Latest / The Quark Side - Quantum Physics Podcast / This New Quantum Memory Solves a Major Quantum Computing Problem

Transcript

This New Quantum Memory Solves a Major Quantum Computing Problem · The Quark Side - Quantum Physics Podcast

• 0:01
  Speaker 1: Welcome to the QuarkSide Quantum Physics Podcast, an exploration of
• 0:05
  Speaker 1: the fundamental structure of reality, where quantum laws govern matter, energy,
• 0:11
  Speaker 1: and information. Here, uncertainty is a feature, not a flaw,
• 0:16
  Speaker 1: and understanding begins at the smallest scales.
• 0:27
  Speaker 2: I want you to imagine, just for a second, Trying
• 0:31
  Speaker 2: to use the exact same brain cells to actively solve
• 0:34
  Speaker 2: this super complex multivariable calculus equation.
• 0:38
  Speaker 3: Okay. Sounds exhausting already. Right.
• 0:40
  Speaker 2: But simultaneously, you have to use those exact same cells
• 0:43
  Speaker 2: to remember a massive grocery list. Oh, wow. At the
• 0:47
  Speaker 2: exact same time. Like, you aren't switching back and forth.
• 0:49
  Speaker 2: You are forcing the neurons doing the heavy mathematical lifting
• 0:53
  Speaker 2: to also act as, you know, a static filing cabinet.
• 0:56
  Speaker 3: Yeah, you'd either completely forget the milk or you'd just
• 0:59
  Speaker 3: totally botch the calculus. I mean, the human brain doesn't
• 1:03
  Speaker 3: work that way.
• 1:03
  Speaker 2: No, not at all.
• 1:05
  Speaker 3: And honestly, classical computers don't work that way either.
• 1:08
  Speaker 2: Right, which makes it totally wild that this brain-melting, overwhelming
• 1:12
  Speaker 2: scenario is precisely how current superconducting quantum computers actually operate.
• 1:17
  Speaker 3: Yeah, it's true.
• 1:18
  Speaker 2: They are literally trying to hold on to the grocery
• 1:20
  Speaker 2: list and do the calculus with the exact same hardware.
• 1:23
  Speaker 3: It's a huge burden on the system.
• 1:26
  Speaker 2: But that era might officially be over. There's this freshly
• 1:29
  Speaker 2: published discovery from September 2026 in Nature Physics, and it
• 1:33
  Speaker 2: finally separates a quantum computer's active thinking from its memory.
• 1:37
  Speaker 3: Which is huge.
• 1:38
  Speaker 2: Essentially, it gives quantum computers their very own version of RAM.
• 1:41
  Speaker 3: Yeah, and to really appreciate how massive of a paradigm
• 1:44
  Speaker 3: shift this is, we should probably look at why classical computers,
• 1:48
  Speaker 3: you know, the laptop or smartphone you're listening to this on,
• 1:51
  Speaker 3: why they are so efficient.
• 1:52
  Speaker 2: Okay, let's unpack that.
• 1:54
  Speaker 3: So, They rely on what we call the von Neumann architecture.
• 1:58
  Speaker 3: The processing, which is your CPU, and the memory, your RAM.
• 2:01
  Speaker 3: They're physically distinct pieces of hardware. Right.
• 2:04
  Speaker 2: They have entirely different jobs.
• 2:06
  Speaker 3: Exactly. The CPU does the math, but it doesn't store
• 2:09
  Speaker 3: the data long term. It offloads that data to the RAM,
• 2:12
  Speaker 3: which acts as a dedicated holding area.
• 2:15
  Speaker 2: But if that architecture is so obviously efficient, and I mean,
• 2:18
  Speaker 2: we've known about it since the 1940s, why didn't we
• 2:21
  Speaker 2: just build quantum computers that way from day one?
• 2:24
  Speaker 1: Well.
• 2:24
  Speaker 3: It wasn't for lack of trying.
• 2:26
  Speaker 2: Were hardware engineers just like ignoring a totally solved problem?
• 2:30
  Speaker 3: Not at all. They were actually fighting against the fundamental
• 2:33
  Speaker 3: laws of quantum mechanics.
• 2:35
  Speaker 2: Oh, right. It always comes back to the physics. Always.
