New AMD Patent Proposes Teleportation to Make Quantum Computing More Efficient

[Lightwreather]

Summary

A team of researchers with AMD have filed a patent application that looks toward a more efficient and reliable quantum computing architecture, thanks to a conventional multi-SIMD (Single Instruction Multiple Data) approach. 

Credit: Shutterstock

Quotes

Quote

According to the application, AMD is researching a system that aims to use quantum teleportation to increase a quantum system's reliability, while simultaneously reducing the number of qubits necessary for a given calculation. The aim is to both alleviate scaling problems and calculation errors stemming from system instability.

There are two major hurdles on the road to quantum development and eventual quantum supremacy: scalability and stability. Quantum states are a fickle matter, so sensitive that they can decohere at the slightest provocation -- and a quantum system's sensitivity tends to increase with the presence of more qubits in a given system.
The AMD patent, titled "Look Ahead Teleportation for Reliable Computation in Multi-SIMD Quantum Processor," aims to improve quantum stability, scalability, and performance in novel, more efficient ways. It describes a quantum architecture based on quantum processing regions: areas of the chip that hold or can hold qubits, lying in wait for their turn on the processing pipeline. AMD's approach aims to improve on existing quantum architectures by actually reducing the number of qubits needed to perform complex calculations -- via the science fiction-esque concept of quantum teleportation.

AMD's design aims to teleport qubits across regions, enabling workloads that would theoretically require in-order execution to become capable of being processed in an out-of-order philosophy. As a quick refresher, in-order execution features dependencies between one instruction and the next, meaning that a workload has to be processed sequentially, with later steps dependent on the previous step being fully processed and its result being known before the chip can move ahead with the computation. 

AMD's patent also includes a look-ahead processor embedded into the architecture,
tasked to analyze the input workload, predict what steps can be tackled in parallel (and those that can't), and appropriately distribute the workload across qubits, using a quantum teleporting technique to deliver them to the required quantum processing, SIMD-based region. 

How this quantum teleportation occurs isn't described in the patent -- it looks like AMD is playing its cards close to its chest on this one. Perhaps we're looking at the beginning of AMD's "Zen 98" design (that's a joke, to be clear), or maybe this will not materialize in an actual product. But it does show, beyond any doubt and surprising no one, that AMD is indeed working on quantum computing. That seems to be the next great computing race. And while AMD may or may not be backing the right horse to ride toward an eventual victory, at least it seems the company does plan to be a part of the race.

 

My thoughts

Well, those are two words I did not expect to hear in a sentence together but here we are. AMD's (at least thinking of) working on quantum computing chips. Now as well all know (right?) that Patents don't always translate to a real product  but there remains the possibility. Now on this tech, does this "teleportation" refer to quantum tunneling or what? But anyway, this is pretty interesting and I hope that we see some progress on this in the future.

Sources

Tom's hardware

Freepatentsonline.com

[Salv8 (sam)]

53 minutes ago, J-from-Nucleon said:

Now on this tech, does this "teleportation" refer to quantum tunneling or what?

from what i understand it's more like the universe's version of wifi, being able to transfer quantum information (like the state of a quark) from one thing to another regardless where they are located. thats the best way i can explain it.

my guess is that AMD wants to have a way to connect to a quantum computer without cables to allow for superfast and instant data transfer from the quantum computer directly into the CPU, GPU or some kind of APU on a server to allow the data to be processed quicker.

[Roswell]

1 hour ago, J-from-Nucleon said:

Summary

A team of researchers with AMD have filed a patent application that looks toward a more efficient and reliable quantum computing architecture, thanks to a conventional multi-SIMD (Single Instruction Multiple Data) approach. 

Credit: Shutterstock

Quotes

 

My thoughts

Well, those are two words I did not expect to hear in a sentence together but here we are. AMD's (at least thinking of) working on quantum computing chips. Now as well all know (right?) that Patents don't always translate to a real product  but there remains the possibility. Now on this tech, does this "teleportation" refer to quantum tunneling or what? But anyway, this is pretty interesting and I hope that we see some progress on this in the future.

Sources

Tom's hardware

Freepatentsonline.com

 

33 minutes ago, Salv8 (sam) said:

from what i understand it's more like the universe's version of wifi, being able to transfer quantum information (like the state of a quark) from one thing to another regardless where they are located. thats the best way i can explain it.

my guess is that AMD wants to have a way to connect to a quantum computer without cables to allow for superfast and instant data transfer from the quantum computer directly into the CPU, GPU or some kind of APU on a server to allow the data to be processed quicker.

Quantum teleportation is related to quantum entanglement. Basically, if you create two identical particles together, they can behave the same at any distance and at any point in time. 

 

This doesn't enable any type of 2-way communication at all, just the knowledge that one particle entangled in one location is behaving in the same way that its counterpart is in another location. 

[Rauten]

I may be completely wrong but, I think this is the first time that a "traditional" computing company has shown such a high degree of interest in quantum computing?

Have Intel or IBM ever involved themselves this much in quantum computing?

[porina]

48 minutes ago, Rauten said:

I may be completely wrong but, I think this is the first time that a "traditional" computing company has shown such a high degree of interest in quantum computing?

