Quantum state of the Union 2022

Its attention-grabbing to keep watch over Quantum computing so I believed to make a break down of what attention-grabbing from the [IBM quantum state of the union 2022](https://www.youtube.com/watch?v=nZu5hutqANk&ab_channel=IBMResearch)

Fast rundown of why is that this attention-grabbing relating to bitcoin:

1. One of many hearts of bitcoin is the sha-256 hashing algorithm. This can’t be cracked by classical computer systems even when Moors regulation stays in impact. Nevertheless for a quantum pc which have – and is ready to make the most of – a single CPU with ~1.9 billion logical qubits (so about 8-12+ instances that in bodily), this turns into trivial.
2. As soon as this, if ever, occurs its not simply that bitcoin is fucked, however all the pieces else you may picture within the digital world can be fucked (the truth is, waaaay extra fucked for the reason that assault vector might be lots broader). Because of this there are already exists a [suit of cryptographic](https://www.ibm.com/docs/en/zos/2.5.0?subject=cryptography-crystals-kyber-algorithm) options to modify to. There are two issues although. One is that the ensuing hash is 4+ instances bigger then sha-256. The opposite is that they clearly can’t be stress examined till now we have such a machine.

**So the place we’re at?**

IBM simply launched they new Osprey Quantum CPU with a whooping 433 bodily qubits! Relying on how a lot error correction you might be capturing for that is about 36-52 logical qubits. The factor is the dimensions of a frying pan, you have to cool it with a proprietary, hand constructed system the dimensions of a freezer room.

To make use of it, you need to break down your very particular downside into elements, the place a few of it may be fed to the Quantum pc, the remainder must be executed by classical computer systems. For our sha-256 instance it will want classical computer systems for orchestration and a shitton of ASIC-s for hash technology as effectively.

**Why is that this tech so gradual?**

First off, watch the video. Even when the techno bla-bla goes over your head, the final feeling of complexity wont.

1. [Because it doesn’t work on room temperature](https://youtu.be/nZu5hutqANk?t=2085). To make the quantum processes even potential, you need to cool shit to close absolute zero. Then you have to eliminate all the warmth that course of generates. Its the identical downside as a btc-miner simply cranked as much as eleven, if you happen to cool on one finish it has to warmth up on the opposite finish. The infrastructure to do that is brutal. Each in dimension and power consumption. Proper now they work on enormous car-sized fridges to chill all the pc part as an alternative of particular person cpu-s.
2. [Then we have error correction](https://youtu.be/nZu5hutqANk?t=737). Quantum computer systems are totally different from classical ones. Actually silly instance: As an alternative of 1 and 0, you’ve got a “curve” of likelihood of state for every qubit. You employ extra then one qubits (8-12) for a similar state to have a strong thought what it represents, the extra you utilize the higher thought you’ve got, however some room for “error” all the time stays. Now qubits, identical to transistors, work in mixtures to characterize one thing of worth. For instance the decimal quantity 13 is 4 transistors flipped to 1 1 0 1 together. You may also “flip” 4 qubits to do one thing however these little remaining error prospects accumulate. Exponentially. They attempt to mitigate this with utilizing classical computer systems however the issue stays: You’ve got a linear progress trajectory of variety of potential qubits within the system, mixed with an exponentially rising error correction want.

**So the factor the place they break sha-256 in 10 years?**

Yeah… except some earth shattering innovation occurs, its a Nope. For those who watch it, its clear they’re already planning to make this right into a multi-processor, modular system, for fixing smaller, low qubit requirement issues in tandem with classical computer systems. Precisely as a result of simply growing the qubits on one CPU is bodily prohibitive. I discover it onerous to picture that this tech goes to “sneak up” on anybody relating to cryptography. If they can double the only CPU qubits yearly (which it appears will not be even deliberate after Condor) they’d get to the theoretically required 19 billion bodily qubits by round ~2047. Lets see.