Skip to content
Sign in
Artwork for Semi Doped
Technology

Semi Doped

Vikram Sekar and Austin Lyons

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

Play
Open in
  • 32 episodes
  • weekly
  • Avg 46 min
  • English

Episodes 32

NewestOldestLongest
All20Unplayed20In progress0Played0
Counted on this page — what you have heard stays on this device, so it is not something the list can be paged by.
  • Tuesday · 39 min

    MicroVCSELs vs MicroLEDs, a MicroExplainer

    Austin Lyons and Vik Sekar host a micro-explainer on the two competing micro-emitter technologies for AI interconnects: microLEDs and microVCSELs. They break down Credo's surprising pivot at OCP from its microLED acquisition to a microVCSEL product, revealing the core physics that give microVCSELs superior speed and reach. The episode establishes the technical reasons why the industry is converging on microVCSELs for the critical 5-50 meter gap in AI data centers. This episode is presented by Crusoe. Crusoe's Serverless Fine-Tuning trains an open model like Llama, DeepSeek, or GLM on your own data, and you pay per million tokens, not per GPU hour. About $2 buys roughly 5M training tokens on smaller models. Get $5 in credit to start: https://semidoped.com/crusoe Key Takeaways: - The new battleground for AI interconnects is the 5-50 meter "reach gap"—too long for copper cables limited to ~7 meters, but too short for expensive, supply-constrained InP long-reach optics. - MicroLEDs are limited by physics to ~3-5 Gbps over 5 meters. Their broad light spectrum causes "chromatic dispersion," smearing the signal and creating a hard ceiling on speed and reach. - To achieve high bandwidth, microLEDs require massive parallelism—bundling ~300 slow lanes to hit 1 Tbps—which introduces significant manufacturing complexity in coupling hundreds of fibers. - MicroVCSELs solve this, hitting 50 Gbps+ over 50 meters. Their "stimulated emission" physics creates a narrow, focused beam that suffers less from dispersion, enabling 10x the speed and reach of microLEDs. - The supply chain is a key advantage. MicroVCSELs use a mature Gallium Arsenide (GaAs) process on 8-inch wafers, shared with cell phones, giving them a cost and scale advantage over exotic InP optics. - MicroLEDs' niche is extreme heat. Their Gallium Nitride (GaN) material is highly temperature-insensitive, making them a candidate for very short-reach, high-reliability optics inside a hot server tray. Chapters: 0:00 The Credo Pivot 3:49 The Talk of OCP 8:13 The 5-50 Meter Reach Gap 12:52 How MicroLEDs Work 21:10 MicroLEDs' Parallelism Problem 22:55 How MicroVCSELs Work 26:57 MicroVCSEL Performance Specs 31:51 The Physics of VCSELs 33:35 The Market Chooses MicroVCSELs 35:45 MicroLEDs' Niche: Heat 38:16 Teasing Future Interviews Mentioned in this episode: - PicoJool’s Al Yuen: The Case for GaAs VCSELs in Scale-Up Interconnects (Semi Doped episode): https://daily.semidoped.com/p/picojools-al-yuen-the-case-for-gaas - A MicroLED Field Trip to Avicena (Vik's Newsletter, paid): https://www.viksnewsletter.com/p/a-microled-field-trip-to-avicena - MicroLEDs vs. Lasers: The Linewidth Tradeoff (Vik's Newsletter, free): https://www.viksnewsletter.com/p/microleds-vs-lasers-the-linewidth-tradeoff - Understanding Energy Bands in Semiconductors (Vik's Newsletter, free): https://www.viksnewsletter.com/p/energy-bands-in-semiconductors Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • October 15 · 49 min

    AI Inference Is a Data Movement Problem: Microsoft's Saurabh Dighe on Maia 300

    Microsoft's Maia 300 is a fully co-designed system, optimized to solve the memory-bound "data movement problem" of emerging agentic AI workloads. Saurabh Dighe, Corporate VP of Azure Hardware Systems at Microsoft, sits down with Austin to detail the architecture, from its TSMC N2P chiplet design enabling 18 TB/s of HBM4 memory bandwidth to its use of AI agents for automated kernel optimization. They discuss how this full-stack approach aims to maximize intelligence per watt and per dollar for Microsoft's planet-scale services. This episode is presented by @CrusoeAI. Crusoe's Serverless Fine-Tuning trains an open model like Llama, DeepSeek, or GLM on your own data, and you pay per million tokens, not per GPU hour. About $2 buys roughly 5M training tokens on smaller models. Get $5 in credit to start: https://semidoped.com/crusoe Key Takeaways: - Microsoft views AI inference as a data movement problem, not a compute problem. The energy cost of moving data to feed the compute engine can be two orders of magnitude higher than the compute itself. - The design of Maia 300 is driven by the shift from chat to agentic AI, which creates memory-bound workloads. The system delivers a 7x token throughput increase per rack over Maia 200, with a 5x increase in rack-level memory bandwidth. - Maia 300 uses a chiplet architecture to "extend the shoreline" of its single TSMC N2P compute die. This allows it to accommodate eight HBM4 stacks, delivering 18 TB/s of memory bandwidth. - Microsoft is using AI agents for "agentic kernel optimization" to automate the creation of low-level software. This strategy aims to bypass the manual optimization bottleneck and enable "Day-zero model availability" for new AI models. - The system is built for deployment flexibility. While servers use direct liquid cooling, a rack-level "heat exchanger unit" allows the 167 kW racks to be deployed in traditional air-cooled data centers. - The network is a two-tiered scale-up fabric built on standard Ethernet. Copper connects the 96 accelerators within a rack, while optics connect racks into superpods of up to 12,288 accelerators. Chapters: 0:00 Intro: Saurabh Dighe 1:39 Microsoft's Silicon Philosophy 5:52 Workload Shift: Agentic AI 9:32 The Data Movement Problem 14:20 Crusoe Ad Read 17:47 Maia 300 Specs 24:07 Chiplets & HBM4 29:12 Prefill/Decode Disaggregation 30:33 Two-Tiered Network 35:45 NPO vs. CPO 36:46 Rack-Scale Power & Cooling 41:35 Agentic Kernel Optimization 49:08 What's Next: Maia 400 Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • October 8 · 42 min

