Semi Doped

Vikram Sekar and Austin Lyons

The business and technology of semiconductors. Alpha for engineers and investors alike.

Episodes

  1. Can Pre-GPT AI Accelerators Handle Long Context Workloads?

    1 DAY AGO

    Can Pre-GPT AI Accelerators Handle Long Context Workloads?

    OpenAI's partnership with Cerebras and Nvidia's announcement of context memory storage raises a fundamental question: as agentic AI demands long sessions with massive context windows, can SRAM-based accelerators designed before the LLM era keep up—or will they converge with GPUs? Key Takeaways 1. Context is the new bottleneck. As agentic workloads demand long sessions with massive codebases, storing and retrieving KV cache efficiently becomes critical. 2. There's no one-size-fits-all. Sachin Khatti's (OpenAI, ex-Intel) signals a shift toward heterogeneous compute—matching specific accelerators to specific workloads. 3. Cerebras has 44GB of SRAM per wafer — orders of magnitude more than typical chips — but the question remains: where does the KV cache go for long context? 4. Pre-GPT accelerators may converge toward GPUs. If they need to add HBM or external memory for long context, some of their differentiation erodes. 5. Post-GPT accelerators (Etched, MatX) are the ones to watch. Designed specifically for transformer inference, they may solve the KV cache problem from first principles. Chapters   - 00:00 — Intro   - 01:20 — What is context memory storage?   - 03:30 — When Claude runs out of context   - 06:00 — Tokens, attention, and the KV cache explained   - 09:07 — The AI memory hierarchy: HBM → DRAM → SSD → network storage   - 12:53 — Nvidia's G1/G2/G3 tiers and the missing G0 (SRAM)   - 14:35 — Bluefield DPUs and GPU Direct Storage   - 15:53 — Token economics: cache hits vs misses   - 20:03 — OpenAI + Cerebras: 750 megawatts for faster Codex   - 21:29 — Why Cerebras built a wafer-scale engine   - 25:07 — 44GB SRAM and running Llama 70B on four wafers   - 25:55 — Sachin Khatti on heterogeneous compute strategy   - 31:43 — The big question: where does Cerebras store KV cache?   - 34:11 — If SRAM offloads to HBM, does it lose its edge?   - 35:40 — Pre-GPT vs Post-GPT accelerators   - 36:51 — Etched raises $500M at $5B valuation   - 38:48 — Wrap up

    38 min
  2. An Interview with Innoviz CEO Omer Keilaf about current LiDAR market dynamics

    5 DAYS AGO

    An Interview with Innoviz CEO Omer Keilaf about current LiDAR market dynamics

    Innoviz CEO Omer Keilaf believes the LIDAR market is down to its final players—and that Innoviz has already won its seat. In this conversation, we cover the Level 4 gold rush sparked by Waymo, why stalled Level 3 programs are suddenly accelerating, the technical moat that separates L4-grade LIDAR from everything else, how a one-year-old startup won BMW, and why Keilaf thinks his competitors are already out of the race. Omer Keilaf founded Innoviz in 2016. Today it's a publicly traded Tier 1 supplier to BMW, Volkswagen, Daimler Truck, and other global OEMs. Chapters   00:00 Introduction   00:17 Why Start a LIDAR Company in 2016?   01:32 The Personal Story Behind Innoviz   03:12 Transportation Is Still Our Biggest Daily Risk   04:28 The 2012 Spark: Xbox Kinect and 3D Sensing   06:32 From Mobile to Automotive: Finding the Right Platform   07:54 "I Didn't Know What LIDAR Was, But I'd Do It Better"   08:19 How a One-Year-Old Startup Won BMW   10:04 Surviving the First Product   11:23 From Tier 2 to Tier 1: The Volkswagen Win   13:47 Lessons Learned Scaling Through Partners   14:45 The SPAC Decision: A Wake-Up Call from a Competitor   16:42 From 200 LIDAR Companies to a Handful   17:27 NREs: How Tier 1 Status Funds R&D   18:44 Why Automotive-First Is the Right Strategy   19:45 Consolidation Patterns: Cameras, Radars, Airbags   20:31 "The Music Has Stopped"   21:07 Non-Automotive: Underserved Markets   23:51 Working with Secretive OEMs   25:27 The Press Release They Tried to Stop   26:42 CES 2025: 85% of Meetings Were Level 4   27:40 Why Level 3 Programs Are Suddenly Accelerating   28:33 The EV/ADAS Coupling Problem   29:49 Design Is Everything: The Holy Grail Is Behind the Windshield   31:13 The Three-Year RFQ: Grill → Roof → Windshield   32:32 Innoviz3: Small Enough for Behind-the-Windshield   34:40 Innoviz2 for L4, Innoviz3 for Consumer L3   36:38 What's the Real Difference Between L2, L3, and L4 LIDAR?   38:51 The Mud Test: Why L4 Demands 100% Availability   40:50 "We're the Only LIDAR Designed for Level 4"   42:52 Patents and the Maslow Pyramid of Autonomy   44:15 Non-Automotive Markets: Agriculture, Mining, Security   46:15 Closing