Have Intel or IBM ever involved themselves this much in quantum computing?

IBM seem to have a quantum computing system you can buy today if your pockets are deep enough, so I'd say that is some interest on their part.

 

https://www.anandtech.com/show/15215/intel-launches-horse-ridge-chip-for-quantum-computing-systems

Above is a product Intel make which is a component that can help implement a quantum computer. I think it safe to say they're doing a lot more behind the scenes like AMD are.

[williamcll]

Too bad quantum teleportation doesn't break the speed of light.

[Mark Kaine]

48 minutes ago, williamcll said:

Too bad quantum teleportation doesn't break the speed of light 

 

"The speed of light is the ultimate speed limit.“

German autobahn: hold my beer

 

 

Also:

Maybe it's something like that?

 

 

Thats besides I'm not sure why anyone would want to go faster, at least until we figured out the braking thing ...

 

But, nah, in quantum physics Einsteins theories aren't set in stone,  in fact quantum phyiscs often go against them, and everyone is just like "whoops!" ...

 

 

 

 

[williamcll]

8 minutes ago, Mark Kaine said:

 

"The speed of light is the ultimate speed limit.“

German autobahn: hold my beer

 

 

Also:

Maybe it's something like that?

 

 

Thats besides I'm not sure why anyone would want to go faster, at least until we figured out the braking thing ...

 

But, nah, in quantum physics Einsteins theories aren't set in stone,  in fact quantum phyiscs often go against them, and everyone is just like "whoops!" ...

 

 

 

 

just gonna have to smash the subatomic particles harder so they break into even smaller pieces

[Mark Kaine]

4 minutes ago, williamcll said:

just gonna have to smash the subatomic particles harder so they break into even smaller pieces

I think a lot of it is down to the "observer effect" they measure something... get the reaction "before" they even started it... "faster than light!"  But, there is a chance there is at least "something" here and quantum computing seems to look into these things more closely,  just too bad we can't game on it, or maybe just randomly generated dungeons kinda games...

[Mark Kaine]

6 hours ago, Salv8 (sam) said:

my guess is that AMD wants to have a way to connect to a quantum computer without cables to allow for superfast and instant data transfer from the quantum computer directly into the CPU, GPU or some kind of APU on a server to allow the data to be processed quicker.

No, this is about "out of order instructions" which is one of the issues with qbit computing,  which they're trying to solve, or improve. 

 

In other words quantum computers are very fast, but they don't work like traditional computers,  so they're good with a specific task, but if you want to use them in a more traditional way, they're actually really bad, because there is a certain amount of randomness or uncertainty  (uncertainty principle) and this aims to improve this situation.  And no i can't really explain out of order execution, it's the opposite of sequential execution obviously,  but beyond that its just super unintuitive and im not a programmer... (and i know for fact a lot of programmers have *huge* issues explaining it themselves lol)

 

Edit: here's more "quantum fun with veritasium" the "hippie prof" explains it really well, partly because derek asks the right questions... 

 

(its a bit old, the qbits stuff is still actual tho)

 

 

[TetraSky]

I won't pretend to understand anything about Quantum computing.

But I was not expecting AMD to be related to that in any way shape or form. Feels like it was always some random companies I had never heard of making stride in that segment of the tech world before. Or maybe they have been working on it for a long time and I'm just living under a rock to have never noticed. Likely that.

[leadeater]

1 hour ago, TetraSky said:

I won't pretend to understand anything about Quantum computing.

But I was not expecting AMD to be related to that in any way shape or form. Feels like it was always some random companies I had never heard of making stride in that segment of the tech world before. Or maybe they have been working on it for a long time and I'm just living under a rock to have never noticed. Likely that.

That's fairly typical as commercial companies are usually publicly traded so it's hard to justify very early research in to unknown areas just for the sake of exploration and science. Also because they are public for profit companies they aren't usually eligible for most research grants.

 

University Theoretical Research Studies -> University Advanced Data Modelling Studies -> Research Institution and/or Public + Private Partnership -> (typically way later) Startup Company -> (probably also way later) Existing Incumbent Commercial Companies

 

Above is not like "correct" but good way to think about it is very general terms as to why it's always companies you've never heard of. IBM is one of the more notable exceptions as they fit in on multiple different stages of the above, they do a lot of early advanced research. Not the only one but the first that comes to mind. Intel as well for their Foundry business branch.

 

[RejZoR]

We had everything VR few years ago. Now we have everything Ai. Prepare for everything being "Quantum"...

[leadeater]

4 minutes ago, RejZoR said:

We had everything VR few years ago. Now we have everything Ai. Prepare for everything being "Quantum"...

Been and gone, Quantum buzz happened like 2-5 years ago.

[RejZoR]

34 minutes ago, leadeater said:

Been and gone, Quantum buzz happened like 2-5 years ago.

It'll come back. VR was around in the 80's too an has returned and faded away several times since.

[porina]

4 hours ago, TetraSky said:

I won't pretend to understand anything about Quantum computing.

