Bloomberg Technology Special: Live From GTC – March 20, 2025

Episode Overview

This special edition of Bloomberg Technology broadcasts live from Nvidia’s GTC (GPU Technology Conference) in San Jose, focused on quantum computing and its intersection with AI and accelerated computing. Host Ed Ludlow moderates conversations with industry leaders, including Nvidia’s Tim Costa, IonQ’s Peter Chapman, and D-Wave’s Alan Baratz. The main theme is examining the state of quantum computing, Nvidia’s role as an “accelerator” rather than a quantum computer builder, and a spirited debate over what “useful” quantum computing looks like—and when it will arrive.

Key Discussion Points and Insights

1. Nvidia’s Role in Quantum Computing

Guest: Tim Costa (Senior Director of Quantum Computing, Nvidia)
Timestamps: [03:02]–[09:06]

• Nvidia’s Vision & Participation

  • Nvidia does not build or sell quantum computers, but provides critical architecture, GPUs, software, and support to quantum companies.
  • Their focus: Integration of quantum processors as data center components alongside CPUs, GPUs, and memory.
  • “Quantum technology offers promise to be very good at certain kinds of computation... we’re focused on helping quantum companies to better develop those technologies.” — Tim Costa [03:02]

• Supporting the Industry

  • Nvidia partners with over 160 quantum-focused groups, accelerating research, simulation, error correction, and device calibration.
  • Quantum devices, tightly integrated with GPUs, show promise in heterogeneous computing—each technology focuses on what it does best.
  • “These are, in some ways, physics experiments upon which we try to compute as much as a computer. Managing that physics experiment ... is actually a complex computational task.” — Tim Costa [03:02]

• The AI-Quantum Symbiosis

  • AI aids quantum by handling device calibration and complex error correction. Conversely, quantum computers could output unique physics data, valuable for AI model training.
  • “There are two sides of AI and quantum... AI for quantum... [and] quantum outputs as training data for AI models.” — Tim Costa [06:15]

• Establishing a Quantum Research Center

  • Nvidia announces a Boston research center to further quantum-accelerated supercomputing.
  • The center will let partners work directly on integrating quantum devices with Nvidia’s infrastructure, focusing on real-world quantum-accelerated computing.
  • “If we’re going to build a new kind of computer... that is a physical endeavor, it has a footprint, a place to do it.” — Tim Costa [07:36]

• Defining “Useful” in Quantum Computing

  • Tim Costa highlights chemistry and biochemistry as the first likely domains for practical quantum breakthroughs.
  • “I do think... chemistry, biochemistry related areas... will be the first area that’s disrupted, or at least I do.” — Tim Costa [08:19]

2. IonQ’s Perspective: Demonstrating Progress & Commercialization

Guest: Peter Chapman (Executive Chairman, IonQ)
Timestamps: [11:38]–[16:24]

• Recalibrating Expectations

  • The session opens by addressing Nvidia CEO Jensen Huang’s earlier, controversial statement: Useful quantum computing is “a decade or more away.” Chapman and others use the day to “bring that timeline in.”
  • “He said on stage, I think twice it was his mea culpa... probably the first CEO in history to invite a panel of people to tell him that he was wrong.” — Peter Chapman [11:51]

• Practical Applications & Partnerships

  • Chapman highlights collaborations with Nvidia, AWS, AstraZeneca, and Ansys:
    • A recent demonstration yielded a 20x improvement on earlier benchmarks.
    • A partnership with Ansys produced a 12% performance gain using IonQ’s quantum computers.
    • “While those numbers are not enough to take over the market...that we did it on our 36-qubit system is remarkable.” — Peter Chapman [12:00]

• Use of Nvidia Technology at IonQ

  • IonQ employs Nvidia DGX GPU clusters for quantum computer design and for simulating small-qubit systems (up to ~35 qubits); higher counts require exponentially more resources.
  • “For small qubit counts...we run the simulation on a GPU... when you get to 64 qubits, you need two and a half billion GPUs...” — Peter Chapman [13:46]

• Pushing Toward Commercialization

  • Chapman agrees with Huang: early “usefulness” must be revenue-generating—chemistry and engineering are the likeliest early customers.
  • “You need to find a set of applications early on that you can start to make money on and...power your R&D.” — Peter Chapman [14:27]

• Handling Market Skepticism

  • Chapman responds to a recent short-seller report as part of being a public company, downplaying its impact.
  • “It’s a short report... these are the kinds of things you have to endure.” — Peter Chapman [15:07]

• Industry Momentum

  • Chapman compliments Nvidia for giving the sector a platform, but cautions that real progress will come from execution, not just messaging.
  • “What really matters is actually going and doing it, not actually getting the message out.” — Peter Chapman [15:49]

3. D-Wave’s Contrarian View: Quantum Usefulness is Here

Guest: Alan Baratz (CEO, D-Wave)
Timestamps: [18:47]–[23:47]

