Hardware Podcasts

Google TPU vs Nvidia GPUs: Token Per Watt Dominance, or the Era of Scaling Is Over?

Built 2 Scale · 26 Nov 2025 · 1h 08m

What this episode covers

Founders Matt Perrott and Scott Wilcox discuss Google's Gemini 3.0 training on TPUs, which bypasses Nvidia's hardware to achieve a structural cost advantage, and the resulting threat to 'vibe coding' startups like Lovable. They analyze the 'age of scaling is over' consensus among researchers like Ilya Sutskever, arguing that future breakthroughs require real-world spatial intelligence rather than just more compute. The hosts also debate office culture, remote work dynamics, and the three critical roles founders must hire early: product, brand, and sales.

Most relevant to

early-stage companies

Topics

Mentioned

Companies
Google, Nvidia, SpaceX
Products
Gemini 3.0, Lovable
Chips
TPU
Also
Ilya Sutskever, Marc Andreessen, Elon Musk

Related episodes from other shows

Other shows on the same ground, closest first.

  • The Circuit · 46m

    EP 143: NVIDIA Earnings! Google and Gemini TPUs FTW

    Ben Bajarin and Jay Goldberg analyze NVIDIA's recent earnings, focusing on the implications of its $26 billion cloud compute pre-purchase commitment and the economic lifecycle of GPUs. They discuss the competitive landscape between NVIDIA's GPUs and custom AI ASICs from Google and Broadcom, noting Jensen Huang's claims about NVIDIA's dedicated focus. The hosts then evaluate Google's release of the Gemini 3 model and the 'Nano Banana' image creation tool, highlighting how these multimodal capabilities are driving practical adoption for non-technical users.

    Pricing, margins & unit economicsChips, FPGAs & semiconductors
    6 chapters

  • The Circuit · 51m

    Episode 88: AI Accellerators, Startup graveyard? With Special Guest Austin Lyons

    Austin Lyons of Chip Strat joins Ben Bajarin and Jay Goldberg to discuss the 'startup graveyard' for AI accelerator companies, noting 67 US startups are struggling against NVIDIA's dominance. The hosts argue that software moves too fast for dedicated ASICs to remain relevant, making flexible GPUs the short-term winner for both training and inference. They highlight Google's TPU success due to vertical integration, while questioning the viability of startups like Groq, Cerebras, and SambaNova that lack hyperscaler backing or control over their target workloads.

    Chips, FPGAs & semiconductorsAI & machine learning
    8 chapters

  • The Circuit · 38m

    Episode 8: AI Infrastructure and Silicon Demands

    Ben Bajarin and Jay Goldberg discuss the immense compute demands of training large foundational models like GPT-3 and GPT-4, noting that Nvidia’s A100 and H100 GPUs dominate this $300 billion infrastructure market due to CUDA’s software moat. They analyze the bottleneck in inference scalability, arguing that while training is a small fraction of the market, inference costs will drop as incumbents like Google, Amazon, and Apple leverage their own silicon, leaving little room for new AI accelerator startups. The hosts conclude that generative AI represents a powerful new category of productivity software rather than a fundamental shift in human-computer interaction, driving data center Capex through natural upgrade cycles rather than complete rebuilds.

    Chips, FPGAs & semiconductorsConnected products & IoT
    8 chapters

  • The Circuit · 41m

    Episode 105: Sanity Check on Custom ASICS?

    Ben Bajarin and Jay Goldberg discuss the debate between custom ASICs and GPUs in AI compute, arguing that GPUs retain significant value due to software flexibility and installed base. They analyze the uneven maturity of hyperscaler silicon efforts, noting Google's TPU success versus the struggles of Amazon's Trainium and Meta's MTIA. The conversation then shifts to analog semiconductors, highlighting their critical role in humanoid robotics and data center power management, with specific mentions of Infineon's inventory buildup linked to NVIDIA's Blackwell.

    Chips, FPGAs & semiconductorsPricing, margins & unit economics
    6 chapters

  • The Circuit · 41m

    Episode 73: Nvidia's Moment, and The "Platformization" of GPUs

    Ben Baharin and Jay Goldberg discuss Nvidia briefly surpassing Apple as the world's most valuable company, analyzing the sustainability of its growth and the 'platformization' of GPUs. They debate whether Nvidia's GPU-centric strategy aligns with hyperscaler preferences or if it serves the emerging market of 'AI factories' and tier-two cloud providers better. The hosts also explore the challenges for AI startups competing against hyperscalers, the reality of multicloud strategies, and the future of inference workloads.

    Chips, FPGAs & semiconductorsGrowth, exits & acquisitions
    7 chapters

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by Built 2 Scale. The episode itself belongs to its publisher.