But I was not expecting AMD to be related to that in any way shape or form. Feels like it was always some random companies I had never heard of making stride in that segment of the tech world before. Or maybe they have been working on it for a long time and I'm just living under a rock to have never noticed. Likely that.

Looking at it from a different direction than Leadeater's reply, tech companies will have to look far forwards to ensure they keep moving forwards. They don't just do one thing, but have a load of things going down the pipeline at various stages.

 

This might not be exactly how it is divided, but AMD might operate with different layers like:

1, manufacture and support of current products (Zen 3)

2, work on getting next gen (Zen 4?) design to be production ready. Basic design would be settled long ago, and they are probably working manufacturability and making adjustments as needed.

3, work on the basic design of the gen beyond that (Zen 5?).

4, look at what known or unknown technologies might go into generations beyond that (Zen 6+?)

 

Quantum computing is one of those things that is far from consumer level, but if you're not looking at it now, you're going to be left behind or have serious catch up to do if/when it does take off. It might not even go into a Zen series CPU but something else on the side. Companies don't like to talk too much about what they're working on in the far future as it may tip off what they are looking at to competitors. That doesn't rule out if they have something that is potentially valuable, it can be protected through patents.

 

To me it is a positive sign that AMD have the resources to look into these areas. It is always a bit of a gamble how much resource you allocate to far future technologies.

[BuckGup]

8 hours ago, Rauten said:

I may be completely wrong but, I think this is the first time that a "traditional" computing company has shown such a high degree of interest in quantum computing?

Have Intel or IBM ever involved themselves this much in quantum computing?

No Microsoft has been heavily into the quantum world for a while now and even released Q# a couple years ago

[Doobeedoo]

Going even over greater beyond!

[Mark Kaine]

3 hours ago, leadeater said:

That's fairly typical as commercial companies are usually publicly traded so it's hard to justify very early research in to unknown areas just for the sake of exploration and science

I mean that's usually probably the case, but with qbit computing i just can't see the big players not investing and researching in it... moores law is deader than dead and they will need something else, and quantum bits just seems like the most likely and logical thing. 

Imagine intel would come up with a new "qbit cpu" that dwarfs even 2nm ones and amd had nothing to show... these guys can't afford not to invest in this imo (as this patent shows,  although they might just have pulled a lets patent a physical law, because why not... )

 

these companies have relied so long on moores law, they must be aware its coming to an end... (the qbits prof estimated around 2025, but I kinda think we're already past, shouldn't we have 2nm or smaller* chips by now as standard already? I know its lagging behind a lot, especially intel) 

*which is afaik simply impossible with current tech,  precisely because of the atoms jumping around uncontrollably due to quantum physical mechanics at a certain scale 

 

3 hours ago, leadeater said:

Quantum buzz happened like 2-5 years ago.

I remember reading about this 30 years ago as the next new best thing... its like nuclear fusion of compute tech, only ~2 years away now!

 

^Ironically both relying on the same tech, super conductors... 

 

[leadeater]

1 hour ago, Mark Kaine said:

moores law is deader than dead and they will need something else, and quantum bits just seems like the most likely and logical thing. 

It's really not though, we're actually a long way off from encountering the real problems with that "law".

 

1 hour ago, Mark Kaine said:

I mean that's usually probably the case, but with qbit computing i just can't see the big players not investing and researching in it

I didn't say they wouldn't, I said they are the last in the chain to enter the research of something like this. AMD is only just getting in to it now, where everything before hand has been essentially what I outlined. There's plenty of research institutions looking in to practical Quantum computing as well as startup companies but as yet IBM is the only big player really doing much at all. Not even Fujitsu who would be my next top expected, HPE bailed out (they aren't doing anything meaningful right now).

 

Purely exploratory research with next to no commercial gain just isn't something these companies do, nor should they.

[leadeater]

1 hour ago, porina said:

Quantum computing is one of those things that is far from consumer level, but if you're not looking at it now, you're going to be left behind or have serious catch up to do if/when it does take off.

Just an extra side note on this, they often also buy out startup companies or rights to patents or are already in partnership with those working on it and then move it in to in house development once it's seen as a more promising not so extreme distance future thing.

 

It's a good way to reduce investment loss risk and have "bad investments" on the books so to speak. Let someone else fail first for you then maybe pickup the pieces or buy out the one that succeed. Technology landscape is littered with buyouts for IP etc.

[leadeater]

1 hour ago, Doobeedoo said:

Going even over greater beyond!

This isn't even my final form!

[Mihle]

Wonder of there will ever be quantum computer that don't require temp close to minium.

 

Also wonder if we will ever have normal CPU/GPUs running on something else than Silicon in my lifetime.

 

 

[Lightwreather]

18 hours ago, williamcll said:

Too bad quantum teleportation doesn't break the speed of light.

If this is related to quantum entaglement,

then it'll probably be faster than the speed of light, maybe......

[CarlBar]

8 hours ago, Caroline said:

Oh not again

To be fair if you could manipulate quantum effects to make very unlikely likely there really is a lot of utterly weird stuff you could do. Now figuring out how or even if it's possibble to o a lot of that is a mess.But as excuses go it's probably not that bad a choice.