TSMC is Working on Quantum Computers

AlphaAtlas

[H]ard|Gawd
Staff member
Joined
Mar 3, 2018
Messages
1,713
Digitimes reports that Taiwan's Ministry of Science and Technology is "aggressively" promoting the development of quantum computers, and claims that Taiwan Semiconductor is the "only industrial player in the projects." IBM's Q cloud quantum computing platform is reportedly involved in the project, and the Taiwanese government is offering an annual subsidy of about $2 million USD to "each approved research program." As a reminder, Taiwan Semiconductor is arguably the most advanced chip manufacturer the world, and will reportedly be the only source for AMD's and Nvidia's 7nm products. Quantum computing, on the other hand, is more of a mixed bag, as this great writeup by IEEE explains. But TSMC hopes to advance the field by developing quantum computing chips on their advanced manufacturing processes.

MOST officials said that quantum computer can solve many problems that existing computers cannot address, and is also able to process extremely huge data. The next-generation computers may be most ideal for application to new-material design, cryptography, chemistry, biomedicine computer simulation and pharmaceutical sectors. Unlike CPU and GPU that require air or water for heat dissipation, quantum computer processor can operate well only under extremely low temperature of around -273 degrees Celsius, thus requiring the support of large-size freezing facilities. Accordingly, quantum computers can hardly be commercialized, unless semiconductor can be smoothly applied to enable quantum processors to operate under normal temperatures. This will be a major technological bottleneck that needs to be tackled to facilitate commercialization of quantum computers, the officials commented.
 
Wait... did I miss something, did D-Wave go belly up and no longer offer Quantum Computers? If IBM could get D-Wave and other key players into the fold that could potentially bring Quantum Computers to power users years if not decades earlier than if just TSMC and IBM go it alone.
 
The TSMC may or may not be working on quantum computing. The probability of them doing so is not zero.
 
Wait... did I miss something, did D-Wave go belly up and no longer offer Quantum Computers? If IBM could get D-Wave and other key players into the fold that could potentially bring Quantum Computers to power users years if not decades earlier than if just TSMC and IBM go it alone.

for all intents and purposes: the D-Wave offerings are not "real" quantum computers.

i believe TSMC is implying that a temperature of 0 (Lord) Kelvins is necessary to "slow down" the subatomic particles such that the states can be read/written (presumably using the more-powerful theory of the riemann sphere instead of the bloch sphere)

it will be easy for the initiated (mathematician; won't make this qualification again, it will be assumed) to see that the theory of the bloch sphere is much weaker than that of the riemannian sphere (projective geometry/space vs complex/real-valued geometry that can be *formulated* to fit in projective geometry).

really happy to see asus move away from mobile and my pals at TSMC looking at a very important issue.

i have confidence their team will get some real work done, unlike D-Wave.
 
for all intents and purposes: the D-Wave offerings are not "real" quantum computers.

i believe TSMC is implying that a temperature of 0 (Lord) Kelvins is necessary to "slow down" the subatomic particles such that the states can be read/written (presumably using the more-powerful theory of the riemann sphere instead of the bloch sphere)

it will be easy for the initiated (mathematician; won't make this qualification again, it will be assumed) to see that the theory of the bloch sphere is much weaker than that of the riemannian sphere (projective geometry/space vs complex/real-valued geometry that can be *formulated* to fit in projective geometry).

really happy to see asus move away from mobile and my pals at TSMC looking at a very important issue.

i have confidence their team will get some real work done, unlike D-Wave.

Whatever nerd, let me know when it can run Crysis.
 
Whatever nerd, let me know when it can run Crysis.

i know you're joking, but: we are a looooooong way away from that.

it's either we keep beefing up the registers of the CPU, or make a serious effort at quantum computing.

i haven't seen appetite for either, especially in the post 2008 landscape.

will be interesting now that TSMC is involved. hopefully this gets others off their ass.
 
It needs to be mentioned in every article talking about quantum computers that they will never replace the desktop computer. For one, they excel at only a narrow range of tasks, and secondly, most likely the tech will never be there to allow them to exist in a cheap home desktop solution.
 
It needs to be mentioned in every article talking about quantum computers that they will never replace the desktop computer. For one, they excel at only a narrow range of tasks, and secondly, most likely the tech will never be there to allow them to exist in a cheap home desktop solution.

