
Quantum processing unit
30
MENTIONS
6
EPISODES
5
PODCASTS
Search complete. 30 mentions across 6 episodes found for "Quantum processing unit".
Nov 19, 2026
Taking an atomic approach to silicon quantum computing
M
6:33Michelle SimmonsGUEST
So in such a way, you have full atomic resolution in all X, Y, and Z.
M
6:37Michelle SimmonsGUEST
But the key essence is in the whole QPU, the quantum processing unit that we make, there are only two atoms, phosphorus and silicon.
M
6:44Michelle SimmonsGUEST
And the hydrogen basically segregates out and goes to the surface as you grow.
H
6:49Hamish JohnstonHOST
And is that something that you make in-house? Is that what SQC does, make these devices? Or do you have somebody else do it for you?
8 MINS LATER
M
14:37Michelle SimmonsGUEST
And so we've been very excited recently that in our field, there's been a massive along the side of us advancements in the GPUs and GPUs for companies like Nvidia.
M
14:48Michelle SimmonsGUEST
They've basically been 3D packaging them so that they can fit lots of them together.
M
14:53Michelle SimmonsGUEST
And so we now can use those kind of 3D packaging techniques to connect to our QPU layers.
M
14:58Michelle SimmonsGUEST
And so we've been waiting, honestly, for the silicon industry to develop that technology alongside us to bring their pitch size down so that it now is compatible with the sizes that we get in our QPU.
Kirk Bresniker on Scaling Quantum the HPE Way
K
18:01Kirk BresnikerGUEST
So even there, it took a little time, but understanding how we adapt the physical data center environment, how we make teams productive, safe, productive, secure, all those things.
K
18:13Kirk BresnikerGUEST
I think there's great lessons to be drawn from the introduction of accelerated computing in GPUs and how we see accelerated computation as we now look to the potential of having multi-physics be everything from Newton and Galileo to Schrodinger and Dirac, because now we have CPUs, GPUs, QPUs, all these other accelerators all being put to purpose and part of that enabling again is AI driven transcoding and coding that will lower that barrier to just the amount of code that we need to generate across all these diverse devices to make them efficiently render up their efficiency.
K
18:59Kevin JacksonHOST
You're pointing to the future here, and I think I'll finish on these two questions.
K
19:04Kevin JacksonHOST
Let's just say we're looking forward to relatively soon, 2030, 31, 32, relatively soon.
QTC's Conversation with Masoud Mohseni
S
6:51Steve LaBarberaHOST
But first, let's set the table for our listeners, because HPE's role in this industry is a little bit different from some of the others.
S
6:59Steve LaBarberaHOST
When you arrived at HPE, you made an assessment about where the company should position itself, and the answer was more of a quantum system integrator rather than a QPU vendor.
S
7:10Steve LaBarberaHOST
Can you walk us through that reasoning and why that's the right fit for HPE? What were some of the arguments for staying out of the qubit building business?
M
7:20Masoud MohseniGUEST
I believe that HPE has the legacy, know-how, and the capability to use advanced processors and make them available in enormous computing infrastructures that work at this scale.
S
8:25Steve LaBarberaHOST
My understanding is that HPE is coming at it from the other end.
S
8:30Steve LaBarberaHOST
Can you explain this top-down approach and what actually changes architecturally when you do it that way?
M
8:36Masoud MohseniGUEST
Yeah, so this is also related to your previous question, why we do not even build QPU in-house, why we are not going to be just another QPU vendors.
M
8:48Masoud MohseniGUEST
So the general history of quantum computing has been beyond the early academic days is to build QPUs given certain capabilities of certain modality, let's say a neutral atom, trapped ion system, superconducting qubits or spin-based qubit, could be photonic.
The Intersection of AI and Quantum Computing
S
44:12speaker_2HOST
Okay, standard stuff.
S
44:13speaker_3GUEST
And physically adjacent to them, connected by ultra-fast fiber optics, you have the dilution refrigerators housing the quantum processing units, or QPUs.
S
44:22speaker_2HOST
And the orchestrator of all this is artificial intelligence.
S
44:26speaker_2HOST
You submit a massively complex problem to the cloud.
Chris Cassin: Building Ultra-Cold Infrastructure for Quantum Tech | The Further, Faster Podcast
C
29:12Chris CassinGUEST
And so we're going to see the same thing with quantum computers.
C
29:15Chris CassinGUEST
Quantum computers have something called a QPU, a quantum processing unit.
C
29:20Chris CassinGUEST
So what we'll see is I believe we'll see data centers with CPUs, GPUs, and QPUs working in concert to solve really complex problems.
C
29:30Chris CassinGUEST
And so what does that mean for us as a cryogenics company? We are really building the critical infrastructure that supports and enables this technology.
C
29:40Chris CassinGUEST
And so, you know, we will have hundreds of millions of dollars worth of infrastructure at a single data center and we will scale with them and support the uptime of their systems.
QTC's Conversation with Claire Le Gall
C
15:44Claire Le GallGUEST
And we address a huge gap in the market because this solution will be part of any future fault-tolerant quantum computing system.
S
15:57Steve LaBarberaHOST
So how mature would you say quantum networking is compared to computing itself? QPUs are already being sold and deployed today in the market.
S
16:06Steve LaBarberaHOST
Where would you say the networking side sits relative to that? And what has to happen to bring these networks to market?
C
16:15Claire Le GallGUEST
Right, so now the field of quantum networking for compute is smaller and earlier.
21 MINS LATER
S
37:06Steve LaBarberaHOST
New Quantum has another product or device called the Quantum Networking Unit.
S
37:12Steve LaBarberaHOST
The first generation product is out now.
S
37:15Steve LaBarberaHOST
In practical terms, how does this help someone looking to operate several QPUs as part of a quantum network?
C
37:23Claire Le GallGUEST
Right.