• 2:39
  Speaker 3: See, in a classical computer, copying data from the CPU
• 2:43
  Speaker 3: to RAM is trivial because classical bits are just ones
• 2:47
  Speaker 3: and zeros.
• 2:48
  Speaker 2: You can just duplicate them.
• 2:49
  Speaker 3: Yeah, you can copy a one as many times as
• 2:51
  Speaker 3: you want. But in the quantum realm, we are dealing
• 2:54
  Speaker 3: with qubits, which exist in these really delicate states of superposition.
• 2:58
  Speaker 2: The classic quantum weirdness.
• 3:00
  Speaker 3: Exactly. And there's a strict rule in quantum mechanics called
• 3:03
  Speaker 3: the no-cloning theorem. It basically states that you cannot perfectly
• 3:08
  Speaker 3: copy an unknown quantum state.
• 3:10
  Speaker 2: Ah.
• 3:11
  Speaker 3: You can't just copy and paste the data. You have
• 3:13
  Speaker 3: to physically move the fragile quantum state itself.
• 3:17
  Speaker 2: So you can't just hit Control-C, Control-V?
• 3:19
  Speaker 3: No, you absolutely cannot.
• 3:21
  Speaker 2: You're moving the actual physical information, which I imagine is
• 3:24
  Speaker 2: incredibly risky when the information degrades if you just look
• 3:28
  Speaker 2: at it funny.
• 3:29
  Speaker 3: That is exactly the problem. Quantum states suffer from decoherence.
• 3:33
  Speaker 3: They collapse incredibly fast. Right. So early quantum engineers decided
• 3:37
  Speaker 3: the safest bet was to just do everything in one place.
• 3:41
  Speaker 2: Keep it all on the processor. Right.
• 3:43
  Speaker 3: Keep the data in the processor. Do the calculations as
• 3:46
  Speaker 3: fast as physically possible before the state collapses.
• 3:52
  Speaker 2: But that created a structural bottleneck, right? And now we're
• 3:54
  Speaker 2: hitting a wall?
• 3:55
  Speaker 3: A literal physical wall, yes.
• 3:58
  Speaker 2: Which brings us to the actual hardware. Because when you
• 4:00
  Speaker 2: say bottleneck, I feel like people imagine a slowdown in
• 4:03
  Speaker 2: software like a loading bar.
• 4:05
  Speaker 3: Sure, yeah.
• 4:05
  Speaker 2: But you're talking about a space problem inside the actual machine. Yes.
• 4:09
  Speaker 3: Because every single qubit in a superconducting quantum computer acts
• 4:13
  Speaker 3: as a processor bit, every single one requires its own
• 4:17
  Speaker 3: dedicated control wiring.
• 4:18
  Speaker 2: Oh, wow. Every single one.
• 4:20
  Speaker 3: Every single one. And we are talking about highly specialized,
• 4:24
  Speaker 3: heavy coaxial cables carrying microwave signals.
• 4:27
  Speaker 2: And all of this has to go inside a dilution refrigerator. Right.
• 4:30
  Speaker 2: These are those giant gold-plated chandeliers you always see in
• 4:33
  Speaker 2: photos of quantum labs. They look like incredibly complicated brass
• 4:36
  Speaker 2: instruments hanging upside down.
• 4:38
  Speaker 3: That's a perfect way to describe them. And those refrigerators
• 4:41
  Speaker 3: have to cool the quantum chips down to about 10 millikelvin.
• 4:44
  Speaker 2: Wait. 10 millikelvin.
• 4:46
  Speaker 3: Yeah, which is just a fraction of a degree above
• 4:49
  Speaker 3: absolute zero.
• 4:50
  Speaker 2: That is colder than deep space.
• 4:52
  Speaker 3: It really is. Now, imagine trying to scale up that computer.
• 4:57
  Speaker 3: If you want to add a single qubit, just to
• 4:59
  Speaker 3: act as memory, you have to run a brand new
• 5:01
  Speaker 3: set of heavy cables all the way from room temperature
• 5:04
  Speaker 3: at the top of the fridge down through all the
• 5:06
  Speaker 3: cooling stages right to the chip.
• 5:09
  Speaker 2: I see the problem. Because every cable you add brings
• 5:12
  Speaker 2: ambient heat down into the fridge.
• 5:14
  Speaker 3: Exactly.