    Synopsys’s Amazon Deal and GPT-Synopsys: Ravi Subramanian

    The demand for workload-specific hardware is forcing a strategic shift in the semiconductor IP business. Synopsys's Ravi Subramanian joins Austin to discuss why the company is moving to a new royalty-based model for its custom "application optimized IP," starting with a billion-dollar deal with Amazon. Ravi also unpacks Synopsys's partnership with OpenAI and its vision for AI "agent engineers" that augment human designers, while explaining why physics-based simulation remains the ultimate ground truth. This episode is presented by Crusoe. Try Crusoe's managed inference for open models with $5 in free credits: https://semidoped.com/crusoe "in physics we trust is because at the end of the day, the ground truth to have the confidence, not hallucinate and be deterministic in your outcome is really what defines is a product going to be successful or not." — Ravi Subramanian, Synopsys Key Takeaways: - The hyperscaler IP model is shifting. Because 1-2 applications drive most revenue, companies like Amazon now demand custom “application optimized IP” instead of standards-based IP. - Synopsys is adopting a new business model for this custom work: a license fee, a “customization fee” to prioritize scarce engineers, and for the first time, a royalty on the final chip. - This new royalty model mirrors the Arm-Amazon partnership from a decade ago, aligning Synopsys with its customer’s 12-month design cycles and expanding its role to system-level bring-up. - Synopsys’s AI strategy isn't just tools, but a new consumption model: customers can “hire” an AI agent engineer with a specific skillset, fitting directly into existing budgets and team structures. - The GPT-Synopsys partnership with OpenAI is a turnkey service bundling a specialized LLM, the necessary compute, and forward-deployed engineers to solve the key adoption hurdles of model quality and compute scarcity. - AI is for exploration, but physics is for verification. Accurately modeling a modern CFET requires 200x more device equations than a planar FET, a complexity that demands physics-based simulators as the ground truth. Chapters: 0:00 Intro: The Amazon & OpenAI Deals 2:39 Application Optimized IP 6:04 The New Royalty Business Model 12:14 How Standard IP Differentiates 18:20 The Arm-Amazon Analogy 22:00 Synopsys's Role in Test Chips 26:06 The Three Stages of Chip Design 31:27 The 'Agent Engineer' AI Model 36:20 The GPT-Synopsys Partnership 39:27 In AI We Believe, In Physics We Trust Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • October 1 · 50 min

    Every Muse User Gets 2 CPUs and 8GB in Meta's Cloud. Your Phone Already Has That

    The push to run AI agents on edge devices will not be driven by model labs or hardware vendors, but by service providers like Meta whose ad-based business models create a powerful financial incentive to offload massive cloud compute costs onto the user's own powerful-but-idle hardware. Austin Lyons and Vik Sekar break down why the failed "AI PC" push has given way to a new, economically-driven strategy for on-device AI. They discuss how Qualcomm's acquisition of Modular provides the key software layer and why the ultimate vision is an "ecosystem of you" where the device acts as a context-aware orchestrator between local and cloud compute. This episode is presented by Crusoe. Try Crusoe's managed inference for open models with $5 in free credits: https://semidoped.com/crusoe "[Model Labs] business model is we make money when we do inference in the cloud. And yet we are saying, actually, there's a world where we would really like to just run this stuff locally... Meta's like, bro, I want to help you at the edge." — Austin Lyons, Chipstrat Key Takeaways: - The driver for edge AI is a business model conflict. Model labs like OpenAI sell cloud inference; service providers like Meta see it as a massive cost center for their ad-supported agents like Muse. - Your phone is an underutilized edge server. Its specs—like 8GB of RAM and 2 vCPUs of power—are comparable to the dedicated cloud VMs that services like Meta's Muse provision for each user. - The "AIPC" push failed because it was a hardware-first strategy without a use case. Today's edge AI push is a software-first strategy driven by the need to cut cloud OpEx for mass-market agents. - The role of the edge device is becoming a "context translation layer." It uses private data like location and biometrics to intelligently decide which AI tasks run locally versus in the cloud. - Qualcomm's acquisition of Modular is a key strategic move. It positions the open-source Mojo language as the "write once, run anywhere" software layer for deploying agents across diverse hardware. - Personal agents will lead to the "end of apps." Fixed UIs, which force a designer's paradigm on the user, will be replaced by interfaces generated on demand via natural language. - Google is missing its opportunity to lead in on-device AI. Despite vertical integration from TPUs to Pixel hardware and a leading ad business, it lacks the product execution of players like Meta. Chapters: 0:00 Who Makes Local AI Happen? 1:25 Making the Intro with AI 4:43 Qualcomm's Agent Vision 6:54 Why AIPCs Failed 8:45 The Phone as a Portal 13:52 Meta's Incentive for Edge AI 24:05 The End of Apps & UIs? 32:31 The Ideal Personal Agent 34:25 The Edge as a Context Layer 38:17 Qualcomm's Modular Acquisition 43:16 The Ecosystem of You 48:01 Google's Missed Opportunity Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • September 29 · 56 min

