Home / Technology / Google quantum computer leaves old-school supercomputers in the dust

Google quantum computer leaves old-school supercomputers in the dust

  Google's Sycamore chip drives a quantum computer

Google's Sycamore chip drives a quantum computer


A Google quantum computer has far surpassed ordinary computer technology. This achievement is called quantum supremacy and is an important milestone in a revolutionary way of computing. Google published the results on Wednesday in the journal Nature. This happened after more than a decade of work at Google, including the use of its own quantum computer chip called Sycamore.

"Our machine performed the target calculation in 200 seconds, and from measurements in our experiment, we determined that this would be the case. The world's fastest supercomputer takes 10,000 years to produce a similar output," Google researchers said in a blog post about the work.

The performance that took center stage in September with a paper-based release does not mean the beginning of classical computers, at least in the eyes of today's quantum computer experts. Quantum computers are expected to perform well on certain tasks – such as optimizing investment portfolios and developing new drugs at the molecular level – but not most of the tasks we do today on computers.

However, Google's findings provide evidence that quantum computers could break out of research laboratories and turn to mainstream computing .

Quantum computer scientist Scott Aaronson likened the landing to the moon in terms of reach. And Google boss Sundar Pichai called it a "big break" in a tweet on Wednesday .

A large industry is dedicated to improving classic computers, but a small number of expensive laboratories at companies such as Google, Intel, Microsoft, Honeywell, Rigetti Computing and IBM are also pursuing general-purpose quantum computing. These are delicate devices that run just a hair's breadth above absolute zero to minimize the likelihood that they will be disturbed. Do not expect to find quantum computers on your desk.

Google's Speed ​​Test offers applications for computer work such as artificial intelligence, materials science, and the generation of random numbers. "

The physicist Jim Preskill, who introduced the concept of quantum supremacy in 2012, summed up this idea: Google's selected test was good for showing the speed of the quantum computer," but otherwise no problem of much practical interest. Preskill said in October after the newspaper's premature release.

Quantum vs. Classic Computers

Almost every digital device, from ENIAC in 1945 to Apple's iPhone 11 in 2019, is referred to as a classic computer Electronics relies on logic circuits that add, for example, two numbers and memory cells to store the results.

  Google Quantum Computer

Google's quantum computer is not like a conventional machine, and it is close to absolute zero cooled.


Quantum computers are completely different and rely instead on the rules of physics that apply to ultra-small objects such as atoms.

Where classical computers store and process data as individual bits, quantum computers use a 1 or a 0 to name another foundation called Qubit. Each qubit can store a combination of different states of 1 and 0 simultaneously through a phenomenon called overlaying. I told you it was weird.

Not only that, but several qubits can be brought together by another quantum phenomenon called entanglement. In this way, a quantum computer can simultaneously examine a variety of possible solutions to a problem.

Exponential Accelerations

In principle, the performance of a quantum computer increases exponentially: adding a qubit more and doubling the number of possible solutions in one fell swoop. For this reason, quantum computer engineers are working to increase the number of qubits in their machines.

"We expect their computing power to continue to grow at a double-exponential rate," Google researchers said in their article. This is even faster than the only exponential improvement that Moore's Law sets for classic computer chips.

Google's computer had 54 qubits, although one did not work properly, so only 53 were available. This corresponds to the number in IBM's most powerful quantum computer .

But the number of qubits is not everything. Unavoidable instabilities cause qubits to lose their data. To counteract this problem, researchers are also working on error correction techniques that allow a calculation to circumvent these problems.

IBM questions Google's quantum results

IBM is a big fan of quantum computers, but questioned Google's prematurely released blog post Monday.

"We argue that an ideal simulation of the same task on a classic system can be done in 2.5 days and with far greater accuracy," wrote IBM researchers. They proposed different algorithms and another classic computer design in their own preprinted paper.

Google welcomed improvements in quantum computer simulation techniques, but said the overall result would be "unaffordable," even for the world's fastest supercomputer, with more double exponential growth. We have already deviated from classic computers and have taken a completely different approach.

Intel did not comment on Google's findings, but said quantum supremacy was "a strategic yardstick."

"We are committed to moving quantum from the lab to commercialization," said Jim Clarke, director not yet

Quantum computing's mathematically proven quantum computing capabilities crack some of today's encryption technologies.

However, much larger quantum computers and new technological breakthroughs are needed to address the issue of error correction [19659006] "In order to realize the full promise of quantum computation (for example, with Shors algorithm for factoring), technical leaps are still required," the researchers said in their article.

At the same time, the US government and others remember "post-quantum" cryptographic methods that exploit the cracking capabilities of Qua withstand computers.

Although quantum computing is a radically different process, there is still no explosion for the tech industry.

First published October 23, 2:15 pm, PT.

Update, 3:09 pm, PT: Adds more details and comments from Google CEO.

Source link