• 5:15
  Speaker 2: At some point, you just physically melt the environment you're
• 5:17
  Speaker 2: trying to freeze.
• 5:18
  Speaker 3: You do. David I. Schuster, he's one of the leading
• 5:21
  Speaker 3: researchers behind this new breakthrough. He actually summed it up perfectly.
• 5:24
  Speaker 2: What did he say?
• 5:25
  Speaker 3: He noted that if every additional qubit costs another set
• 5:28
  Speaker 3: of control hardware, scaling stops being a physics problem and
• 5:32
  Speaker 3: becomes an engineering one.
• 5:33
  Speaker 2: Because you just run out of space.
• 5:35
  Speaker 3: You simply run out of physical space and cooling capacity
• 5:38
  Speaker 3: for all those wires. If you want a quantum computer
• 5:41
  Speaker 3: with a million qubits... You can't have a million cables
• 5:44
  Speaker 3: running into the fridge. It's impossible.
• 5:46
  Speaker 2: Okay, so that's the wall we've hit. We desperately need memory,
• 5:49
  Speaker 2: but we can't afford the wires to support it.
• 5:51
  Speaker 3: Exactly.
• 5:52
  Speaker 2: So how do these researchers from Stanford, the University of Chicago,
• 5:55
  Speaker 2: and the Slack National Accelerator Laboratory actually solve this?
• 5:59
  Speaker 3: It's pretty brilliant.
• 6:00
  Speaker 2: If they aren't running a million new wires, what are
• 6:02
  Speaker 2: they doing?
• 6:02
  Speaker 3: They designed a completely new shared infrastructure using something called
• 6:06
  Speaker 3: a multimode aluminum cavity.
• 6:08
  Speaker 2: Okay, I'm going to need you to break down what
• 6:10
  Speaker 2: a multimode cavity is, because cavity just makes me think
• 6:13
  Speaker 2: of a hole in a tooth.
• 6:14
  Speaker 3: Fair enough, yeah. Physically, it's a highly polished, hollowed-out block
• 6:18
  Speaker 3: of superconducting aluminum.
• 6:20
  Speaker 1: Okay.
• 6:20
  Speaker 3: When it's cooled down to those extreme temperatures, the aluminum
• 6:23
  Speaker 3: loses all electrical resistance. Got it. Now, inside this hollow space,
• 6:29
  Speaker 3: you can inject microwave energy.
• 6:31
  Speaker 2: So we have a super-cold metal box filled with microwaves.
• 6:34
  Speaker 2: Where does the memory come in?
• 6:36
  Speaker 3: It comes down to standing waves. Think of the cavity
• 6:39
  Speaker 3: like the body of a flute.
• 6:41
  Speaker 2: Like the instrument. Yeah.
• 6:43
  Speaker 3: When you blow air into a flute, it can produce
• 6:45
  Speaker 3: different harmonic frequencies, right? Different distinct pitches.
• 6:48
  Speaker 2: Sure.
• 6:49
  Speaker 3: In our aluminum cavity, the microwave energy forms distinct patterns
• 6:53
  Speaker 3: of oscillation. We call those modes.
• 6:56
  Speaker 2: Modes. Okay.
• 6:57
  Speaker 3: And because of the specific geometry of the cavity, it
• 6:59
  Speaker 3: can support multiple different modes or frequencies at the exact
• 7:03
  Speaker 3: same time without them interfering with one another.
• 7:06
  Speaker 2: So each specific frequency acts as a separate filing cabinet?
• 7:09
  Speaker 3: Precisely. The big breakthrough here is that each distinct mode
• 7:13
  Speaker 3: inside this single physical cavity can act as a totally
• 7:17
  Speaker 3: separate addressable memory cell.
• 7:19
  Speaker 2: You know, Schuster actually used a musical analogy in the
• 7:21
  Speaker 2: research that really helped this click for me.
• 7:23
  Speaker 3: Oh, the chords.
• 7:25
  Speaker 1: Yeah.
• 7:25
  Speaker 2: Imagine a single metal box, and inside that box you
• 7:28
  Speaker 2: have multiple separate musical notes ringing at the exact same time.
• 7:32
  Speaker 3: Right.
• 7:32
  Speaker 2: Because an A-sharp and a C-natural are completely different frequencies...
• 7:36
  Speaker 2: They don't destroy each other. They just coexist in the
• 7:38
  Speaker 2: same space.
