I really need a quantum computer for browsing firefox with my dark reader extension enabled, I'm willing to pay ANY price... I need multiple youtube tabs open and pronto! let's get this to market! xD
my first computer was a Sinclair ZX81 and my first job i worked on an IBM System 34, 8inch disks and an addon rampack that was the size of a fridge… and now we are here, so yes, its quite impressive :)
How about 1 trillion qubits made using graphene 270mm² microprocessor size made using nano technology working in room temperature achieving quantum coherence by nanotechnology isolation
I love how the decided 10,000 years wasn't long enough for comparison, NAY! we need septillions of years! let's reach for the stars themselves and beyond xD
Each R&D, in order to be granted the next funding round, must either present a product or promise an earlier product, even if the product never materializes.
We've been hearing about quantum being just around the corner for the last 20 years, why should we be anything but extremely sceptical at this latest announcement? Notably absent from this announcement is any hint of practical applications or practical results.
They speak at length in their blogpost on the topic how this explicitly has no practical applications yet. It is a significant and incredible advancement on the state of the art. I'm sorry it can't toast your bread quicker yet
"The numbers all go to eleven. Look, right across the board, eleven, eleven, eleven. It's one better, isn't it? It's not ten. You see, most blokes, you know, will be computing at ten. You're on ten here, all the way up, all the way up, all the way up, you're on ten on your computer. Where can you go from there? Where? Nowhere. Exactly. What we do is, if we need that extra push over the cliff, you know what we do? Put it up to eleven. Exactly. One better."
I actually wrote my MSc thesis on the state of Quantum a couple years ago (admittedly from a business perspective rather than a physics one). The general consensus amongst academics and professionals is that we will start seeing quantum use cases sometime around 2030-2035. The reality is that the quantum is never going to replace classical computing. It’s only going to be useful at quite specific mathematical computations. So don’t expect to see Quantum computers appearing at scale anytime soon. That’s unlikely to ever happen.
I don't imagine they'll replace classical computers, since there's things classical computers are better at than quantum computers, but imagine a future where some combination of AI and quantum computer simulation allows for us to find the holy grail: a room temperature superconductor. At that point, I feel like scaling quantum computers becomes much easier and much more rapid. Given some time, I could see a future where classical computers come with a QPU or Quantum Processing Unit, similar to GPUs in today's computers. They would be especially useful for AI, and given that it looks like AI is here to stay, that could mean a paradigm shift. I think AGI specifically might become possible if this scenario where we can massively expand quantum computers. The ability to process multiple things at once, to quickly search unorganized and chaotic data, etc may unlock new aspects of AI that aren't in reach today. The future of computing seems as exciting as it is quickly approaching!
@@AbadonBIack AI right now is a marketing buzzword for a really fancy auto-correct. I am sure one day we will reach levels of AI that will be able to do the things we think it can, but current LLMs based on transformers will not be able to reach that level, IMO. I could be wrong of course, but seeing as how chatGPT is being touted as the solution to my problems but so far it has only been able to draft my emails for me (poorly, as of late), like a glorified outlook template, I'm not holding my breath. Companies that laid off workers to try out AI solutions had had to re-hire some of the people I know personally, and had to increase salaries to entice them back lol
Also writing a thesis on Quantum Computing and QEC theory, theoretical physics perspective. We already have Quantum Computers at scale, though quite error prone, and it was never intended for them to replace classical computers only to work with them in large super computer applications.
optimization problems, search problems, simulating other quantum systems, yeah.. things like gene sequencing, material sciences, machine learning.. a lot of cool stuff 😮
Random numbers. That’s literally it. They are analogue random numbers generators, the reason they are fast is because ‘classical’ computers are deterministic and can’t do randomness easily.
YOU ARE google you should be able to translate this video in all language, i mean perfect si titled and using sound, not only subtitle. Your new challenge is to use quantum computing to find targets and cure for Myalgic Encephalomyelitis disease.