Near future, sure. Never say never though, 60 years ago people probably said the same thing and here we are with building sized computational power in a thimble sized package.
 
for all intents and purposes: the D-Wave offerings are not "real" quantum computers.

i believe TSMC is implying that a temperature of 0 (Lord) Kelvins is necessary to "slow down" the subatomic particles such that the states can be read/written (presumably using the more-powerful theory of the riemann sphere instead of the bloch sphere)

it will be easy for the initiated (mathematician; won't make this qualification again, it will be assumed) to see that the theory of the bloch sphere is much weaker than that of the riemannian sphere (projective geometry/space vs complex/real-valued geometry that can be *formulated* to fit in projective geometry).

really happy to see asus move away from mobile and my pals at TSMC looking at a very important issue.

i have confidence their team will get some real work done, unlike D-Wave.
D-Wave is pretty much the current leader in the tech according to "Lockheed Martin, Google, NASA, USC, USRA, Los Alamos National Laboratory, Oak Ridge National Laboratory, Volkswagen, and many others" they have launched their cloud computing platform along side a free opensource dev kit for it. Supposedly people can register free accounts and get access to run their experiments on the hardware though I have not looked into it. With this they are hoping to get the Universities onboard as more than 300 of them are currently using the IBM Q systems and the 2000Q's are supposedly a few factors faster than them.
 
D-Wave is pretty much the current leader in the tech according to "Lockheed Martin, Google, NASA, USC, USRA, Los Alamos National Laboratory, Oak Ridge National Laboratory, Volkswagen, and many others" they have launched their cloud computing platform along side a free opensource dev kit for it. Supposedly people can register free accounts and get access to run their experiments on the hardware though I have not looked into it. With this they are hoping to get the Universities onboard as more than 300 of them are currently using the IBM Q systems and the 2000Q's are supposedly a few factors faster than them.

I just don't think you fully appreciate what a real quantum computer is, but that's okay. it takes time. no news article will articulate the difficulties and how these are realised in the machine. here's a good start to the issue i'm alluding to, which states it is no faster than a high-end Intel cpu performing the simulated annealing task.

edit: i don't know why you shared that quote. it's like you're saying the big players in the "military industrial complex" are honest and know what they're talking about, which is not necessarily the case *here*. the MIC has fucked up enough stuff to have earned joe schmoe's reserved judgement.
 
I just don't think you fully appreciate what a real quantum computer is, but that's okay. it takes time. no news article will articulate the difficulties and how these are realised in the machine. here's a good start to the issue i'm alluding to, which states it is no faster than a high-end Intel cpu performing the simulated annealing task.

edit: i don't know why you shared that quote. it's like you're saying the big players in the "military industrial complex" are honest and know what they're talking about, which is not necessarily the case *here*. the MIC has fucked up enough stuff to have earned joe schmoe's reserved judgement.
I copy and pasted it from another site because I was too lazy so that's why its in quotes, but yes nobody has gotten their quantum computers to the point where they are faster than their traditional CPU equivalents but they do come in cheaper to operate and they are getting faster so possibly in another 5 years we may see them faster than a traditional CPU for specific task loads. Point is TSMC see's a future where the x86 and ARM processors can't compete for that market set and they want to get in on it, in another 10+ years it could be a very lucrative market, and if the Taiwanese government is willing to hand out funds to get competitive in that space then all the power too them.
 
I copy and pasted it from another site because I was too lazy so that's why its in quotes, but yes nobody has gotten their quantum computers to the point where they are faster than their traditional CPU equivalents but they do come in cheaper to operate and they are getting faster so possibly in another 5 years we may see them faster than a traditional CPU for specific task loads. Point is TSMC see's a future where the x86 and ARM processors can't compete for that market set and they want to get in on it, in another 10+ years it could be a very lucrative market, and if the Taiwanese government is willing to hand out funds to get competitive in that space then all the power too them.

you've surmised correct. as someone with a mac pro 5,1, which is "old", i'd like a quantum computer for my matrix calculations because the precision would be infinite compared to 64 bit.

there are other perks too, of course, but to me it's all about the probabilistic model of the atom and leveraging its predictions of a particle's location to improve the resolution of my algorithm.

the brain is a quantum computer...
tozcXYF.gif

FdprESh.gif

A3ltCV1.gif

5gC5ByB.gif


these are all in vivo as well
too bad there could only be "one winner"
Mitt is determned to hang on to canada as hispersonal bank account
 
I remember reading a Taiwanese business journal about a decade back where the head honcho of TSMC lamented that they play third (not even second fiddle) to the big boys of Intel and AMD. It's kind of amazing that TSMC is now at the vanguard of the industry.
 
everyone uses the term quantum, that i don;t even know what the hell is quantum suppose to be anymore.

but hell there is sure a lot of probabilistic modeling being called quantum.

wasn't quantum about entanglement? you flip one photon on Earth, and another paired copy will flip in Pluto instantaneously. ?
 
for all intents and purposes: the D-Wave offerings are not "real" quantum computers.

i believe TSMC is implying that a temperature of 0 (Lord) Kelvins is necessary to "slow down" the subatomic particles such that the states can be read/written (presumably using the more-powerful theory of the riemann sphere instead of the bloch sphere)

it will be easy for the initiated (mathematician; won't make this qualification again, it will be assumed) to see that the theory of the bloch sphere is much weaker than that of the riemannian sphere (projective geometry/space vs complex/real-valued geometry that can be *formulated* to fit in projective geometry).

really happy to see asus move away from mobile and my pals at TSMC looking at a very important issue.

i have confidence their team will get some real work done, unlike D-Wave.

