Hardware Podcasts

CTO Interview: Kunle Olukotun of SambaNova

EE Times Current · 29 Oct 2021 · 39m

What this episode covers

Kunle Olukotun, chief technologist at SambaNova, explains the company's dataflow architecture designed to overcome the memory and bandwidth limitations of traditional GPU-based AI systems. He details how SambaNova's hardware and Sofware stack, SambaFlow, enable efficient simultaneous training and inference, allowing for dynamic model adaptation and the handling of massive embedding tables. Olukotun also introduces the concept of 'dataflow threads' to bridge the gap between regular dense matrix operations and irregular sparse computations, addressing the unsustainable growth rate of model parameters.

Most relevant to

companies scaling up

Topics

Mentioned

Companies
SambaNova, Argonne National Lab, NVIDIA
Products
SambaFlow, Google TPU
Tools
PyTorch, TensorFlow, CUDA
Also
Kunle Olukotun

Related episodes from other shows

Other shows on the same ground, closest first.

  • Oxide and Friends · 1h 31m

    AI Disruption: DeepSeek and Cerebras

    Andy Hock and James Wang of Cerebras discuss how their wafer-scale silicon executes open-weight models like DeepSeek R1 Distill Llama 70B faster than any number of GPUs, achieving 57x speedup by keeping all 44GB of model weights in on-chip SRAM. They explain that the release of efficient, open models like DeepSeek and Llama transformed Cerebras from an exotic supercomputer vendor into a provider of commodity inference APIs, standardizing the unit of sale as tokens. The guests highlight that hardware-software codesign and the shift toward reasoning models allow smaller teams to compete with well-funded incumbents by optimizing algorithms for specific architectures rather than relying on brute-force scaling.

    AI & machine learningChips, FPGAs & semiconductors
    10 chapters

  • The Circuit · 52m

    EP 163: Breaking the Memory Wall: Micron’s Strategy for the AI Era

    Jeremy Werner, SVP and GM of Micron’s Core Data Center Business Unit, explains how AI is shifting the memory industry from a cyclical model to sustainable growth driven by the 'memory wall' in inference. He details Micron's strategy to break this bottleneck through innovations like HBM4 and ultra-high-capacity SSDs, which address critical data center constraints of power and physical space. The conversation covers the technical hierarchy of memory, from HBM to context storage, and Micron's aggressive capacity expansion plans to meet surging demand for AI training and inference workloads.

    Chips, FPGAs & semiconductorsPCB design & layout
    7 chapters

  • Moore's Lobby · 43m

    Ep. 29 | NVIDIA CTO Michael Kagan on the New Age of AI and Supercomputers

    Michael Kagan, CTO of NVIDIA, discusses the evolution of high-performance computing, tracing his career from Intel and the 860 processor to co-founding Mellanox and its eventual acquisition by NVIDIA. He explains how NVIDIA has shifted from a graphics company to an accelerated computing platform, integrating GPUs, CPUs, and DPUs to support AI workloads. Kagan emphasizes that modern data centers function as unified computing units, delivering processing as a service rather than standalone hardware, driven by the massive computational demands of training neural networks.

    Chips, FPGAs & semiconductorsConnected products & IoT
    8 chapters

  • The Circuit · 48m

    Episode 109: Unpacking the Nuances of Deep Seek with Austin Lyons

    Austin Lyons, founder of ChipStrat, joins Jay Goldberg to analyze Deep Seek, a self-funded AI lab from China that emerged from the quantitative hedge fund High Flyer. They discuss Deep Seek's technical innovations, including its efficient Mixture of Experts (MoE) architecture and the R1 reasoning model, which demonstrated that high-performance AI is possible despite US chip export controls limiting access to NVIDIA H100s. The conversation explores how Deep Seek's training efficiency and open-weight models have shifted the industry focus from pure pre-training scaling to inference optimization and post-training techniques.

    AI & machine learningChips, FPGAs & semiconductors
    7 chapters

  • The Circuit · 41m

    Episode 85: OpenAI leaves AI Silicon in the Dust

    Ben Bajarin and Jay Goldberg discuss how OpenAI's new model, which uses reinforcement learning and multi-level reasoning, drastically increases inference compute demands, leaving current silicon architectures behind. They analyze the competitive landscape, noting that while Nvidia dominates training, the inference market is becoming more open due to high costs and the need for edge deployment. The hosts also highlight Google's recent demos, including a podcast generator and a game engine, as evidence of rapid software innovation that continues to outpace hardware development.

    AI & machine learningChips, FPGAs & semiconductors
    7 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.