Quantum Programming 101: Solving a Problem From End to End | D-Wave Webinar
Вставка
- Опубліковано 27 лис 2024
- Want to learn how to program a quantum computer? In this webinar, we explain how to do so by running through a complete, simple example. We explain how to formulate the problem, how to write it, and how to tune it for best results.
This webinar is intended for those with little or no experience programming on a D-Wave quantum computer. After watching, get free time on Leap, the quantum cloud service at cloud.dwavesys...
The explanation that you show from 24:00 is the most valuable thing of all your videos. We need more detailed implementations like this. This is the missing link. There are many articles videos about theory and implementations without actually showing or braking down the formulas. Make more videos like this!
Great video, really good explanations on how it works from theory to implementation!
I'm still dying to visit USC's V-lab to checkout their DWaveSys. I'd love to pick the grad/postgrad students' brains. Thanks Joel for the amazing insight. You guys ROCK !
At 10.15 it shows: By split up, we mean that they have an edge connecting them.
After that, I didn't really know what it was about but eventually I realised that they would be split up if they have different coloured shirts and an edge connecting them. That seems obvious now but it went on to the flow of numbers and I only worked it out after the end. I would need to watch it again.
Nice presentation, clear and understandable. Now to fold problems into models ...
It might help to see how it works through if a different set of friendship lines were to be shown, say called Scheme 2. Then, frames labelled Scheme 2 could be shown next to frames labelled Scheme 1, That might be in a section at the end of a video, where a viewer could pause and click from one such frame to the other. Numbers would change in the view and it would show more about what they mean.
How about a variant with 3 teams also?
Great presentation!
I think a choice of words will throw some people, not others. In my first comment here, I called the lines, Friendship Lines, maybe searching for a way to understand the scheme. It's not the lines, or edges, that split them up. It's the different coloured shirts!
It may be of interest that Paul Erdös is said not to have been able to understand the Monty Hall problem - from the 3-box choice game.
Wikipedia:Paul Erdös:
"Paul Erdős was a renowned Hungarian mathematician. He was one of the most prolific mathematicians and producers of mathematical conjectures of the 20th century."
Could it be that the mathematical faculty can be quite separate from the treatment of the real objects in the world? Maybe in D-Wave's line of business it is especially valuable to be able to bridge any such divide.
A little wild theory of mine.
The Universe is Quantum and like everything else everything inside the Universe and the Universe itself tends to mimic.
So here's a wild correlation.
Like a Quantum computer, a superposition state is a unknown 1 or 0 classical outcome.
The fifth force:
So think of 'dark matter' as a 'superposition' state of 1 and 0, but more accurately like this -1 and 1 simultaneously unknown outcomes, until and external Magnetic field is applied ( the magnetic universe).
This is called the Double Well Potential and the outcome is a classical state of 1 or 0 depending what was chosen, matter arises. In turn the the Universe is assigned always '1' because it's an open system.
Before the Big bang a '0', until it NReset again.
So all dark matter is, is unknown outcomes, until and outcome happens. All Possibilities.
So they'll never find 'Dark Matter' because it's not a result,.only an unknown..
GOOD ARGUMENT. HAVE YOU SEEN CHARLIE ROSE SPEAK ABOUT OPENING OTHER DIMENSIONS AND TAKING THEIR RESOURCES WITH QUANTUM COMPUTERS? ALSO I HOLD STOCK IN D WAVE SYSTEMS
@@arniecunningham4220
I have heard Geordie Rose, one of the founders of D-Wave speak about Dimensional Entities, and Vancouver Canada will be the center for AI Civilization...
I will give you the link.. trust me it's a great watch..
@@arniecunningham4220
ua-cam.com/video/cD8zGnT2n_A/v-deo.html
Check it out...
Alien in the Title just means Change..
@@arniecunningham4220
You have a link?
Plus I cant find Stock in D-Wave or Kindred AI..
Where did you find the stocks for D-wave or Kindred AI..?
@@raspberrypi4970 Sorry to reply back late bro. My girl just told me that we're invested in Google stock not d-wave. By the wary you analyse and write you sound like you have a good head on your shoulders. Stay safe my friend
Can dwave only solve Ising hamiltonian(QUBO also) or any arbitrary hamiltonian is suitable for dwave QPU?
Lagrangian in QUBO
In Lagrangian relaxation, an optimization variable sisi is split into two variables, s(1)isi(1) and s(2)isi(2), and the constraint s(1)i=s(2)isi(1)=si(2) added. A similar technique can be used to map an optimization variable sisi onto a set of one or more qubits {q(1)i,⋯,q(k)i}{qi(1),⋯,qi(k)}. Because all qubits (assumed to be ±1±1 valued) represent the same problem variable, impose the constraint q(j)i=q(j′)iqi(j)=qi(j′)for all pairs (j,j′)(j,j′) occurring as edges in a spanning tree across the qubits. The equality constraint q(j)i=q(j′)iqi(j)=qi(j′) can be encoded as the Ising penalty:
−Mq(j)iq(j′)i−Mqi(j)qi(j′)
where M>0 is the weight of the penalty. If M is sufficiently large, the lowest energy state in the full problem always has q(j)i=q(j′)iqi(j)=qi(j′) because that feasible assignment is 2M lower in energy than the infeasible assignment q(j)i≠q(j′)iqi(j)≠qi(j′).
In this way, strings of qubits are related to each other to create chains that can connect arbitrary vertices. To create these chain-like connectors, weigh the penalties large enough so that low-energy configurations do not violate the equality constraints. Balancing this, use the smallest possible penalty weight that enforces the constraints to prevent precision issues, and to foster efficient exploration of the search space. An iterative procedure, which incrementally updates weights until the equality constraints are satisfied, is effective.
My understanding sir, is that the BQM, (binary quadratic model) comes in 2 flavors. 1)QUBO, or 2), Ising. For QUBO, the variables are 0 or 1. For Ising they are, (interestingly enough) 1 or -1.
Either way, the goal is to find values that will minimize the BQM. A conversion from QUBO Ising is somewhat trivial. In the Red-team vs Blue-team webinar, a few weeks ago, they use the QUBO algo'.
In your question posed above, when you say, "...or any arbitrary Hamiltonian...", I think what is important is the conformation of form. The form needs to be correct for the problem to be solved correctly.
I hope this helps.
Cool - thanks guys.
Are ready for this ????????????
this guy sounds like donald trump. good explanation
Spiritual Manipulation, Occult Science 101 are you Sirius! :-)))
Also can you explain Demonic Programming..😁
Jesus is God
is that the reason that my pug dogs food bowl is always blank? I ask for him but think i understand. Amazing video, i thought i hate mathematics, till now, no, i love it. I am now 56 years old btw. Best regards CdH PS; Great video!