• 7:39
  Speaker 3: Exactly.
• 7:40
  Speaker 2: So each unique note ringing in that box holds one
• 7:43
  Speaker 2: qubit's worth of quantum information.
• 7:45
  Speaker 3: It's such an elegant way to visualize it. You have
• 7:47
  Speaker 3: this one physical footprint, the single block of aluminum, but
• 7:52
  Speaker 3: it provides multiple storage slots.
• 7:54
  Speaker 2: That's amazing.
• 7:55
  Speaker 3: In this specific prototype, they managed to hold seven distinct
• 7:58
  Speaker 3: memory cells in one cavity. It means you're replacing seven
• 8:03
  Speaker 3: individually wired physical qubits with just one shared space.
• 8:08
  Speaker 2: But wait, here's where it gets really tricky. If I
• 8:10
  Speaker 2: have seven delicate notes ringing in this box, how do
• 8:13
  Speaker 2: I actually use them without ruining everything?
• 8:16
  Speaker 3: What do you mean?
• 8:17
  Speaker 2: Well, the processor doing the math, the CPU side. That's
• 8:20
  Speaker 2: made of a different type of qubit, right?
• 8:22
  Speaker 3: Yes, the processor relies on a superconducting qubit called a transmon. Right.
• 8:27
  Speaker 2: And from what I understand, transmons are incredibly noisy, aggressive components.
• 8:32
  Speaker 3: They are very active, yes.
• 8:33
  Speaker 2: They are literally designed to compute, which means they are
• 8:36
  Speaker 2: designed to force quantum states to interact.
• 8:38
  Speaker 1: Right.
• 8:39
  Speaker 2: So if I connect a transmon directly to this beautiful,
• 8:42
  Speaker 2: delicate symphony of memory notes, doesn't the transmon just smash
• 8:46
  Speaker 2: into the cavity and scramble all the data?
• 8:49
  Speaker 3: You've just identified the single greatest obstacle that has historically
• 8:52
  Speaker 3: prevented quantum RAM from working.
• 8:54
  Speaker 2: Really? Oh, yeah.
• 8:56
  Speaker 3: Transmonds are made using Josephson junctions, which gives them strong anharmonicity.
• 9:00
  Speaker 2: Okay, in plain English.
• 9:02
  Speaker 3: In plain English, that means they are highly interactive by nature.
• 9:05
  Speaker 3: If you wire a transmond directly to a multimode memory cavity,
• 9:09
  Speaker 3: the interactions bleed over instantly.
• 9:11
  Speaker 2: So it's an immediate crash.
• 9:12
  Speaker 3: Totally. The zero-point energy fluctuations alone would obliterate the delicate
• 9:16
  Speaker 3: quantum states you were trying to store.
• 9:18
  Speaker 2: Which is great for doing math, but terrible for the
• 9:20
  Speaker 2: grocery list.
• 9:21
  Speaker 3: Exactly. So the researchers' solution was simply to, well, not
• 9:26
  Speaker 3: connect them directly.
• 9:27
  Speaker 2: Oh, just keep them apart.
• 9:29
  Speaker 3: Yeah. They built a cascaded device made of four distinct components.
• 9:33
  Speaker 2: Okay, what are they?
• 9:34
  Speaker 3: You have the transmon acting as the CPU. You have
• 9:37
  Speaker 3: the multimode cavity acting as the RAM. And sitting between them,
• 9:40
  Speaker 3: you have a tunable coupler and a buffer cavity.
• 9:43
  Speaker 2: I was trying to conceptualize this whole buffer thing, and
• 9:46
  Speaker 2: it feels a lot like a carpenter's workbench.
• 9:48
  Speaker 3: Okay, let's hear it.
• 9:50
  Speaker 2: Think of the multimode cavity as... as a fragile glass
• 9:53
  Speaker 2: cabinet where you store perfectly carved, really delicate pieces of wood.
• 9:58
  Speaker 3: I like this.
• 9:59
  Speaker 2: Now, you have a circular saw. That's your transmon processor.
• 10:03
  Speaker 2: You would never turn on the saw while it's sitting
• 10:06
  Speaker 2: inside the glass cabinet.
• 10:08
  Speaker 3: The vibrations alone would shatter everything.
