Hardware Podcasts

IBM Used Mathematics as Compass on Journey to NorthPole

EE Times Current · 2 May 2025 · 49m

What this episode covers

Dr. Dharmendra Modha, IBM’s chief scientist for brain-inspired computing, explains how his team moved from the TrueNorth chip to the NorthPole inference system by prioritizing mathematical axioms over biological replication. He details how NorthPole achieves 75 times greater energy efficiency than GPUs by fitting large language models entirely on-chip, leveraging a 16-axiom architecture that optimizes time, space, and energy within silicon constraints. The discussion covers the transition from analog to digital neurons, the specific hardware organization of NorthPole, and the strategic decision to focus solely on inference to minimize capital and operating costs. Post-interview commentary from Prof. Ralph Etienne-Cummings and Dr. Julia D'Angelo highlights the importance of academic leadership and the trade-offs between biological realism and computational efficiency in neuromorphic design.

Most relevant to

large enterprises

Topics

Mentioned

Companies
IBM, DARPA, UC San Diego, Nvidia, Cerebras
Products
TrueNorth, NorthPole, Golden Gate, Blue Raven, BrainScaleS

Related episodes from other shows

Other shows on the same ground, closest first.

  • Moore's Lobby · 53m

    Ep. 34 | The Latest from the Lab: How IBM Research Is Inventing What's Next

    Dr. Jeffrey Welser, IBM's Vice President of Exploratory Science and University Partnerships, discusses the intersection of fundamental research and industrial application. He details IBM's progress on the 2 nm chip, the shift from classical scaling to multi-core and specialized architectures to manage power, and the potential of analog computing for AI workloads. The conversation also covers the mechanics of superconducting quantum computers, the importance of open-source collaboration in quantum software, and the role of government funding in sustaining semiconductor innovation.

    Chips, FPGAs & semiconductorsAI & machine learning
    8 chapters

  • Moore's Lobby · 58m

    Ep. 27 | Two Google Senior VPs of Engineering: From Shipping Containers to Today's Data Centers

    Google Fellows Luiz Barroso and Amin Vahdat discuss the evolution of Google's data center infrastructure, from early 'broom closet' colocation facilities and shipping container prototypes to modern, energy-efficient warehouse-scale systems. They explain how Google shifted from relying on commodity hardware to designing custom domain-specific accelerators like TPUs for machine learning and video processing to overcome the limits of Moore's Law. The conversation also covers the critical importance of energy proportionality, Power Usage Effectiveness (PUE), and the cultural philosophy of 'healthy hubris' that drives Google's engineering decisions.

    Energy & climate techChips, FPGAs & semiconductors
    8 chapters

  • IoT For All Podcast · 19m

    The Evolution of AI in IoT | BrainChip's Nandan Nayampally | Internet of Things Podcast

    Nandan Nayampally, CMO at BrainChip, discusses the evolution of AIoT and the shift of AI processing from the cloud to the edge. He explains how neuromorphic computing enables real-time, private, and efficient inference on billions of devices, citing PwC's $15 trillion GDP impact estimate. Nayampally highlights key adoption markets like healthcare, agriculture, and infrastructure, and advises companies to focus on solving acute pain points ('headaches') rather than nice-to-have features ('vitamins') to gain traction.

    AI & machine learningConnected products & IoT
    5 chapters

  • The Circuit · 51m

    Episode 70!!! NPUs!!!!!!!!

    Ben Bajarin and Jay Goldberg host Austin Lyons and Paul Karazuba to discuss the architecture, market, and future of Neural Processing Units (NPUs). The group debunks the metric 'TOPS per watt' as meaningless, arguing that real-world performance depends on specific network workloads and latency. They explore how chipmakers like Qualcomm, Intel, and Apple are designing NPUs, often by licensing IP or evolving DSPs, and predict that NPUs will become ubiquitous in devices from smartphones to microcontrollers within three to four years.

    Microcontrollers & dev platformsChips, FPGAs & semiconductors
    8 chapters

  • Moore's Lobby · 51m

    Ep. 42 | Groq CEO and Ex-Googler Jonathan Ross on the Petaflop AI Chip and First Ever TPU

    Jonathan Ross, founder and CEO of Groq, discusses his journey from designing Google's first TPU to building Groq's deterministic AI chip, which achieves petaflop performance using older process nodes through unique architecture. He explains how Groq's hardware solves latency and throughput issues in AI inference by providing deterministic timing, enabling applications in autonomous vehicles, finance, and nuclear fusion. Ross also reflects on his time at Google X, the importance of empathy in innovation, and the cultural dynamics of building a high-performance computing company.

    Chips, FPGAs & semiconductorsAI & machine learning
    8 chapters

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by EE Times Current. The episode itself belongs to its publisher.