• Disagreeing with the Decadal Timeline

  • Baratz maintains that D-Wave’s unique “annealing” approach already delivers practical results, contrary to Nvidia’s long-term timeline.
  • “We are actually able to support useful, important applications today.” — Alan Baratz [18:47]

• Scientific & Commercial Use Cases

  • D-Wave published results in Science: calculating properties of magnetic materials that classical computers cannot, enabling swift materials discovery.
  • Demonstrated quantum-powered blockchain/proof-of-work, enabling energy-efficient crypto mining.
  • “We can now create a blockchain that uses a quantum computer to do the proof of work...” — Alan Baratz [19:47]
  • “Cryptocurrency mining could be at a fraction of the energy cost...” — Alan Baratz [19:53]

• Readiness & Timelines

  • D-Wave is already fielding prototypes for quantum blockchain and commercial clients (e.g., NTT DoCoMo for cell tower optimization).
  • Baratz estimates blockchain commercialization is a year or two away—not 10–20 years.
  • “I think we’re looking at a year or two.” — Alan Baratz [20:30]

• Distinct Approach from Nvidia’s Narrative

  • D-Wave’s annealing systems don’t require the GPU-powered error correction that Nvidia offers; they self-calibrate their large (5,000+ qubit) machines.
  • For D-Wave’s gate-model (more conventional) research, Nvidia’s offerings may be more relevant.
  • “We calibrate them ourselves. We don’t need GPU power to calibrate those systems. So currently we do not.” — Alan Baratz [21:57]

• Reflections on Industry Debate

  • Baratz is ambivalent about the value of the GTC debate for the industry, emphasizing maturity and practical value over narrative control.
  • “I don’t think this event was all that helpful... We're still at the beginning of a learning curve with respect to how Nvidia and Jensen interact with quantum computing companies.” — Alan Baratz [23:21]

Notable Quotes & Memorable Moments

• Jensen Huang (Nvidia CEO), on January 8th’s Infamous Caution:
  “We were more than a decade away from quantum computers being able to do anything useful.”
  [Referencing earlier comments, context at [01:08], direct quote replayed [02:01]]

• Tim Costa (Nvidia) on Nvidia’s role:
  “We do not build a quantum computer, but we have a vision about where quantum computing will be useful...” [03:02]

• Peter Chapman (IonQ) on Timelines and Product Milestones:
  “The fact that we managed to do it on our 36-qubit system is really quite remarkable.” [12:00]
  “You need to find a set of applications early on that you can start to make money on...” [14:27]

• Alan Baratz (D-Wave) on Quantum Usefulness:
  “We are useful today. Blockchain is one of our newest application areas. And yes, we’re just starting to roll that.” [21:07]
  “I don’t think this event was all that helpful to the industry or to D-Wave.” [23:21]

• Ed Ludlow (Host) on Defining ‘Useful’:
  “I don’t think... we’ve reached a definitive agreement on what useful is.” [07:07]

Important Segment Timestamps

• [03:02]–[09:06]: Tim Costa (Nvidia) on supporting quantum innovation, integration, and defining quantum “usefulness”
• [11:38]–[16:24]: Peter Chapman (IonQ) on timeline revision, practical breakthroughs, commercialization, and dealing with market critiques
• [18:47]–[23:47]: Alan Baratz (D-Wave) on real-world applications, differentiating D-Wave’s approach, and what’s needed for industry credibility

Summary Table: Companies & Perspectives

| Company | Position on Usefulness | Nvidia Relationship | Key Use Cases Highlighted | |-------------|-----------------------------|----------------------------------|------------------------------------| | Nvidia | “Useful” is years off, but integration is imminent | Provides architecture, software, GPU access; supports quantum ecosystem | Simulation, calibration, error correction, chemistry, AI models | | IonQ | Timeline is shortening; demonstrated partial wins, needs revenue use cases | Leverages Nvidia GPUs for design/simulation | Chemistry, engineering, Ansys partnership, R&D-focused | | D-Wave | Useful quantum is here, via annealing architecture | Minimal dependency; self-sufficient calibration | Scientific research (magnetics), blockchain, NTT DoCoMo partnership |

Overall Conclusion

This GTC special underscores both the dramatic progress and persistent debate in quantum computing. Nvidia aims to accelerate quantum’s arrival by supporting the entire ecosystem but maintains a cautious outlook on commercialization timelines. IonQ seeks to bridge the lab-market gap, demonstrating incremental progress and partnerships. D-Wave argues forcefully that its annealing approach is already unlocking real-world value, challenging the dominant industry narrative.

The episode is a snapshot of an industry in lively transition, balancing optimism, rivalry, and the relentless push for real, scalable quantum utility.

For further listening: Find the full episode on the Bloomberg Terminal or wherever you get your podcasts.