Hardware Podcasts

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

The Circuit · 16 Mar 2025 · 48m · Episode 110

What this episode covers

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.

Most relevant to

large enterprises

Topics

Mentioned

Companies
Deep Seek, High Flyer, NVIDIA, OpenAI, ChipStrat, Meta, Google, Anthropic
Products
ChatGPT, Llama, Mixtral, Gemini, Copilot
Chips
H100, H800

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

  • EE Times Current · 40m

    CEO Interview: Flex Logix’ Geoff Tate on AI | Open Source Hits Hardware

    Geoff Tate, CEO of Flex Logix, explains the distinction between AI training and inference, arguing that engineers must benchmark their specific neural network models on target chips rather than relying on generic metrics like TOPS, which he demonstrates can be misleading compared to real-world performance on models like YOLO V3. Industry analyst Kevin Krewell discusses the rise of open technologies in hardware, tracing the evolution from open source software like Linux to open RAN in networking and the growing ecosystem around the open-source RISC-V architecture as an alternative to proprietary instruction sets like ARM and x86.

    AI & machine learningChips, FPGAs & semiconductors
    6 chapters

  • Hackaday Podcast · 59m

    Ep 306: Bambu Hacks, AI Strikes Back, John Deere Gets Sued, and All About Capacitors

    Hosts Elliot Williams and Dan Maloney discuss the recent DeepSeek AI release, noting its open-source nature and impact on the AI market, alongside security concerns like prompt injection. They review Joshua Wise’s work hacking the Bambu Lab X1 Carbon 3D printer to run custom firmware, highlighting the ongoing battle between the community and the manufacturer's lock-down measures. The episode also covers a hobbyist's backyard antenna farm for deep space communications, technical insights into capacitor selection and orientation, and a legal battle where the FTC sues John Deere over right-to-repair restrictions.

    Embedded firmwareWireless & RF
    7 chapters

  • EE Times Current · 39m

    CTO Interview: Kunle Olukotun of SambaNova

    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.

    AI & machine learningMicrocontrollers & dev platforms
    6 chapters

  • Built 2 Scale · 1h 24m

    Google CRUSHES Everyone + Elon's $10B Anime Crisis + The Age of Agents

    Hosts Matt Perrott and Scott Wilcox discuss Google's recent market dominance, highlighting its Pixel phone launch, instant voice translation, and video generation capabilities that undercut OpenAI's costs. They analyze the strategic implications of DeepSeek being confirmed as a Chinese state operation and Elon Musk's controversial pivot to edge computing and LIDAR-free autonomous vehicles. The episode also covers the rise of AI agents, including Anthropic's Claude browser integration and HeyGen's digital twin technology, while debating the future of corporate AI implementation and Gen Z's role in the industry.

    Connected products & IoTChips, FPGAs & semiconductors
    10 chapters

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