What’s the point of making extremely rapid calculations if the result is wrong or you don’t know if it is wrong. When they solve the error correction problem then quantum computers may have some practical use. That is not going to happen any time soon.
Quite inspiring your work amazing technology congratulations! Just a quick note, "Main street" people understood long time ago that technology no matter how powerful will become is not able to solve their problems since always is used to empower even more the few. so your work is really nice to see how far humankind can get in terms of knowledge Not in terms of welfare.
"comercial value" vs ethical value it's exactly like firefighters putting "comercial value" before ethical/2+2=4 values plus efficiency value plus prevention value. which is symptoms of insanity/chaos(2+2=5) clearly.
Here's one of my favorite: In order to find cures and create medicines, they have to model molecules, chemical compositions, and the like and then just 'try it' over and over and over with differences each time until it works to combat or create what they need -- (this is a very simplified example, and I am no expert) and this takes a very long time and is super complex and power hungry. But, with quantum computing, they can model the same things with many different variations, account for changes, and do it at insanely fast speeds. Trial and error testing would happen more accurately and so much more quickly so we could get results and likely cure a lot of ailments or diseases in a much shorter time frame.
@@skeleton.wizard Yes, I get how quantum computing can do stuff like modeling molecules way faster and more efficiently, but it makes me wonder-if it’s so powerful, why isn’t it used for other general things like a regular CPU? Is it just because the tech is still new, or is there something about how quantum works that makes it more specialized?
one simple question, are you driving quantum-computer with a classical-computer? if yes? then there is a huge problem. it is not a quantum-computer, it is a quantum-processor, you have to adjust jargon accordingly.
Preskill is a total killjoy ahahahahaha Yeah knowing of the research in the field of QC/QEC I'm sure most agree a cautious optimism is most warranted, question is just how to best phrase that to investors
So those benchmarks are deliberately designed to be hard for a normal computer vs a quantum one. Nice test. Now excuse me while I declare C++ source code compilation speedrun to be the new IQ test.
This will not scale. Spacetime has limits. You can't pack too many coherent states into a small volume without applying exponentially large amounts of energy.
What about "Far Future Super /(Ultra)/Hyper Quantum Computer"?maybe Tetrational growth or Pentational Growth or Beyond that.if 1000 second or 1 million second = "Google"plex
5:16 Misleading figure, I think. The latest demonstration has 100 physical qubits (2019 had 54) that enables them to create 1 logical qubit using the surface code. If for useful quantum computation you need 10^6 logical qubits, then you would require 10^12 physical qubits. For context, the current WIllow processor has only 100 physical qubits. Also, this is a logical quantum memory, meaning it can store quantum states for a short amount of time with error correction. The paper does not include qubit operations.
They are planning to tile each of these logical qubits. So for example if Willow stores one logical qubit using 100 physical qubits, to have 10^6 logical qubits you might only need (10^6) * 100 = 10^8 physical qubits, not 10^12 physical qubits (unless I'm missing something?). But Willow has an error rate of 1 in 1000, and to achieve an error rate of 1 in a trillion (which is their stated goal), with the current architecture, my rough calculations say that they'd need ~100 billion physical qubits for 1 stable logical qubit. If they can do that cheaply, to have 10^6 logical qubits will just involve scaling up production and tiling them together.
Willow Quantum Chip 2024 10^25 Years of Classical Supercomputer Problem-Solving in Just 300 Seconds Sycamore Quantum Chip 2024 10000 Years of Classical Supercomputer Problem-Solving in Just 200 Seconds
lol, a selected benchmark that favors the Quabtum computer. Now let's take a benchmark like the 3D mark, a PC will be million times faster. Hah! See it goes both ways.
Still haven't personally seen any interesting quantum programs. Not one. Not saying it will never happen - I've just never seen one. Wake me up if you see one. Meanwhile I keep on coding on my C64 : )
I'd encourage you to look into Quantum Computing applications in molecule simulation, encryption breaking, optimization problems and various modified machine learning protocols. There is promise but it's hard to make a fault tolerant Q computer capable of realizing the corresponding algorithms.
