Diagram from towarddatascience.com for more detail on this complex topic see https://www.towarddatascience.com
Bloomberg Businessweek April 13, 2020 Technology “Is This The Breakthrough Quantum Computing Has Been Waiting For?” “A secretive team in Palo Alto says it’s working a device several orders of magnitude more powerful than Google’s or IBM’s.
Companies Mentioned
PsiQuantum
Honeywell
IBM
GlobalFoundries-Chipmaker
Playground Global-Investor in PsiQuantum along with others including Atomico, Founders Fund, Redpoint Ventures, BlackRock Advisors, and Microsoft
What is the promise of quantum computing?
Reduce computing time, for even the most complex calculations, from decades or years to minutes. Hope being that this level of speed will facilitate new insights impossible with transistor-based computing. Two examples of problems that computers can't solve currently are prime factorization of large numbers and most molecular modeling. The former is key to data security and the latter is key to understanding molecular biology and applying solutions for biopharma etc.
How is quantum technology different than traditional super computers?
Super computers, that fastest currently, still run on transistors-binary switches that are at any one moment either on or off, a bit, a 1 or 0. The transistors are grouped and computing instructions are strings of 0 and 1; 011101100101010100010001 etc.
Quantum computers are based on quantum mechanics and therefore use light or Photons-single particles of light. Something all, including Einstein, take on "faith" is that qbits are both 1 and 0 simultaneously-a property known as superimposition. Think of a string between points a and b and think of that string being a very fast wave in all dimensions. The point is not to measure a qbit but if you could halt the light the qbit would be either 1 or 0 just like a classic transistor. qbits interact with other qbits creating entanglement-once two quantums systems interact they are forever entangled, regardless of how far they are apart-called by Einstein "Spooky action at a distance". If qbit A has an up spin then qbit B, it's entangled partner, has a down spin and vice versa at any point in time and space. Instead of transistors, think of Entangled qbits grouped as holding instructions and that adding qbits to a Quantum Computer gives rise to exponentially increasing possibilities or paths (See above N qbits 2^N paths [read 2 to the power of N {an exponential function}]) . Somehow, nobody really knows, the thinking goes that, I am paraphrasing-the full spectrum of possible computational outcomes are evaluated simultaneously and the best solution is found-in a single operation. The more qbits the greater the computing power only if the qbits behave reproducibly or give the same accurate function. Having a low error rate is key to having usable computing power. Another term dubbed Quantum Volume refers to the number of useful qbits.
Is this for real?
Not even easy to say but harder to understand, so "show me" we might say! IBM and Google have demonstrated Quantum Computing. Most recently, "...Google showed that its 54 qbit machine needed only three minutes to perform a calculation that would take a traditional supercomputer 10,000 years”. PsiQuantum claims they have achieved a 1,000,000 qbit Quantum Computer capable of general uses, not just designed for a answering a specific question, that with error correction yields 100-300 “useful” qbits. PsiQuantum notes, of the Quantum Computing "arms race", that Google’s feat was accomplished with a device specifically for Google’s test-calculation rather than general use.
PsiQuantum founded in 2015 has gathered more than $200M in funding (See Above), currently has 100 employees in Palo Alto but just now has given it’s first notable interview (This BBW Piece). “It’s also declining to publish academic papers that can be reviewed for their merits or to let outsiders evaluate it technology”. Regardless…”a working device [from any entity] is still likely years away”.
Kommentare