In fairness, it's not unreasonable to expect that the technology could explode exponentially at some point as one or more persistent issues are finally solved in a scalable way.

Sure but who will be able to afford, 'own' or even operate a QM code breaking machine even if the scalability (some are based on the basic physics of entanglement coherence being stable at the scale needed, not just the engineering of building things) issues are solved? It's IMO doubtful that cracking cryptocurrency is on the list of those that could operate one if they existed.

It's not unreasonable to also believe the experiments will fail because some fundamental physics law (akin to impossible Perpetual Motion machines) can't be broken, we just do not have enough data to make a scientific prediction on the path to follow.

 

And when will a Desktop version be available? .... Quantum Laptop, anyone?

Depends on the users lap.

 

REALLY!!! I was not trying to be funny at all!!. To solve any problem there must be an adequate description of the problem, and then a means to pose that problem to the computer. And possibly even a scheme to recover the solution when the computer achieves it. ALL of those bothersome details make the problems even bigger.

 

What operating system will these new computers use?? and who will create the software to do the computing?? It is important to always understand that the hardware is the smallest and least expensive part of the system. It is the software that constantly need repair that costs so very much. (at least the one that has to send out weekly repairs and bug fixes)
GOOD software costs a whole lot more than that!!

Work in those areas has been going on in parallel. We can simulate quantum processors just like we can simulate electronic circuits or other CPUs, so researchers into quantum software are not having to wait in the wings until hardware is available. Of course, like any CPU, the development of the specific quantum instruction set architecture will go hand in hand with the development of the final quantum processor, but marrying that to the higher layers in the stack should be pretty much proforma.

 

I cannot possibly imagine what the instruction set for such a computer looks like ...

 

I cannot possibly imagine what the instruction set for such a computer looks like ...

When I was at the Academy one of the big players (can't remember who) gave a workshop on quantum computing specifically from the perspective of the programmer. Their intent was to encourage faculty and cadets to get involved in writing quantum programs that would be run on their simulators and, potentially, even on their hardware as it was developed. I wasn't able to attend because it conflicted with my teaching schedule, so I don't know if they talked about what the underlying instruction set looked like or not. This was well over a decade ago, so I'm sure it's progressed a lot since then.

 

I already get phone calls from computers trying to sell me medicare or an extended warranty on my Corvette, which I never owned or even rented one. So what sort of grief can we expect from a quantum computer???

In fact, what actual benefit will quantum computers bring to the majority of humanity??
Already, computer data centers are wasting huge amounts of electrical power. And what benefit are THEY providing for us??

 

In fact, what actual benefit will quantum computers bring to the majority of humanity??

I expect the fields of chemistry and medicine will be revolutionized among many other things (material science, fluid dynamics, etc). IF QC is ever made to work as expected.

 

I already get phone calls from computers trying to sell me medicare or an extended warranty on my Corvette, which I never owned or even rented one. So what sort of grief can we expect from a quantum computer???

In fact, what actual benefit will quantum computers bring to the majority of humanity??
Already, computer data centers are wasting huge amounts of electrical power. And what benefit are THEY providing for us??

Actually, there are a number of problems of practical significance that quantum computers could make a huge difference in. In addition to physics, chemistry, and medical research, there are big problems of practical commercial interest, such as scheduling and routing for parcel transportation, rail traffic, airline scheduling, and much more.

These are all NP-Hard problems in which not only does finding a solution take exponentially long time on a non-deterministic Turing machine, but so does even verifying that a proposed solution is correct. As a result, beyond very small problem sizes, it is impossible today to find optimal solutions and we rely on heuristic approaches to find solutions that are often poor approximations (the ones that can be arrived at in usable time frames on practical computers), but are the best we can do.

Now, quantum computing does not turn arbitrary NP-Hard problems into polynomial-time problems (at least most people think that's the case, but it hasn't been proven one way or the other), but Grover's algorithm reduces the scale size from N to √N, which is a HUGE improvement. For instance, if the largest size of a particular problem you could solve today is 10,000 points, now you can solve a 100,000,000 point problem in a roughly comparable amount of time (depends on what the scaling factor is).

 

The traveling salesman!

 

The traveling salesman!

Many of these problems are variants of TSP (the Traveling Salesman Problem), but quite a bit more complex because instead of one salesman visiting N cities exactly once and ending up back at the start, you have things like a fleet of N aircraft that need to service M airports, some multiple times a day. And they don't just have to visit the airport, but have to do it in time windows that minimize both the number of flights and layover times for as many passengers as possible. And it has to take fuel limitations into account. And allow for the fact that neither aircraft nor crews have to return to the same point on a given day, but crews do have to get back to their point of origin at least so many days. And a host of other considerations, such as how many aircraft can be handled at the terminals at any given time without causing problems. Makes the classic TSP problem tame.