Ouch that shows just how little I know about the Quantum Computing architectures and fall into the group of those who just generalize things.

Either way D-Wave's input could be helpful in the long run. Also if you want to see part of the internal workings of the D-Wave "Quantum" computers, Linus Tech Tips actually shows some of how it's built/operated.

All I know is they could potentially do some amazing things going forward... from mapping weather patterns to who knows maybe mapping our garbage in space so we can clean it up and see commercial space flight or even locate and calculate the routes of all Asteroids/Comets in our solar system. *cough* all of this being an external co-processor to make Crysis even better lol j/k
 
Sadly, so few know this nor understand the implications it has.
very true.

but the way it is considered as a quantum computer, is interesting. Read this essay by dr koch and hepp called "quantum mechanics in the brain".

Essentially, brain *function* as taught/learned by a psychiatrist/neurologist is *macroscopic*. Meaning, we can understand the rudimentary functions (which part of the brain is activated while performing a task) from a "classical physics" perspective (without quantum mechanics).

However, since it's electrical, there is quantum mechanical phenomena at the *microscopic* scale (nano or even Pico, not sure). We could look at the macroscopic function as a consequence of multiple superpositions that result from quantum mechanical phenomena at the nanoscopic level, but we're not that far yet.

Sir Roger's quantum or consciousness gets ripped in dr koch's essay, but an idea *like* it or close to it may be necessary to understand how exactly the brain wires/fires together (which is not really required to understand psychiatric disease since those are biomarkers, aka physical targets with activity associated with a certain task).

my mentor from the the motherland, dr karl friston, was just in wired magazine recently, too, and he deserves it. His theory reconciles thermodynamics with brain function, which is also an integral part of the puzzle.
 
everyone uses the term quantum, that i don;t even know what the hell is quantum suppose to be anymore.

but hell there is sure a lot of probabilistic modeling being called quantum.

wasn't quantum about entanglement? you flip one photon on Earth, and another paired copy will flip in Pluto instantaneously. ?


Strictly speaking, a "quantum" is the smallest measurement that can be detected. it's related to the Planck length.

probabilistic modeling, if done right, *can* apply to quantum mechanics, but not always. truth is the mathematical basis for probability theory is equipped to deal with infinites on the [0, 1] interval.

this means a lot of good quantum theory work will share the same machinery, and sometimes similar concepts, to the work in probability theory (but not always).

regarding entanglement: can't help you here. im good at the mathematical, biological, computing and engineering stuff, but not so hot at the theoretical physics (never been a huge fan of armchair theorists).
 
Strictly speaking, a "quantum" is the smallest measurement that can be detected. it's related to the Planck length.

probabilistic modeling, if done right, *can* apply to quantum mechanics, but not always. truth is the mathematical basis for probability theory is equipped to deal with infinites on the [0, 1] interval.

this means a lot of good quantum theory work will share the same machinery, and sometimes similar concepts, to the work in probability theory (but not always).

regarding entanglement: can't help you here. im good at the mathematical, biological, computing and engineering stuff, but not so hot at the theoretical physics (never been a huge fan of armchair theorists).


hmm? I was operating under the impression that quantum entanglement has been re-created in labs all over the place and over some pretty incredible distances at that.
 
hmm? I was operating under the impression that quantum entanglement has been re-created in labs all over the place and over some pretty incredible distances at that.

you may very well be right; all i was saying is that i am not familiar with quantum entanglement at my self-imposed standards to speak on it, that's all.

i know the high-level idea; my familiarity with it starts at Bell's inequality, which i was told is just a reformulation of Boole's conditions of possible experience.
here was my participation in that discussion a few years ago on pubpeer. [1],[2],[3]

i may have posted more; just look for the all lowercase poster and that's probably me ;)
 
you may very well be right; all i was saying is that i am not familiar with quantum entanglement at my self-imposed standards to speak on it, that's all.

i know the high-level idea; my familiarity with it starts at Bell's inequality, which i was told is just a reformulation of Boole's conditions of possible experience.
here was my participation in that discussion a few years ago on pubpeer. [1],[2],[3]

i may have posted more; just look for the all lowercase poster and that's probably me ;)

All that stuff is way over my head, I'm just going off what stupid little science news articles I've seen. My personal knowledge on physics tanks hard once you get past the 19th century with Maxwell and Coulomb.
 
Let me know when we can get Quantum Computers to this point:
Timeline-1.jpg


(From the movie "Timeline") ;)
 
Back
Top