Are practical, large-scale quantum computers actually achievable?

[SeekerOfTruth007]

I saw an article on MSN yesterday outlining a new QCCD based quantum computer which has a whopping 96 Qubits. The article stated that in order to do real calculations, they would need a computer with at least 1 million Qubits (Because about 85% of them are used for error checking). I calculated that if Moore's Law applies to these devices, they would need an additional 14 doublings of Qubits at the same or similar costs in space and money, and this would take on average another 28 years if Moore's Law doesn't "slow down" for Quantum Computers... assuming this reasoning is even valid. As things stand, they spent millions and millions of dollars making the 96 Qubit computer, and the only thing it can do is generate random numbers.

lol

Updates

8 d

Oh yeah, Moore's Law for Classical Electronic Computers will hit its real physical limit in about 8 to 10 years, and Classical Electronic Computers will be virtually impossible to further improve beyond that, except maybe minor architecture improvements, which is why we have a ~2-year pad here.

Besides, Spintronic Computers will also hit their limit in about 8 to 10 years and are expected to overtake electronic computers at almost exactly that same time.

Updates

8 d

True Quantum Computers won't actually be better than Spintronic Computers for likely another 28 years, because they need to somehow be miniaturized from the size of a room to the size of a laptop during that time period.

[Jersey2]

I don’t know much about how they actually work beyond keywords such as entanglement etc. They seem specialized for certain tasks that most people currently have no need for in daily life, perhaps that will change with space tourism and travel needs although cosmic rays are an issue for these machines so that will need to be solved beyond just cooling them to near zero… specialized shielding would be an obvious first step.

My real question is will a new path open up making Quantum Computing a short lived side path like steam cars vs combustion engine cars. To push Quantum Computing to the end state you seek I think it would be need government money and research facilities to continue their work to make the inroads until solved or something better is uncovered.

So in short, I dunno. Lol. We humans just need to keep pushing it forward until the problem is resolved or something better is found. That is how we always have solved such problems.

[Jersey2]

In regards to your update, I heard the same crap about Moores Law 20 years ago. We always seem to find a way.

[Jersey2]

Ahh, you’re thinking about AI usage.

[SeekerOfTruth007]

We're down to below 1nanometer transistor gates. Atoms are 0.1nm, so there is no way to get transistor gates smaller than one atom. So, we really are going to hit the real physical limit to both electronic and spintronic computers in about 8 to 10 years, or less.

... Unless someone figures out how to make transistor gates out of something besides atoms.

[SeekerOfTruth007]

Quantum computers are only "better" than classical computers if the problem can be framed as one or more types of Quantum Algorithms. Otherwise, they are no better.

What will probably be the case is classical computers will query a few quantum computers strategically placed around the world. We'll probably see them used in Video Games more often than real-world problem solving.

[SeekerOfTruth007]

Actually, I'm most interested in Weather forecasting, specifically for forecasting Hurricane Track and especially intensity, as well as tornado likelihood in the 24 to 48 hours range.

Global Weather models currently use 3km x 3km horizontal resolution plus 64 layers in vertical resolution. If we improved computers by a factor of another 32, that would allow 0.75kmx0.75km horizontal resolution while doubling vertical layers to 128.

Current "hurricane models" use 2kmx2km resolution and 64 layers. If we could improve computers by a factor of 32 (5 more doublings of Moore's Law,) then we could have 0.5kmx0.5km resolution and 128 layers in the hurricane model. This would greatly improve both track and intensity forecasts.

[Jersey2]

I remember with regard ti Moores Law that there was a physical limit to the amount that could be stored on a chip but then they developed a new layering technique that changed everything.

Yes, a lot of variable to co sider when it comes to weather. We have so many models for hurricanes and yet they can be quite different in paths. The European model seems to be one of the better ones. What’s interesting is that the need for more data collection will be needed as well then to improve accuracy. Not just satellites which can produce amazing information but also ground and sea data from sensors. It could save lives if people actually listened to the warnings. It could be used on other planets like Mars and Jupiter as well.

Anyway, good luck with that.

[SeekerOfTruth007]

Satellites will need to improve to 0.5km resolution for infrared. Space-based radars will need to improve to 0.5km resolution as well, and we'll need more of them. I have in the past proposed that we have polar orbiting radars and infrared satellites on every 15 degrees meridian, so that we can track the atmosphere much more effectively. Currently, we only have one functional polar orbiting space-based radar, and it is around 15 years old, so very outdated compared to what is physically possible. Our Geostationary satellites are also outdated, and could use some updated replacements. I'd like to see multi-spectral satellites capable of scanning in Terahertz band in addition to the existing bandwidth, to better detect dust and rain drops inside storms, instead of just what we are doing now.

[SeekerOfTruth007]

The return to "layering" nano-components is limited due to heat waste concerns. A 3-d space radiates heat waste very poorly compared to a 2-d surface, so there is a very finite limit to how many layers you can use, even when using liquid coolant, such as water (normal) or Liquid Nitrogen (very impractical).

[Jersey2]

There will need to be a business case for your proposal. Given competing priorities I am not sure they will fund it in the near term. Much of the population and trillions in investment are on the East Coast. Perhaps that is a start but what practical difference will a more accurate model make? But beyond that, the current capability of the one polar satellite is not only out of date but is at risk of failing by the time we get a new satellite or set of satellites launched. So the time is now as it will take years to get a new one in orbit. Luckily SpaceX has made launches so much cheaper than when the current one was sent up.

[SeekerOfTruth007]

Further reduction in excess evacuations would pay for itself in the lifetime of the system of satellites, just as the previous generation did. If we could narrow the evacuation zone down even more for hurricanes, it would save billions per year on average in reduced "excess evacuations". It's worth it. We just need the funding from Congress or one of these tech giants might be able to fund it independently.

[Jersey2]

It could also increase compliance with evacuation orders and many people ignore them as they have moved inland in the past only to have it be a non-event where they lived. They will only do that so many times and then ignore them warnings.

It can also help in keeping supplies like wood to board up houses with in the areas that need it instead of everyone trying to board up using all the limited supply and leading to price gouging. The same with gasoline for people trying to evacuate and hotels for those that actually need them.

Electric companies could be a valuable Allie in this fight with Congress. They have lobbyists and a vested interest. They need to prepare for the post hurricane problem of fixing the grid and they need to best preposition their trucks and crews. Same can be said of State Police, hospitals (evacuations and providing supplies to), gas stations / Gas Companies, Fire/Rescue, National Guard, Insurance companies for assessors etc, road and bridge closures, much more.

Get your allies together who have money and pull in Washington.

[jshm2]

Yeah, but they're largely useless... for now.

The scale, complexity, power requirements and costs are just not worth anything going on.

Hence why photonic computing is a much more viable option instead.