    47 min

  3. 19 JAN

    LiDAR, Explained: How It Works and Why It Matters

    Austin and Vik discuss why LiDAR is important for autonomy, how modern systems work, and how the technology has evolved. They compare Time of Flight and FMCW architectures, explain why wavelength choice matters, and walk through the tradeoffs between 905 nm and 1550 nm across eye safety, cost, and performance. The discussion closes with a clear-eyed look at competition, Chinese suppliers, and supply chain risk. Chapters (00:00) Introduction to LiDAR and why it matters (05:40) The case for LiDAR in autonomous vehicles (12:41) Wavelengths, eye safety, and system tradeoffs (15:38) How LiDAR works: Time of Flight vs. FMCW (20:12) Mechanical vs. solid-state LiDAR designs (27:31) Market dynamics, competition, and geopolitics

    36 min

  4. 12 JAN

    Nvidia CES 2026

    Episode Summary Austin and Vik break down NVIDIA’s CES 2026 keynote, focusing on Vera Rubin, DGX Spark and DGX Station, uneducated investor panic, and physical AI. Key Takeaways DGX Spark brings server-class NVIDIA architecture to the desktop at low power, aimed at developers, enthusiasts, and enterprises experimenting locally.  DGX Station functions more like a mini-AI rack on-prem: Grace Blackwell for inference and development without full racks The historical parallel is mainframes to minicomputers, expanding compute TAM rather than displacing cloud usage.  On-prem AI converts some GPU rental OpEx into CapEx, appealing to CFOs  NVIDIA positioned autonomy as physical AI with vision-language-action models and early Mercedes-Benz deployments in 2026.  Vera Rubin integrates CPU, GPU, DPU, networking, and photonics into a single platform, emphasizing Ethernet for scale-out. (Where was the Infiniband switch?) The new Vera CPU highlights rising CPU importance for agentic workloads through higher core counts, SMT, and large LPDDR capacity.  Rubin GPU’s move to HBM4 and adaptive precision targets inference efficiency gains and lower cost per token.  Context memory storage elevates SSDs and DPUs, enabling massive KV cache offload beyond HBM and DRAM.  Cable-less rack design and warm-water cooling show NVIDIA’s shift from raw performance toward manufacturability and enterprise polish.

    47 min

  5. 8 JAN

    Insights from IEDM 2025

    Austin and Vik discuss key insights from the IEDM conference.  They explore the significance of IEDM for engineers and investors, the networking opportunities it offers, and the latest innovations in silicon photonics, complementary FETs, NAND flash memory, and GaN-on-silicon chiplets.  Takeaways Penta-level NAND flash memory could disrupt the SSD marketGaN-on-Silicon chiplets enhance power efficiencyComplementary FETsOptical scale-up has a power problemThe future of transistors is still bright

    42 min

  6. 5 JAN

    Nvidia "Acquires" Groq

    Key Topics What Nvidia actually bought from Groq and why it is not a traditional acquisitionWhy the deal triggered claims that GPUs and HBM are obsoleteArchitectural trade-offs between GPUs, TPUs, XPUs, and LPUsSRAM vs HBM. Speed, capacity, cost, and supply chain realitiesGroq LPU fundamentals: VLIW, compiler-scheduled execution, determinism, ultra-low latencyWhy LPUs struggle with large models and where they excel insteadPractical use cases for hyper-low-latency inference:Ad copy personalization at search latency budgetsModel routing and agent orchestrationConversational interfaces and real-time translationRobotics and physical AI at the edgePotential applications in AI-RAN and telecom infrastructureMemory as a design spectrum: SRAM-only, SRAM plus DDR, SRAM plus HBMNvidia’s growing portfolio approach to inference hardware rather than one-size-fits-all Core Takeaways GPUs are not dead. HBM is not dead.LPUs solve a different problem: deterministic, ultra-low-latency inference for small models.Large frontier models still require HBM-based systems.Nvidia’s move expands its inference portfolio surface area rather than replacing GPUs.The future of AI infrastructure is workload-specific optimization and TCO-driven deployment.

    41 min
  • 6 Episodes

About

The business and technology of semiconductors. Alpha for engineers and investors alike.

Information

  • Creator
    Vikram Sekar and Austin Lyons
  • Years Active
    2K
  • Episodes
    6
  • Rating
    Clean
  • Copyright
    © 2026 Semi Doped
  • Show Website
    Semi Doped
  • Provider