Did you run probabilistic algorithms on the classical hardware when comparing the RCS on the quantum hardware? If not, running a deterministic on classical hardware is not a fair benchmark imo
Don’t worry.. it is good reason this came across meaningless, as they are completely hyped and without any applications in the foreseeable future. They might crack encryption for really small numbers that no one encrypts with anymore, and that’s about it.
So that LLM can possibly predict every possible questions you can make and prepare all possible set of solution it can answer and then also you can cross question and so on... and share a response which you might not be even able to understand :)
Just don't wast it to make batteries better, use it to create some resource efficient way to convert energy directly from the quantum foam or plank field, skip fusion.😮
With that much computing power we are on the verge of being able to have 5 open Chrome tabs with zero hit to performance.
This has nothing to do with chrome but appreciate the satirical attempt.
I really need a quantum computer for browsing firefox with my dark reader extension enabled, I'm willing to pay ANY price... I need multiple youtube tabs open and pronto! let's get this to market! xD
@@aaronbiederman1256🤓☝️
Does this mean we get more ads on TH-cam?
The ads are there to annoy you to buy premium.
Premium is worth it. You get music streaming also.
10 times more ads
done use pie
WHAT A TIME TO BE ALIVE!!!!!!
my first computer was a Sinclair ZX81 and my first job i worked on an IBM System 34, 8inch disks and an addon rampack that was the size of a fridge… and now we are here, so yes, its quite impressive :)
Hold onto the papers 📖📑
@@nyxstrix0 Just two more papers down the line.
Relaxxxx
@@nyxstrix0😅
using quantum computers to simulate quantum systems at useful sizes and speeds is the most exciting application of them.
how about using AI to regulate the errors.
How about 1 trillion qubits made using graphene 270mm² microprocessor size made using nano technology working in room temperature achieving quantum coherence by nanotechnology isolation
Incredible. That's one of these moments that people will read about in 100 years.
People no longer read today. Who will still read 100 years from now?
people read and learn much more than they did 100 years ago - just in more diversified ways. Not to mention the pace of generating content.
I love how the decided 10,000 years wasn't long enough for comparison, NAY! we need septillions of years! let's reach for the stars themselves and beyond xD
Wow! Much respect to the team!
That is great.
What warms my heart the most is that he is actively thinking how to make people on main street happy...
We are in safe hands indeed
Each R&D, in order to be granted the next funding round, must either present a product or promise an earlier product, even if the product never materializes.
We've been hearing about quantum being just around the corner for the last 20 years, why should we be anything but extremely sceptical at this latest announcement? Notably absent from this announcement is any hint of practical applications or practical results.
if this is real, we will need at least a year for real proofs and use to be announced, that's not simple as "train a model and launch"
They speak at length in their blogpost on the topic how this explicitly has no practical applications yet. It is a significant and incredible advancement on the state of the art. I'm sorry it can't toast your bread quicker yet
Time will filter all the BS, we need to wait and see.
It really says a lot about you that you think it's your approval they are after.
We should judge the technical achievements based on their merits. Of course, that requires a level of understanding.
A Sergio se le ve muy centrado, sabe lo que quiere y sabe como conseguirlo. Es uno entre 100000000
Get back to me when you can crack 2048 RSA
legit
they wont tell you when they actually can
The second they can do that the NSA is going to swoop in and confiscate everything.
Not happening for at least a decade more.
NSA: Hold my beer.
"The numbers all go to eleven. Look, right across the board, eleven, eleven, eleven. It's one better, isn't it? It's not ten. You see, most blokes, you know, will be computing at ten. You're on ten here, all the way up, all the way up, all the way up, you're on ten on your computer. Where can you go from there? Where? Nowhere. Exactly. What we do is, if we need that extra push over the cliff, you know what we do? Put it up to eleven. Exactly. One better."