    Flash Hit Its Physical Limit. ReRAM Is the Replacement: WeeBit Nano's Ilan Sever

    As embedded flash hits a physical and economic scaling wall below 22nm, a new on-chip memory is needed. Ilan Sever of WeeBit Nano joins to explain how Resistive RAM (ReRAM) is positioned to become the new standard by offering a low-cost, high-endurance alternative. Ilan details how ReRAM's simple back-end-of-line integration, automotive qualification, and future applications in AI in-memory compute create a compelling replacement for the incumbent technology. "In geometries below 28 or 22 nanometer, flash reached its physical limit, actually. So physically you cannot implement embedded flash anymore and the industry had been looking for alternative technologies for a few decades now." — Ilan Sever, WeeBit Nano Key Takeaways: - The inflection point for on-chip memory is physical. Embedded flash cannot scale below the 28/22nm node because its floating gate becomes too small to reliably hold charge. - The economic problem with eFlash is its 25-30% cost increase, a result of adding 8-10 extra masks to a standard CMOS process—a non-starter for cost-sensitive MCUs. - ReRAM's manufacturing advantage is its simplicity. It's built in the back-end-of-line (BEOL) metal layers, adding only two masks and not interfering with the sensitive transistors below. - The physics are different: ReRAM isn't storing charge, it's changing resistance by using voltage to form or dissolve a conductive filament of oxygen vacancies in a dielectric. - Automotive qualification is the key market validation. WeeBit's ReRAM is qualified for the AEC-Q100 standard and demonstrates 100,000 write cycles—10x that of typical eFlash. - The business model isn't just a recipe; it's a full IP module with the bit cell, analog circuits, digital logic, and programming algorithms, which is what SoC designers actually need. - Beyond replacing flash, ReRAM (as a memristor) is a foundational technology for future in-memory compute (IMC) AI accelerators that perform calculations inside the memory to save power. Chapters: 0:00 The Overlooked Memory 4:15 The eFlash Scaling Wall 6:03 Four Uses of NVM 23:19 The Cost of eFlash 27:29 Introducing ReRAM 32:25 How ReRAM Works 37:12 Fab-Friendly Materials 42:14 Automotive Qualification 44:30 The Vertically Integrated IP Model 51:13 Future: In-Memory Compute 54:19 Innovation Story: Self-Trimming Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • September 24 · 20 min

    Qualcomm's Durga Malladi: HBC vs HBM, Dragonfly AI 250, and Winning on TCO

    Qualcomm re-entered the data center inference market late — competitive GPUs, XPUs, and custom accelerators are already out there. Durga Malladi (EVP/GM, Technology Planning, Edge & Data Center) explains why they didn't build another me-too accelerator and instead placed a differentiating technical bet: High Bandwidth Compute (HBC). The difference between HBC and HBM: HBM is stacked DRAM that shuttles data across wide buses to a separate accelerator — very high bandwidth, but expensive and power-hungry. HBC bonds stacked DRAM directly on top of a custom logic die using wafer-on-wafer techniques, so a lot of the compute runs literally next to memory. The result is lower latency and lower power per bit — and 18x effective bandwidth on the AI 250 versus the AI 200 at the SAME 768 GB LPDDR capacity and the SAME 160 kW per rack. Why the late-mover bet works: Qualcomm's differentiation isn't just architectural. Their foundry and memory-vendor relationships let them credibly promise the multi-megawatt-to-gigawatt supply hyperscalers need. Memory vendors are developing their own HBC-equivalent variants as partners rather than pure competitors, and Qualcomm expects the two technologies to co-evolve. Silicon is back in the lab; proof points land in Q1/Q2 2027. Chapters: 0:00 Introduction 2:49 The Memory Wall Problem 8:05 HBC Physical Architecture 9:33 HBC vs. Custom HBM 11:50 Silicon Proof Points Coming 12:08 Managing Thermals 13:34 Dragonfly AI 250 14:48 Trillion-Parameter Model on One Card 16:56 Tokenomics and TCO 18:26 Why TCO is Key 19:26 Manufacturing at Scale Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • September 18 · 36 min

