to achieve greater system." & Rdquo;
IBM has invested a lot of time and resources into quantum computing, and IBM Q represents the next step. The company has the ambition to make quantum computing a commercial product, and if it can find the right partner at this critical stage, IBM Q will be able to play a key role in allowing quantum computing out of the lab, in the real world Dig out the application potential.
ViaDigitalTrendLei Feng network compiler