 

OK, I see three posts listing what math problems could be solved, and those applications seem like Quantum Computers could solve them. The actual benefit from getting the solution is a lot less obvious. It seems a bit like a friend of mine who had a car that could do 150 MPH. It was really cool, but there were not many roads with that high a speed limit.

 

OK, I see three posts listing what math problems could be solved, and those applications seem like Quantum Computers could solve them. The actual benefit from getting the solution is a lot less obvious. It seems a bit like a friend of mine who had a car that could do 150 MPH. It was really cool, but there were not many roads with that high a speed limit.

So, you don't see any benefit to more efficient transportation systems? Less pollution? Cheaper service? Less congestion?

 

So, you don't see any benefit to more efficient transportation systems? Less pollution? Cheaper service? Less congestion?

I don't see any benefit in making the computers for those systems a thousand times faster. The present systems are not limited by the computer speeds, although if they run the common BLOATED OS, that will slow them a whole lot. The same processors running good machine code are many times faster. Most people do not see how much that bloated operating system slows the process. Back in about 1981 I WATCHED a programmable electrical tester running a 6809 processor with an 8 megahertz clock vastly outrun a next generation tester with a 33 megahertz windows processor on a similar test. Of course machine code needs to be written and then compiled and assembled, but it is usually very fast. So it is the big operating system that slows things down. Like a "FUNNY CAR" having to pull a travel trailer.

 

I don't see any benefit in making the computers for those systems a thousand times faster. The present systems are not limited by the computer speeds, although if they run the common BLOATED OS, that will slow them a whole lot. The same processors running good machine code are many times faster. Most people do not see how much that bloated operating system slows the process. Back in about 1981 I WATCHED a programmable electrical tester running a 6809 processor with an 8 megahertz clock vastly outrun a next generation tester with a 33 megahertz windows processor on a similar test. Of course machine code needs to be written and then compiled and assembled, but it is usually very fast. So it is the big operating system that slows things down. Like a "FUNNY CAR" having to pull a travel trailer.

It's not 1981. Your 'reasoning' is extremely simplistic and IMO uninformed about the realities of OS system design and basic programming in general.

 

Every one of the past few OS releases has demanded much faster clock rates. I am not in a position to go thru the code and show why,, but not everybody buys a faster computer just to celebrate a new OS release.

 

Seriously, Bill ... I don't get your point. Are you saying you find no advantages to multi-parallel processing? (yeah, I know that the word "multi-parallel" sounds kind of dumb, but you know what I mean)

 

I don't see any benefit in making the computers for those systems a thousand times faster. The present systems are not limited by the computer speeds, although if they run the common BLOATED OS, that will slow them a whole lot. The same processors running good machine code are many times faster. Most people do not see how much that bloated operating system slows the process. Back in about 1981 I WATCHED a programmable electrical tester running a 6809 processor with an 8 megahertz clock vastly outrun a next generation tester with a 33 megahertz windows processor on a similar test. Of course machine code needs to be written and then compiled and assembled, but it is usually very fast. So it is the big operating system that slows things down. Like a "FUNNY CAR" having to pull a travel trailer.

Several times, or a thousand times, is not the same as billions and trillions of times faster -- and for many problems, trillions of times doesn't come close to how much the speed up factor is.

 

A billion times faster is quite a claim indeed.
It seems that we will need to wait to see if that is actually the case.

 

@MisterBill2:

The speed difference between coding in a higher level language and machine language is linear. If it speeds up a short task from 2 seconds to 1, a longer one speeds up from 2 hours to 1 hour.

The speed up of quantum computing is exponential. This is far different. The larger the problem, the larger the speed ratio.

Taking the traveling salesman problem, it might look something like this, on a conventional computer:

2 cities: 100 seconds
3 cities: 1000 seconds
4 cities 10000 seconds
10 cities: 10,000,000,000 seconds

And with quantum, it would instead grow linearly:

2 cities: 100 seconds
3 cities: 150 seconds
4 cities: 200 seconds
10 cities: 500 seconds

So no, more efficient coding does not come close to replicating the speedup you get with quantum computing.

The problem with realizing quantum speedup is that, in practical realizations, the complexity seems to also go up exponentially with the problem size (number if Qbits)