    Pacing AI Means More Compute, Not Less

    The debate over "pacing" AI for safety is a red herring, as the practical outcome of more safety work is not a slowdown but a net increase in demand for compute hardware. Austin Lyons and Vik Sekar unpack Dario Amodei's call to slow down frontier AI development, arguing that the true governors on AI are not voluntary agreements but physical data center bottlenecks and geopolitical realities. They conclude that the push for more safety and interpretability will ultimately drive more sales of both training (GPUs) and inference (XPUs) hardware. Today's sponsor is G2i — expert training data, evals, and RL environments for AI labs: https://fandf.co/4wZvKHS Chapters: 0:00 Introduction: Pacing AI 1:36 Recapping Dario Amodei's Essay 7:42 Hot Take: Anthropic vs. OpenAI 11:34 The John Deere Playbook 15:22 The Problem with Interpretability 17:53 Pulling the Plug 19:03 The Real Bottleneck: Data Centers 22:03 The Geopolitical Blind Spot 23:41 Counter-narratives: Meta & Open Source 27:26 The Investment Thesis: Pacing Increases Compute 32:29 The Financial System Analogy 36:05 AI Safety as the New Cybersecurity Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • September 14 · 50 min

    TSMC Will Buy High NA After All: ASML's New 6x12-Inch Mask

    The emergence of user-friendly AI agents is creating massive new hardware demand, which in turn is forcing the semiconductor industry into a complex, multi-year transition to High NA EUV lithography and an entirely new 12-inch photomask standard. Austin Lyons and Vik Sekar unpack the physics and economics of ASML's move to High NA, explaining why it necessitates a shift from 6-inch to 6x12-inch masks and what this means for the timelines of TSMC, Intel, and Samsung. "This is where you have to get the whole supply chain coordinated around this. And this is why you really ultimately need ASML's biggest customers to stand up and say, 'We're going to buy this.'" — Austin Lyons, Chipstrat Key Takeaways: - ASML's High NA EUV (0.55 NA) solves resolution but creates a new problem: anamorphic optics (4x by 8x demagnification) cut the exposure field in half, effectively doubling the cost per wafer. - The industry's fix for High NA's halved output is a new 6x12-inch photomask standard — the 12-inch dimension compensates for the 8x demagnification, restoring the full 26mm x 33mm reticle size. - This shift to 12-inch masks is a full supply chain problem, which is why TSMC is waiting until 2030 for high-volume manufacturing while the ecosystem matures over the next 5-7 years. - Intel's aggressive first-mover strategy on High NA — with over 1 million wafers processed to date — is a direct reaction to its disastrous delay in adopting the previous generation of EUV. - Nearly 30% of ASML's revenue comes from recurring services (€2.8B of €9B in Q2), giving it a more stable financial profile than a typical equipment manufacturer. - The new wave of AI agents (Astra, Muse, Instinct) is moving from simple instruction-following to proactive partnership, creating hundreds of millions of new users for server CPUs and memory. - Integrating AI agents into ubiquitous text platforms like WhatsApp is key for mass adoption, as it solves the 'blank text box problem' for non-technical users who can simply text a request. Chapters: 0:00 Episode Opening 2:23 The Rise of AI Agents 6:00 Instinct: The Autonomous Agent 10:50 Productizing AI for Mass Adoption 14:03 From AI Agents to ASML 24:29 The Photomask Stitching Problem 31:49 High NA's Anamorphic Optics Tradeoff 36:39 The Cost of a Halved Reticle 37:26 The 6x12-Inch Mask Solution 39:08 A Full Supply Chain Problem 41:17 TSMC, Samsung & Intel Timelines 48:24 Intel's EUV History Lesson Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • August 27 · 54 min

    OpenAI’s Jalapeño! Feeling Hot Hot Hot!

    Austin and Vik react to OpenAI's Jalapeño announcement at Hot Chips. Plus extra spicy questions like should OpenAI sell it, how much of this was AI-written RTL, and where is Anthropic's chip? Key Takeaways: - The chip's core design philosophy is "dark silicon is cheaper than idle accelerators" — using one balanced chip and power-gating unused blocks is more efficient than a two-chip (e.g. GPU + LPU) solution. - The unprecedented nine-month RTL-to-tapeout cycle was enabled by AI for EDA tools, serving as a wake-up call that small, expert teams can now develop Rubin-class chips in under a year. - Jalapeño's key innovation is a NUMA-style architecture that gives each accelerator a local HBM slice, solving the memory contention that throttles performance in unified memory systems. - OpenAI chose Broadcom's ESUN for its scale-up network to connect 128 chips in the rack at 600 GB/s and up to 2,048 chips across 16 racks at 200G --- all scale up! - The design's "regret factor" principle justifies generality — the opportunity cost of being unable to support a future model is far higher than the marginal cost of adding hardware flexibility upfront. Chapters: 0:00 Hot Chips Reaction 2:32 Designing for User Experience 11:16 A Generalized Inference Chip 14:18 The Foundry-IDM Analogy 18:42 The 'Regret Factor' 21:02 The 9-Month Design Cycle 23:45 Challenging the Two-Chip Solution 35:08 Solving HBM Underutilization 36:46 The NUMA Architecture Solution 39:28 System-Level ESUN Networking 42:06 Dark Silicon vs. Idle Accelerators 49:08 A Wake-Up Call for the Industry 52:59 Where's Anthropic's Chip? Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • August 24 · 44 min