Creating a dataset of the 𝟺₂₁-polytope would be a step forward in the research of new forms of matter.
The AIs will find this very useful.
This universe is wild.
the universe is as it is, we label it as wild because of our inadequate brains
This Simulation is wild.
Wild is ❤
Will chrome still lag on these?
They are so enjoyable! As a software engineer, I hope I can join your team in the future
Im so excited for this guys!!! Congrats!!!
what computations can you do?
Simulating reality in ways impossible before.
none that are useful
(You aren't supposed to ask that 😅)
You can quite literally predict the future
So what was the computation
by the beginning of the video, seems like the problem was not performance in 2019, what's holding up for us to have this level of permance now?
I actually wrote my MSc thesis on the state of Quantum a couple years ago (admittedly from a business perspective rather than a physics one).
The general consensus amongst academics and professionals is that we will start seeing quantum use cases sometime around 2030-2035. The reality is that the quantum is never going to replace classical computing. It’s only going to be useful at quite specific mathematical computations.
So don’t expect to see Quantum computers appearing at scale anytime soon. That’s unlikely to ever happen.
I love to see a voice of reason in the comments. Thank you :D
I don't imagine they'll replace classical computers, since there's things classical computers are better at than quantum computers, but imagine a future where some combination of AI and quantum computer simulation allows for us to find the holy grail: a room temperature superconductor. At that point, I feel like scaling quantum computers becomes much easier and much more rapid. Given some time, I could see a future where classical computers come with a QPU or Quantum Processing Unit, similar to GPUs in today's computers. They would be especially useful for AI, and given that it looks like AI is here to stay, that could mean a paradigm shift.
I think AGI specifically might become possible if this scenario where we can massively expand quantum computers. The ability to process multiple things at once, to quickly search unorganized and chaotic data, etc may unlock new aspects of AI that aren't in reach today.
The future of computing seems as exciting as it is quickly approaching!
@@AbadonBIack AI right now is a marketing buzzword for a really fancy auto-correct. I am sure one day we will reach levels of AI that will be able to do the things we think it can, but current LLMs based on transformers will not be able to reach that level, IMO. I could be wrong of course, but seeing as how chatGPT is being touted as the solution to my problems but so far it has only been able to draft my emails for me (poorly, as of late), like a glorified outlook template, I'm not holding my breath. Companies that laid off workers to try out AI solutions had had to re-hire some of the people I know personally, and had to increase salaries to entice them back lol
Thank you! A voice of reason!
Also writing a thesis on Quantum Computing and QEC theory, theoretical physics perspective. We already have Quantum Computers at scale, though quite error prone, and it was never intended for them to replace classical computers only to work with them in large super computer applications.
What do they calculate ?
optimization problems, search problems, simulating other quantum systems, yeah.. things like gene sequencing, material sciences, machine learning.. a lot of cool stuff 😮
@@GNARGNARHEAD no, what DID they calculate. Not what could they theoretically do when these things are actually useful.
Random numbers. That’s literally it. They are analogue random numbers generators, the reason they are fast is because ‘classical’ computers are deterministic and can’t do randomness easily.
@@womp6338that’s a good framing, thank you 😊
EVERYTHING
Can anyone specify what kind of computation ie what kind of problem was solved during the benchmark testing?
This has got to be alien technology to jump so fast from classical 2024.
Cures and Longevity around the corner
So next year Nobel and Oscar goes to Google😅
YOU ARE google you should be able to translate this video in all language, i mean perfect si titled and using sound, not only subtitle.
Your new challenge is to use quantum computing to find targets and cure for Myalgic Encephalomyelitis disease.
"One must be optimistic for great things to happen" Haha love it.
hartmut neven is the berlin final boss
Use this chip to make next generation one and so on
Now i will rewatch devs the series about quantum computer
crazy to imagine that humans can achieve such an achievement !
How much power does it consumes?
What’s the point of making extremely rapid calculations if the result is wrong or you don’t know if it is wrong.