• 10:10
  Speaker 2: Right. The environmental noise would cause immediate decoherence. So instead,
• 10:14
  Speaker 2: you use a buffer.
• 10:16
  Speaker 3: Exactly.
• 10:16
  Speaker 2: You carefully pull the specific piece of wood out of
• 10:19
  Speaker 2: the cabinet. You place it securely on your workbench, which
• 10:22
  Speaker 2: is the buffer, and then you bring the saw down.
• 10:25
  Speaker 3: You do the work outside the cabinet.
• 10:27
  Speaker 2: You do your noisy, messy work on the workbench, and
• 10:30
  Speaker 2: when the cuts are made, you put the finished piece
• 10:32
  Speaker 2: safely back into the cabinet. The saw never even gets
• 10:35
  Speaker 2: near the storage.
• 10:36
  Speaker 3: That is a fantastic analogy. I really like the workbench idea,
• 10:41
  Speaker 3: but we need to look at how that pulling out
• 10:43
  Speaker 3: the wood actually happens physically.
• 10:46
  Speaker 2: Right, because there are no tiny robotic arms in there.
• 10:49
  Speaker 3: Exactly. In a quantum circuit, you have to move the
• 10:52
  Speaker 3: information using microwave pulses.
• 10:54
  Speaker 2: So how does the tunable coupler actually work as the middleman?
• 10:57
  Speaker 2: Because my first instinct is that adding a coupler and
• 10:59
  Speaker 2: a buffer just adds more hardware to a system we
• 11:02
  Speaker 2: are desperately trying to shrink.
• 11:04
  Speaker 3: It does sound counterintuitive, but the tunable coupler is actually
• 11:08
  Speaker 3: the secret to scaling down. Think of it as a
• 11:11
  Speaker 3: frequency bridge.
• 11:12
  Speaker 1: A bridge.
• 11:13
  Speaker 2: Yeah.
• 11:14
  Speaker 3: Normally, the memory modes in the cavity and the buffer
• 11:16
  Speaker 3: cavity change. are operating at completely different frequencies. They are
• 11:20
  Speaker 3: entirely mismatched. Okay. Because they are mismatched, they can't even
• 11:24
  Speaker 3: see or interact with each other. The cabinet is effectively locked.
• 11:29
  Speaker 2: So the data is safe because the physics literally prevent
• 11:31
  Speaker 2: the two sides from communicating.
• 11:33
  Speaker 3: Exactly. But when the processor needs a specific piece of
• 11:36
  Speaker 3: information from one of the seven memory cells, the engineers
• 11:39
  Speaker 3: apply a very specific microwave control tone to the coupler.
• 11:43
  Speaker 3: Like a key. Basically. They pump it with energy that
• 11:47
  Speaker 3: perfectly matches the frequency difference between the chosen memory cell
• 11:50
  Speaker 3: and the buffer.
• 11:52
  Speaker 1: Oh.
• 11:53
  Speaker 2: It's like turning an old-school radio dial until you catch
• 11:55
  Speaker 2: the exact frequency of the station you want.
• 11:58
  Speaker 3: That is precisely the mechanism. It's called a stimulated parametric process.
• 12:03
  Speaker 2: A stimulated parametric process.
• 12:06
  Speaker 3: Right. When you hit that exact difference frequency, the coupler
• 12:10
  Speaker 3: creates a bridge, and a single photon carrying the quantum
• 12:13
  Speaker 3: state smoothly swaps from the memory mode... into the buffer cavity.
• 12:17
  Speaker 2: And how fast does that swap happen? I assume we
• 12:19
  Speaker 2: are racing against decoherence.
• 12:21
  Speaker 3: Oh, incredibly fast. About half a microsecond. Wow.
• 12:26
  Speaker 2: So they dial up the specific musical note they want,
• 12:29
  Speaker 2: pull just that note out onto the workbench in half
• 12:32
  Speaker 2: a microsecond, the transmon does its calculations, and then they
• 12:35
  Speaker 2: reverse the process to put the updated note back in
• 12:38
  Speaker 2: the box.
• 12:39
  Speaker 3: That is the complete cycle, yes. And crucially, while the
• 12:42
  Speaker 3: transmon is doing all that heavy lifting on the workbench,
• 12:45
  Speaker 3: The tunable coupler is turned off for the other frequencies.