    Grok Bots and How CPUs are used in Agentic AI

    The rise of user-friendly agentic AI platforms will create a massive new demand category for dedicated, high-core-count "agentic CPUs" to execute tasks in parallel, fundamentally reshaping the server CPU market beyond just feeding GPUs. Key Takeaways: - The 'Mac Mini Craze' wasn't about having a GPU on your desk — it was also about security, as users needed a sandboxed machine to run untrusted agent code like OpenClaw, a problem cloud VMs solve too. - In AI servers, the GPU is the 'genius' doing the thinking, while the host CPU is the 'assistant' whose primary job is keeping the GPU fed, requiring high single-core performance. - Agentic tasks create a 'spillover' of parallel work that overwhelms the host CPU, creating a new demand category for dedicated, high-core-count 'agentic CPUs' in separate racks. - The procurement decision for agentic CPUs becomes about cost-per-core, or 'cost per employee' — balancing core count (like AMD's 256-core chips) against single-core speed. - Intel's P-rack (Performance) and E-rack (Efficiency) offerings are a direct response to this need for heterogeneous CPU solutions tailored to different agentic workloads. - The mass adoption of agentic AI could create demand for a billion new CPU cores in the cloud, driven by the convenience of 'easy button' platforms over self-hosting. - A key bottleneck to this heterogeneous future is the orchestration software needed to schedule jobs across different CPUs and accelerators from multiple vendors. Chapters: 0:00 Introducing Grok bot 3:45 Grok bot's Cloud VM Architecture 6:28 The 'Mac Mini Craze' Explained 14:38 Three CPU Deployment Models 16:02 GPU as Genius, CPU as Assistant 21:36 The Limits of the Host CPU 25:20 The 'Office Building' Analogy 29:09 Cost-Per-Core is the Metric 31:20 Intel's P-rack and E-rack 37:32 The Orchestration Bottleneck 40:19 Where Grok bot's VM Lives 44:22 The Unanswered Question Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • August 18 · 55 min

    Tensordyne's R K Anand: HPE Juniper Fabric, Logarithmic Math, MoE Inference, Air Cooling, 3nm

    Tensordyne co-founder and CPO R K Anand joins Austin to discuss the company's strategy for disrupting AI inference. RK explains how Tensordyne combines power-efficient logarithmic math with a battle-hardened networking fabric from partner HPE Juniper. The result is a high-density, air-cooled system designed to efficiently run massive Mixture-of-Experts models in existing data centers. Key Takeaways: - The core innovation isn't just log math, it's the patented method for accumulation. This turns expensive multiplications into cheap additions, freeing die space for a massive on-chip SRAM cache. - Networking is a partnership, not a project. Tensordyne leverages HPE Juniper's 7th-gen router fabric, skipping development cycles to get a 1-2 microsecond latency solution ideal for random MoE traffic. - The power and density claims are radical. By combining log math silicon with an air-cooled fabric, Tensordyne packs 72 chips into a 13U chassis at just 30 kW — a quarter of the space and power of an NVL72. - One go-to-market advantage is air cooling. The 30 kW, 19-inch rack system can be deployed in existing 'brownfield' enterprise and telco data centers that cannot support liquid cooling. - Partnerships de-risk the aggressive timeline. Broadcom provides access to TSMC 3nm and HBM, while strategic investor HPE Juniper provides the carrier-grade fabric with 'five nines' reliability. Chapters: 0:00 Introducing Tensordyne 5:32 The Juniper vs. Cisco Playbook 11:29 Origin Story: Automotive Power Constraints 15:37 The Secret Sauce of Log Math 18:02 Pivoting to the Data Center 22:08 Leveraging a Router Backplane for AI 27:22 Why Router Fabrics Suit MoE Models 34:12 The Three Phases of Inference Hardware 37:40 How One Chip Handles Pre-fill & Decode 40:34 The 'Too Good to Be True' System Specs 43:31 Go-to-Market: The Air-Cooled Advantage 48:21 De-risking with Strategic Partnerships 52:37 Solving the Software Problem with AI Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/ - The software moat is eroding. Tensordyne argues that modern agentic AI workflows can now automate the generation of optimized software kernels, solving the classic adoption problem for new hardware.

    • Transcript
  • August 11 · 34 min

    NEWS TAKE: China's Optical Ban, Volta's $10B Anthropic Deal, AMD's Earnings

    Austin Lyons and Vik Sekar break down three stories that hit semis last week. They analyze a proposed US ban on Chinese optical transceivers that threatens to cut off 50% of the global supply, the market's reaction to AMD's surprise $800M CapEx spend despite strong earnings, and how a new company called Volta Infrastructure landed a $10B compute deal with Anthropic by pioneering a new financial model for AI. Key Takeaways: - The proposed ban on Chinese optical transceivers is based on a flawed security rationale—the components are simple signal converters, not a meaningful vector for malware. - A ban would be self-defeating, as it would cut off the ~50% of global transceiver supply assembled in China, creating the very data center disruption it claims to prevent. - AMD's successful pivot to a data-center-first company (58% of revenue) is being scrutinized for its high CapEx—$800M vs an expected $200-300M—revealing the hidden costs of securing supply. - Nvidia's use of on-chip SRAM for inference's decode phase highlights a strategic gap for AMD, which lacks a compelling SRAM-based solution to compete on disaggregated workloads. - Volta Infrastructure's $10B deal with Anthropic is an innovation in finance, not tech; it applies low-cost 'infrastructure debt' to AI compute by framing clusters as predictable 'token factories'. - The Volta deal was necessary because all existing CSP capacity is allocated; it acts as a 'clean balance sheet' SPV to secure low-cost debt for Anthropic's new, dedicated Nvidia capacity. Chapters: 0:00 Intro: News Take Format 0:40 The China Optical Ban 4:02 A Flawed Security Rationale 11:30 Market & Supply Chain Impact 15:07 Investment Paralysis 15:52 AMD's Earnings Scrutiny 24:46 AMD's Missing SRAM Strategy 25:33 Volta's $10B Anthropic Deal 28:55 Volta's 'Toll Road' Model 32:19 Why the Volta Deal Was Necessary 33:38 The 'One Customer' Counterpoint 34:27 Wrap: Tech & Financial Innovation Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • August 7 · 39 min