When they solve the error correction problem then quantum computers may have some practical use. That is not going to happen any time soon.
Quite inspiring your work amazing technology congratulations! Just a quick note, "Main street" people understood long time ago that technology no matter how powerful will become is not able to solve their problems since always is used to empower even more the few. so your work is really nice to see how far humankind can get in terms of knowledge Not in terms of welfare.
"comercial value" vs ethical value it's exactly like firefighters putting "comercial value" before ethical/2+2=4 values plus efficiency value plus prevention value. which is symptoms of insanity/chaos(2+2=5) clearly.
So Willow is faster than Aurora?
you can't have sunglasses that cool and not wear them. gonna need you guys to reshoot and upload
Isn't that enough to put in to use now?
if I understand correctly, apparently not, but it s an important milestone...I guess you can think of it as a baby taking its first steps
Amazing!
Is it possible to apply as a research intern at Google?
NGL man, I dont know or understand where is this used for
Here's one of my favorite: In order to find cures and create medicines, they have to model molecules, chemical compositions, and the like and then just 'try it' over and over and over with differences each time until it works to combat or create what they need -- (this is a very simplified example, and I am no expert) and this takes a very long time and is super complex and power hungry.
But, with quantum computing, they can model the same things with many different variations, account for changes, and do it at insanely fast speeds. Trial and error testing would happen more accurately and so much more quickly so we could get results and likely cure a lot of ailments or diseases in a much shorter time frame.
@@skeleton.wizard Yes, I get how quantum computing can do stuff like modeling molecules way faster and more efficiently, but it makes me wonder-if it’s so powerful, why isn’t it used for other general things like a regular CPU? Is it just because the tech is still new, or is there something about how quantum works that makes it more specialized?
Crazy future is near!
how much willow would take for solving the same problem that took old version 200 seconds?
If it was much better, they wiuld have shown growth from old to new comp
@supastazz who knows🤷♂️
The comment section is so depressing.
Guys, you have forgotten the IBM solution to the 2019 problem that solved it much much faster than 10000 years, look it up.
one simple question, are you driving quantum-computer with a classical-computer? if yes? then there is a huge problem. it is not a quantum-computer, it is a quantum-processor, you have to adjust jargon accordingly.
Preskill is a total killjoy ahahahahaha
Yeah knowing of the research in the field of QC/QEC I'm sure most agree a cautious optimism is most warranted, question is just how to best phrase that to investors
That is mind boggling, sadly the time will come one day that AI realizes it no longer needs us, and it will kill us all.
where is API? for python, please
Are we not borrowing time from other parallel universes to do this? Does that not affect them?
Don't wallow in negativity. We'll release every quantum computing results, starting with Sycamore.
Thank you.
42?
don't panic!
What does that mean? 42 means Area 42?
@pratyushpralimpradhan2867 all means 42!
Google got a decillion dollar fine, but they are a septillion years ahead.
Haha😂
So can it test an infinite amount of passwords to unlock an email account or bank account?
"I've made a robot that screams"
Beautiful
So those benchmarks are deliberately designed to be hard for a normal computer vs a quantum one. Nice test.
Now excuse me while I declare C++ source code compilation speedrun to be the new IQ test.
This will not scale. Spacetime has limits. You can't pack too many coherent states into a small volume without applying exponentially large amounts of energy.
What about "Far Future Super /(Ultra)/Hyper Quantum Computer"?maybe Tetrational growth or Pentational Growth or Beyond that.if 1000 second or 1 million second = "Google"plex
This is how Skynet began
A Quantum Leap.
Thanks ☺️ Merry Christmas 🎄🎁 from Me 😊😊😊 .
Germans in America at the forefront of technological innovation. We are so back.
Call the creator of the Doom video game and the creator of Linux, they can create software and the director of NVIDIA and AMD can make the hardware.