• 12:49
  Speaker 2: So the rest of the data is safe?
• 12:50
  Speaker 3: Yep. The other six memory cells sitting in the cavity
• 12:53
  Speaker 3: remain completely undisturbed by the processor's noise.
• 12:57
  Speaker 2: The isolation really is brilliant.
• 12:59
  Speaker 3: It's beautiful engineering.
• 13:01
  Speaker 2: But let's get down to the actual reality of this,
• 13:03
  Speaker 2: because theoretical physics is really fun until you build it
• 13:07
  Speaker 2: and the machine just throws a bunch of errors.
• 13:09
  Speaker 3: Oh, of course.
• 13:09
  Speaker 2: How did this prototype actually perform when they ran this
• 13:12
  Speaker 2: process in the lab?
• 13:13
  Speaker 3: Well, the numbers from researchers Zikan Li and Ish Gupta
• 13:17
  Speaker 3: are incredibly revealing. They ran thousands of tests.
• 13:21
  Speaker 1: And?
• 13:22
  Speaker 3: When they measured just the swap mechanism, so moving the
• 13:25
  Speaker 3: data from the memory to the buffer and back without
• 13:27
  Speaker 3: doing any computation on it, they achieved a 99.3% fidelity rate.
• 13:33
  Speaker 2: Okay, that sounds fantastic for a prototype, but what about
• 13:35
  Speaker 2: a full cycle, doing the math?
• 13:38
  Speaker 3: When they tested the complete read-compute-write cycle, pulling the data,
• 13:42
  Speaker 3: having the transmon actively operate on it, and writing it back,
• 13:46
  Speaker 3: the cycle introduced an average error of about 1.2% per
• 13:50
  Speaker 3: memory cell used.
• 13:51
  Speaker 2: Okay, stop right there. What? In classical computing, if my
• 13:55
  Speaker 2: RAM had a 1.2% error rate every single time it
• 13:57
  Speaker 2: read or wrote data, my laptop would be a literal paperweight.
• 14:00
  Speaker 2: It would crash instantly.
• 14:01
  Speaker 3: Oh, absolutely.
• 14:02
  Speaker 2: Why are they celebrating a 1.2% error rate? Is that
• 14:06
  Speaker 2: a passing grade?
• 14:07
  Speaker 3: Because in quantum computing, you aren't comparing it to a
• 14:10
  Speaker 3: classical laptop. You are comparing it to the threshold of
• 14:13
  Speaker 3: quantum error correction, or QEC.
• 14:16
  Speaker 2: Okay, explain that.
• 14:18
  Speaker 3: In the quantum realm, information is so fragile that errors
• 14:20
  Speaker 3: are just inevitable. The strategy isn't to build a flawless
• 14:24
  Speaker 3: physical machine because you can't. Right. The strategy is to
• 14:28
  Speaker 3: build a machine where the errors are low enough that
• 14:30
  Speaker 3: software algorithms can catch them and correct them on the fly.
• 14:35
  Speaker 2: Ah, so the hardware just needs to be, like, good
• 14:38
  Speaker 2: enough for the software to clean up the mess.
• 14:40
  Speaker 3: Exactly. And the most common framework for this, it's called
• 14:43
  Speaker 3: the surface code, requires physical error rates to be around
• 14:46
  Speaker 3: 1% or lower to function properly.
• 14:48
  Speaker 2: Oh.
• 14:49
  Speaker 3: So a 1.2% error rate on a brand new, completely
• 14:52
  Speaker 3: unoptimized architecture proves that this memory system is hovering right
• 14:57
  Speaker 3: on the edge of viability.
• 14:58
  Speaker 2: That's incredibly close. Yeah.
• 15:00
  Speaker 3: With just minor engineering tweaks, it crosses the threshold.
• 15:04
  Speaker 2: That makes a lot of sense. But it really begs
• 15:06
  Speaker 2: the question, where is that 1.2% error actually coming from?
• 15:11
  Speaker 2: Did the radio dial slip? Did the photon get lost
• 15:14
  Speaker 2: on the bridge?
• 15:15
  Speaker 3: That is actually the most fascinating part of the study.
• 15:18
  Speaker 3: They didn't just track the active cell moving to the workbench.
• 15:21
  Speaker 3: They tracked all the idle memory cells that were left
• 15:24
  Speaker 3: behind in the cavity.