    GlobalFoundries Thomas Barber: CPO, Silicon Photonics, 300mm, SiGe, OCI, NRZ

    The physical limits of copper are forcing a shift to optical interconnects in AI data centers. Austin sits down with GlobalFoundries' Thomas Barber to unpack why GF thinks it can lead that transition twice over: once with its long-running silicon photonics platform, and again with the specialty Silicon Germanium process the industry needs to drive it. "The enemy to me right now is copper. I'm trying to beat copper, right? If TSMC wins and we win, that's great because we're both displacing copper. And until all the copper is gone, there's plenty of market to go around." — Thomas Barber, GlobalFoundries Key Takeaways: - Copper's usable range halves every time the data rate doubles, and 200 Gbps/lane inside a rack-scale AI cluster is already past the point copper can handle. - CPO's real win isn't speed, it's power: saving 20-25 pJ/bit frees part of a data center's fixed 50-100 MW budget to go toward compute instead of moving bits. - GlobalFoundries leads photonics revenue for an unglamorous reason: it moved to 300mm wafers early, which yields 2.25x more die per wafer than the 200mm lines rivals still run. - CPO can end up more reliable than the pluggables it's replacing, not less, because it deletes the physical plug connector, and dust at that connector is the leading cause of field failures. - The OCI MSA picks NRZ over faster PAM4, deliberately going wide and slow, because NRZ's native bit error rate is a million times lower, which simplifies the receiver and cuts power. - GlobalFoundries stacks two specialty processes into one edge: micro-mirror couplers on the photonic side, and Silicon Germanium transistors hitting 350-400 GHz on the electrical side driving them. - The real competitor for silicon photonics isn't another foundry, it's copper itself — and that market is big enough that GlobalFoundries and TSMC can both win without taking share from each other. Chapters: 0:00 GlobalFoundries in Photonics 1:21 GF's Photonics Strategy 3:02 GF's Market Leadership 5:06 300mm Wafer Advantage 6:55 Copper's Range Limits 9:18 Pluggable to CPO 13:48 CPO Reliability 19:22 OCI MSA Explained 26:20 GF's Scale Platform 29:04 Micro Mirror Technology 31:32 Photonics vs. Copper 38:25 Silicon Germanium Advantage Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • August 1 · 47 min

    How Retimers Built an $80B Company: The Story of Astera Labs

    A retimer is a small, unglamorous chip. Astera turned it into a high-margin franchise. Austin and Vik cover why copper gives up at PCIe speeds, what a retimer does that a redriver can't, and how one H100 socket became a moat. Then Scorpio: can Astera take switch share from Broadcom? Key Takeaways: - 30cm of PCB trace at Gen 5's 32 GT/s and the signal is gone. Signal integrity is an inside-the-server problem, not just a rack-to-rack one. - Gen 6 swaps NRZ for PAM4. Four voltage levels instead of two — a third the eye height, ~9.5 dB of SNR gone. - A redriver amplifies, noise included. A retimer recovers the clock and data, rebuilds the signal, and resets the jitter budget. - Astera won H100 by shipping PCIe 5 retimers at volume first, then bundling COSMOS on top. The chip became a fleet telemetry sensor. - Sticky socket: H100 carried straight into an Aries 6 design-in on Blackwell. - Scorpio moves up the stack. Same signal conditioning IP, higher ASP, pointed at Broadcom's PCIe switch business. - Astera granted Amazon a warrant — 3.26M shares at $142.82, vesting against up to $6.5B of purchases through 2033. Equity flowing to the customer, and it sizes the Scorpio P-Series commitment behind Trainium 2 and 3. Chapters: 0:00 The Astera Labs Franchise 3:32 The PCIe Copper Problem 6:53 Signal Smearing and Jitter 11:52 Eye Diagrams Explained 14:59 Equalization: Fixing the Signal 23:52 Redriver vs. Retimer 29:30 How Astera Won Nvidia 35:22 The Blackwell Scare 38:50 Scorpio: Moving into Switches 42:52 The UALink vs. Ethernet Battle 45:58 Taurus and Leo Product Lines Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • July 29 · 17 min

