News Microsoft's latest Quantum computing claims have been named 'unreliable' by scientists

[Admin]

Several scientists expressed concern that Microsoft's Majorana 1 chip is deeply flawed.

Microsoft's latest Quantum computing claims have been named 'unreliable' by scientists : Read more

 

[Pierce2623]

While have zero understanding of the tech underpinning quantum computers, I do know that Microsoft doesn’t throw billions of dollars at a project they don’t believe will ever work.

 

[ttquantia]

Quantum computers are not for everything. The only really credible application so far is factoring. Factoring is useful for breaking encryptions such as RSA. But not all encryption methods. And that's about it. There is a theoretical speed-up obtainable by Grove's algorithm for NP-complete problems, but it does not make solving NP-complete problems efficient. It only makes solving them less exponential in the theoretical worst case. This includes hard optimization problems arising in lots and lots of fields of computing. However, this is only a theoretical improvement, as none of the solution methods for these types of optimization methods actually typically run as slow as the theoretical worst-case predicts, but much much faster. So much faster, that quantum computing has no chance of getting even close. That's why it is broadly believed by experts that "lots of applications" for quantum computers as they are currently known, is just hype, with far fewer real-world applications than what some people would like us to believe. Who wants to decrypt encrypted messages? Not me. I have no use for that at all. Almost nobody does.

 

[Alex/AT]

Another problem is that it is not just any 'processing power' in any form of application.
It's only a power to quickly [attempt to] [randomly] bruteforce input values for given output over specific combination action, so it's highly specialized & not applicable to classical computing.

 

[Mr Majestyk]

Quantum computers are not for everything. The only really credible application so far is factoring. Factoring is useful for breaking encryptions such as RSA. But not all encryption methods. And that's about it. There is a theoretical speed-up obtainable by Grove's algorithm for NP-complete problems, but it does not make solving NP-complete problems efficient. It only makes solving them less exponential in the theoretical worst case. This includes hard optimization problems arising in lots and lots of fields of computing. However, this is only a theoretical improvement, as none of the solution methods for these types of optimization methods actually typically run as slow as the theoretical worst-case predicts, but much much faster. So much faster, that quantum computing has no chance of getting even close. That's why it is broadly believed by experts that "lots of applications" for quantum computers as they are currently known, is just hype, with far fewer real-world applications than what some people would like us to believe. Who wants to decrypt encrypted messages? Not me. I have no use for that at all. Almost nobody does.

You better tell the researchers using them for weather/climate modelling and materials science then they are doing it wrong.

 

[Alex/AT]

You better tell the researchers using them for weather/climate modelling and materials science then they are doing it wrong.

Yes, it has its uses. But for general computing it means no change.

 

[ttquantia]

You better tell the researchers using them for weather/climate modelling and materials science then they are doing it wrong.

Which researchers? Can you cite a research paper that shows that some computational problem in weather/climate or materials was done with quantum computers, and that computation could not have been very easily done with classical computers instead? No such thing exists.
You are a victim of quantum hype and disinformation. Quantum computers so far have not even done any factoring that had not been earlier easily done with classical computers.
All current stories about applications (including yours) are about possible future applications, not about the reality today or even the next years. There is a very big difference between current actual reality, and possible applications 20 years from now.