• 15:26
  Speaker 2: Oh, checking to see if the glass cabinet shook while
• 15:28
  Speaker 2: the saw was running.
• 15:29
  Speaker 3: Precisely. And they found that the errors weren't coming from
• 15:32
  Speaker 3: natural decay. And they weren't coming from bad control pulses.
• 15:36
  Speaker 3: Then what was it? The errors were coming from something
• 15:39
  Speaker 3: called the cross-cur effect.
• 15:42
  Speaker 2: The cross-cur effect. Okay, what is the mechanism there?
• 15:46
  Speaker 3: Remember that all those memory modes, the different frequencies, they
• 15:49
  Speaker 3: share the exact same physical aluminum walls. Right. So even
• 15:53
  Speaker 3: though they operate at different frequencies, their electric fields still
• 15:57
  Speaker 3: subtly interact with the physical cavity.
• 15:59
  Speaker 2: Okay, they aren't perfectly isolated.
• 16:02
  Speaker 3: Exactly. When a photon occupies one mode, it slightly perturbs
• 16:06
  Speaker 3: the whole electromagnetic environment.
• 16:09
  Speaker 2: Oh, I see.
• 16:10
  Speaker 3: And that causes the frequencies of all the other idle
• 16:13
  Speaker 3: modes to shift just a tiny, tiny bit, like maybe
• 16:16
  Speaker 3: a few kilohertz.
• 16:18
  Speaker 2: Oh, it's like jumping on a trampoline.
• 16:20
  Speaker 3: How so?
• 16:21
  Speaker 2: Well, if I'm bouncing on one side, even if you
• 16:23
  Speaker 2: are standing totally still on the other side, you're going
• 16:25
  Speaker 2: to feel the canvas shift under your feet.
• 16:28
  Speaker 3: That is a great physical intuition.
• 16:30
  Speaker 2: The modes are literally shifting under each other. And that
• 16:33
  Speaker 2: causes the error.
• 16:35
  Speaker 3: Yeah, because quantum operations rely on perfectly timed microwave pulses.
• 16:39
  Speaker 3: If a memory mode drifts by a few kilohertz and
• 16:42
  Speaker 3: your control software doesn't know it drifted, the next time
• 16:45
  Speaker 3: you try to read that cell, your pulse will be
• 16:47
  Speaker 3: slightly out of tune.
• 16:48
  Speaker 2: You won't get a perfect swap.
• 16:50
  Speaker 3: Exactly. That phase drift is where the 1.2% error comes from.
• 16:54
  Speaker 2: But wait, the beauty of that is that it's a measurable,
• 16:57
  Speaker 2: predictable problem. If you know exactly how the trampoline canvas stretches,
• 17:02
  Speaker 2: you can just adjust for it.
• 17:05
  Speaker 3: You've hit the nail on the head. That is the
• 17:07
  Speaker 3: ultimate prize for a hardware engineer. Knowing exactly how a
• 17:11
  Speaker 3: system fails gives you the exact blueprint to fix it.
• 17:15
  Speaker 3: They now know that for the next generation, they either
• 17:18
  Speaker 3: need to design better couplers to aggressively isolate the memories,
• 17:21
  Speaker 3: or they need software that dynamically tracks the phase drift
• 17:25
  Speaker 3: and adjust the control pulses in real time.
• 17:27
  Speaker 2: Which perfectly transitions into the real-world impact of all this,
• 17:31
  Speaker 2: because we aren't just talking about a neat little lab experiment.
• 17:34
  Speaker 2: We are talking about fundamentally redesigning the architecture of quantum processors.
• 17:39
  Speaker 3: It is a massive structural pivot.
• 17:43
  Speaker 2: Think about a modern classical processor chip. Okay. Over 99%
• 17:47
  Speaker 2: of the physical transistors on that chip are dedicated strictly
• 17:51
  Speaker 2: to memory, just caching data so the logic gates can
• 17:55
  Speaker 2: do their jobs smoothly.
• 17:56
  Speaker 3: 99%. Yeah. But before this breakthrough, superconducting quantum processors had
• 18:03
  Speaker 3: essentially zero dedicated memory. Every single qubit was trying to
• 18:06
  Speaker 3: do both jobs.