This stuff seems so underwhelming unless you can actually use it the way classical computers are used
They probably said the same stuff when classical room-sized computers were first rolled out. Progress takes time
@@amellowblue this guy gets it.
@@amellowblue will take a lot longer to even reduce noise and have fault-tolerant systems
Here we go again now we gotta find a better encryption algorithm again.
Still not enough but good enough
5:16 Misleading figure, I think. The latest demonstration has 100 physical qubits (2019 had 54) that enables them to create 1 logical qubit using the surface code. If for useful quantum computation you need 10^6 logical qubits, then you would require 10^12 physical qubits. For context, the current WIllow processor has only 100 physical qubits. Also, this is a logical quantum memory, meaning it can store quantum states for a short amount of time with error correction. The paper does not include qubit operations.
So how many qubits did they reach now?
They are planning to tile each of these logical qubits. So for example if Willow stores one logical qubit using 100 physical qubits, to have 10^6 logical qubits you might only need (10^6) * 100 = 10^8 physical qubits, not 10^12 physical qubits (unless I'm missing something?).
But Willow has an error rate of 1 in 1000, and to achieve an error rate of 1 in a trillion (which is their stated goal), with the current architecture, my rough calculations say that they'd need ~100 billion physical qubits for 1 stable logical qubit. If they can do that cheaply, to have 10^6 logical qubits will just involve scaling up production and tiling them together.
Quantum computer 2024🔥❤️
Quantum superposition is not reality it's just apparent result of measurement.
The only practical use thus far has been to simulate a…Quantum System. Of which it *already* consists. Not exactly earth-shattering.
Willow Quantum Chip 2024
10^25 Years of Classical Supercomputer Problem-Solving in Just 300 Seconds
Sycamore Quantum Chip 2024
10000 Years of Classical Supercomputer Problem-Solving in Just 200 Seconds
Brilliant - Can we watch Cat's with it though ?
Im happy to see this,
ASI right around the corner
Crypto cooked though
We all know bench marks vs real world are two different things😂 probably no need to upgrade my pc
lol, a selected benchmark that favors the Quabtum computer. Now let's take a benchmark like the 3D mark, a PC will be million times faster. Hah! See it goes both ways.
Still haven't personally seen any interesting quantum programs. Not one.
Not saying it will never happen - I've just never seen one. Wake me up if you see one. Meanwhile I keep on coding on my C64 : )
What brand are the middle guy glasses? Swag
But i still gotta work…
Doesn't this prove the many worlds theory??!
Wildddd
Amazing achievement but please enlighten us how this can be applied to real world application instead of just benchmarking against models
I'd encourage you to look into Quantum Computing applications in molecule simulation, encryption breaking, optimization problems and various modified machine learning protocols. There is promise but it's hard to make a fault tolerant Q computer capable of realizing the corresponding algorithms.
Is it still just the RNG from then? If yes, that's the 'Hello world' of quantum computers. Better than nothing, but not worth the hype.
wow
quantum future
Welp, that's one way to waste some money 💰
Did you run probabilistic algorithms on the classical hardware when comparing the RCS on the quantum hardware? If not, running a deterministic on classical hardware is not a fair benchmark imo
So this is why bitcoin dropped today
Don’t worry.. it is good reason this came across meaningless, as they are completely hyped and without any applications in the foreseeable future. They might crack encryption for really small numbers that no one encrypts with anymore, and that’s about it.
Without application?
Ever heard of "AI"?
What they are gonna use it for ?
Wake me up when the quantum guys compute something useful. 💤
Can wait to run local 400b LLM to a Quantum Personal Pc 🤪 Good job Google!!!
So that LLM can possibly predict every possible questions you can make and prepare all possible set of solution it can answer and then also you can cross question and so on... and share a response which you might not be even able to understand :)
Just don't wast it to make batteries better, use it to create some resource efficient way to convert energy directly from the quantum foam or plank field, skip fusion.😮
అచ్ఛా
💖❤️💖
Yah, but when do I get this in my laptop?
just buy d-wave already