    News Take: Hyperscaler CDS, SK Hynix Earnings, China's DUV

    Austin Lyons and Vik Sekar break down the market's recent volatility. They discuss the rising fear around hyperscaler debt, visible in credit default swap (CDS) premiums, and the counter-argument that GPU ROI is actually understated. They then analyze the paradox of SK Hynix's massive stock drop despite record growth, and explain why China's new DUV lithography tool makes 3D chip stacking essential to their strategy. Key Takeaways: * Hyperscaler debt fears are quantified by rising credit default swap (CDS) premiums, signaling investor nervousness that the AI CapEx firehose is shifting from cash flow to riskier debt. * The counter-take on debt is that hyperscalers are under-earning on GPU assets — as long-term contracts re-price to spot rates 2x higher, the hardware's ROI will justify the financing. * SK Hynix's 20% stock drop despite 257% YoY revenue growth shows a market priced for perfection, where a small miss against consensus triggers a panic disconnected from fundamentals. * The memory market paradox: commodity DRAM is currently more profitable than HBM because HBM's production is 3x less bit-efficient per wafer, consuming scarce supply while DRAM spot prices soar. * China's reported immersion DUV breakthrough is equivalent to ASML's late 2000s-era technology, a step toward self-sufficiency but not a leap to the leading edge. Chapters: 0:00 The Market Is Freaking Out 0:21 Hyperscaler Debt and Credit Default Swaps 3:56 The Counter-Take: Under-Earning on GPUs 6:18 SK Hynix: Record Growth, Market Carnage 8:10 The HBM vs. DRAM Profitability Paradox 10:35 Leverage, Retail, and Panic Selling 12:05 China's Immersion DUV Breakthrough 13:28 The Limits of DUV: 7nm and Multi-Patterning 14:14 Logic Folding and 3D Stacking 16:12 A Long Journey to Self-Sufficiency 17:05 Surviving Semiconductor Cyclicality Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/ Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr

    • Transcript
  • July 25 · 49 min

    Datacenter Interconnects: Copper vs. Optics, Nvidia's 78-Layer PCB, Co-Packaged Optics (CPO)

    Vik Sekar and Austin Lyons tackle the biggest bottleneck in inference: moving data. They break down the three tiers of datacenter networking — scale up, scale out, scale across — and the core engineering trade-off at each layer: copper vs. optics. Topics include Nvidia's extreme measures to keep scale-up fabric electrical (a 78-layer mid-plane PCB), why Co-Packaged Optics is the "holy grail" everyone wants and no one can ship, and the serviceability problem standing in its way. Key Takeaways: A single 72-GPU scale-up rack needs 5,000+ cables spanning ~2 km — at that density, the power and cost of every connection becomes a system-level constraint, not an implementation detail. The scale-up rule is "copper when you can, optics when you must": each pluggable optical module adds ~30W, and with thousands of links in the fabric, that penalty compounds fast. Nvidia is pushing copper past its usual limits with a 78-layer mid-plane PCB — 3x+ the layer count of a typical complex board — specifically to avoid paying the optics power tax in its scale-up fabric. Scale-up isn't just a training problem. Frontier MoE models need 72-GPU domains to hit memory bandwidth targets, which pulls high-performance interconnect into the inference conversation. Scale-up has the highest connection density of the three tiers, making it the largest TAM and the sharpest three-way fight between NVLink, UA-Link, and Ethernet. Co-Packaged Optics could cut interconnect power by two-thirds — but a single failed laser could brick an entire multi-thousand-dollar GPU package, and that serviceability risk is what's keeping it on the roadmap instead of in racks. Chapters: 0:00 The Biggest Problem in Computing 7:14 The Three Tiers of Networking 13:49 Scale Up: Copper vs. Optics 17:59 Front-End vs. Back-End Networks 22:12 The Physical Scale of Cabling 28:46 Nvidia's 78-Layer Mid-Plane 33:16 How Optical Transceivers Work 39:16 The Power Penalty of Pluggables 42:09 The Business of Speed Transitions 44:40 The Promise and Peril of CPO 51:45 The Holy Grail of Networking Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/austinsemis Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free: https://daily.semidoped.com/

    • Transcript
  • July 16 · 49 min

    PicoJool's Al Yuen: The Case for GaAs VCSELs in Scale-Up Interconnects

    Al Yuen, CEO of PicoJool, talks with Austin about using VCSELs for scale-up optical interconnects in AI data centers. Al explains why gallium arsenide (GaAs) supply is unconstrained while indium phosphide (InP) is limited, and how PicoJool can leverage existing supply chains to ship in the millions per month. They cover the roadmap from 1.6T to 12.8T, the trade-offs among 8×200G, 16×100G, and 32×50G, and the push to train the next generation of photonics engineers. The spec that changed is error rate. AI needs error-free links (10⁻¹⁰–10⁻¹² vs Ethernet's old 10⁻⁶) because GPUs act as one system — one error stalls the whole job. Single-mode optics can't hit the volume. Built for ~100K city-to-city links; datacenters need millions/month — a 10–50× gap GaAs vs InP is the whole supply story. VCSELs run on unconstrained GaAs; single-mode on constrained InP. 1M units: 8–10 weeks vs a sold-out 8–18 months. One platform, three flavors to 1.6T: 8×200G "fast and narrow," 16×100G LPO low-power, 32×50G NRZ "slow and wide" — pick on power, cost, and BER. 12.8T needs no new tech — just more of the same: 64 channels (4×16 array in a finger-sized connector) × 200G, scaling channels, bi-di wavelengths, and lane speed. Capacity isn't the time consuming part — qualification is. WIN runs ~1,000 wafers/week at ~240K VCSELs each, so 1M units = ~10 wafers. The wait is tier-1 (>6 mo) and tier-2 (~3 mo) quals. Chapters: 0:00 Meet Al Yuen and PicoJool 2:29 Inventing the active optical cable 5:03 Engineering mindset, copper limits 8:43 Why VCSELs 13:45 Scale-up and bit error rate 20:09 Unconstrained vs constrained supply 21:53 Indium phosphide bottleneck 25:49 VCSEL design and foundry handoff 31:59 Product road map, 200G launch 34:16 Path to 3.2T and 12.8T 40:10 Ordering a million VCSELs 45:10 Ramp timing and training new engineers Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free! Sign up: https://daily.semidoped.com/