• 18:08
  Speaker 2: They were trying to do the calculus and remember the
• 18:10
  Speaker 2: grocery list at the exact same time.
• 18:12
  Speaker 3: Exactly. And now suddenly there is a working blueprint where
• 18:17
  Speaker 3: one single active quantum processor can reliably manage seven dedicated
• 18:22
  Speaker 3: storage cells, all without scrambling the data.
• 18:25
  Speaker 2: And that ratio will only improve, right?
• 18:27
  Speaker 3: Oh, for sure.
• 18:28
  Speaker 2: Because going from seven cells to 70 or 700 in
• 18:31
  Speaker 2: a single multimode cavity drastically reduces that tangled nightmare of
• 18:36
  Speaker 2: control wires.
• 18:37
  Speaker 3: Oh, absolutely. You no longer need an individual coax cable
• 18:41
  Speaker 3: running all the way down into the fridge for every
• 18:45
  Speaker 3: single memory qubit.
• 18:46
  Speaker 2: But the hardware isn't the end of the road, right?
• 18:48
  Speaker 2: Because earlier you mentioned quantum error correction. I imagine the
• 18:52
  Speaker 2: software folks are going to have an absolute field day
• 18:54
  Speaker 2: with this new architecture.
• 18:55
  Speaker 3: Oh, they already are. The scientific community is aggressively pivoting
• 18:59
  Speaker 3: to hardware-software co-design.
• 19:03
  Speaker 2: Because you need both. Right.
• 19:04
  Speaker 3: The whole point of creating dense quantum RAM is to
• 19:07
  Speaker 3: enable robust error correction. An algorithm can now use the
• 19:11
  Speaker 3: processor to constantly check physical qubits for errors.
• 19:15
  Speaker 2: And use the memory to track them. Yes.
• 19:18
  Speaker 3: Use this new multimode cavity to store those error logs,
• 19:21
  Speaker 3: we call them syndrome measurements, while it computes the actual corrections.
• 19:26
  Speaker 3: You finally have the hardware and the software working in
• 19:28
  Speaker 3: lockstep to scale up.
• 19:30
  Speaker 2: It really feels like we are watching the quantum equivalent
• 19:33
  Speaker 2: of the moment classical computers stopped using vacuum tubes and
• 19:36
  Speaker 2: started using integrated circuits. Like we are watching the architecture
• 19:40
  Speaker 2: mature in real time.
• 19:42
  Speaker 3: It is the exact same evolutionary leap. happening right now
• 19:46
  Speaker 3: under the laws of quantum mechanics.
• 19:49
  Speaker 2: Which leaves us with a truly mind-bending reality to think about,
• 19:52
  Speaker 2: because this buffer system proves definitively that we can separate
• 19:56
  Speaker 2: quantum memory from quantum processing.
• 19:59
  Speaker 3: The physics works.
• 20:00
  Speaker 2: The isolation works. It is no longer a theoretical debate.
• 20:04
  Speaker 2: It is officially an engineering race. And.
• 20:07
  Speaker 3: Engineering races always, always drive miniaturization.
• 20:10
  Speaker 2: Exactly. So what happens when this quantum memory inevitably shrinks?
• 20:16
  Speaker 2: What happens when it scales down to the physical footprint
• 20:18
  Speaker 2: of classical memory?
• 20:21
  Speaker 3: It's hard to even picture.
• 20:22
  Speaker 2: When quantum computers finally break free from the constraints of
• 20:25
  Speaker 2: these massive wire-stuffed deep-freezed refrigerators and start operating with the
• 20:30
  Speaker 2: physical density of a standard silicon chip.
• 20:33
  Speaker 3: The computational power that will be unlocked is... almost difficult
• 20:38
  Speaker 3: to comprehend. It is.
• 20:39
  Speaker 2: So the question for you listening right now is this.
• 20:42
  Speaker 2: When that tipping point happens, when quantum computers don't have
• 20:45
  Speaker 2: to use the exact same hardware to solve the math
• 20:48
  Speaker 2: equation and remember the grocery list, what historically impossible problem
• 20:53
  Speaker 2: in your industry, in global health, or in digital security
• 20:56
  Speaker 2: will suddenly become solvable overnight? Because the moment this architecture
• 21:01
  Speaker 2: scales up, the speed at which it will change our
• 21:03
  Speaker 2: world is going to be truly unimaginable.