    • Transcript
  • July 10 · 1 hr 4 min

    WEKA's Val Bercovici: KV Cache, DeepSeek V4, HBF, SLC vs QLC NAND, CXL, NVLink, Tokenomics

    Vik welcomes Val Bercovici from Weka to discuss the rapidly evolving landscape of AI memory and storage. Val explains how Weka's architecture leverages high-bandwidth networks to make storage faster than motherboard DRAM. They dive into KV cache optimizations, the future of NAND flash tiers, and the role of CXL in AI inference. The episode concludes with a look at predictive memory offloading and the AI flywheel. Chapters: 0:00 Welcome Val Bercovici, Weka 1:59 Memory situation and model routing 3:50 KV cache offloading to CMX 6:10 Network faster than motherboard 13:10 Weka as AI memory infrastructure 14:45 Inference market is different 16:06 Memory hierarchy and KV cache 19:40 KV cache optimizations and demand 25:20 DeepSeek's cache read pricing 34:49 NAND flash tiers: SLC vs QLC 43:01 High Bandwidth Flash (HBF) 49:59 CXL versus other interconnects Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free! Sign up: https://daily.semidoped.com/

    • Transcript
  • July 3 · 50 min

    Micron's Record Profits, Apple's CXMT Plea: AI is Eating All the Memory

    Austin and Vik break down the current memory crisis, where AI demand is driving unprecedented profits for memory makers like Micron, while simultaneously causing inflation and price hikes for consumer electronics. They discuss how even premium brands like Apple are struggling to secure supply, leading them to controversial suppliers. The hosts also explore the long-term implications of AI's insatiable memory appetite and its impact on both data center and consumer markets. Chapters: 0:00 Memory crisis hits 1:20 AI impacting consumers 3:00 AI causing inflation 6:48 Consumer demand drop? 8:49 AI demand inelastic 10:56 Long-term memory outlook 11:02 GoPro's memory woes 12:24 Apple's pricing power 18:20 Apple seeks CXMT DRAM 21:38 Shrinkflation for phones 23:16 Korea's memory investment 26:10 Micron's killing profits 33:00 Why AI needs so much DRAM 40:27 Future of AI training 44:34 Cost-optimizing inference Follow Chipstrat: Newsletter: https://www.chipstrat.com X: https://x.com/chipstrat Follow Vik: Newsletter: https://www.viksnewsletter.com/ X: https://x.com/vikramskr Follow Semi Doped: Get more of Austin and Vik daily, free! Sign up: https://www.semidoped.com/

    • Transcript
  • June 29 · 59 min

    Qualcomm's HBC Memory, Alphawave, Modular, and more

    Qualcomm flew Austin to New York for its investor day, where the communications company laid out a plan to make data center, automotive, and IoT two-thirds of its business by FY29. Austin was in the room and asked Cristiano Amon a question on the record. Vik watched the whole thing on YouTube. Together they break down what actually changed. The technical centerpiece is High Bandwidth Compute (HBC): stacking LPDDR on top of logic to expose the whole face of the chip for interconnects, claiming up to 100x more lanes and a path around the HBM bandwidth bottleneck. Austin and Vik dig into what's really under that memory, why "no advanced packaging needed" just moves the hard problem somewhere worse, and how it compares to d-Matrix. And as Austin put it: don't let anyone tell you HBM is dead. MOAR memory. Then the roadmap: the AI200/250/300 accelerators, the C1000 server CPU (5 GHz, 250+ cores, Meta as a customer), the Alphawave and Modular acquisitions, and Chris Lattner's Mojo. The most interesting takeaway might not be the data center at all. Qualcomm's edge play — AI-defined vehicles, cars as token generators, and a $1T robotics opportunity by 2040 — could be where High Bandwidth Compute matters most. Chapters: 0:00 Communications? That's just the start 4:08 Inside Qualcomm's investor day 9:16 Can Qualcomm build a data center business? 13:09 Disaggregated inference opens the door 17:57 High Bandwidth Compute: memory on the XPU 30:29 "No advanced packaging" just moves the problem 36:20 The roadmap, Alphawave, and Modular 46:00 The C1000 CPU and the agentic shortage 50:40 Cars as token generators, the $1T robotics bet 57:32 The memory market: MOAR Follow Semi Doped: Get more of Austin and Vik daily, free! Sign up: https://www.semidoped.com/ Connect with Vik and Austin: Vik's Paid Substack: https://www.viksnewsletter.com Austin's Paid Substack: https://www.chipstrat.com

    • Transcript
Showing 1–20 of 32 episodes