Envisioning an Energy Efficient Tech Future

OnTrack · 14 May 2024 · 54m · Episode 266

What this episode covers

Greg Bailey, President of GreenArrays, discusses the company's approach to designing energy-efficient custom silicon by building from first principles rather than using standard libraries or FPGAs. He explains how GreenArrays creates full-custom chips with asynchronous logic to minimize leakage and power consumption, contrasting this with the inefficiencies of modern multicore architectures and the high energy cost of clocks. Bailey critiques the industry's reliance on hyper-specialization and software reuse, arguing that true innovation requires understanding both hardware physics and application-level requirements to solve problems like IoT power waste and AI simulation efficiency.

Other shows on the same ground, closest first.

  • The Circuit · 40m

    Episode 28: A Whole Lot of Custom Going On

    Ben Bajarin and Jay Goldberg discuss the hidden growth of custom silicon, highlighting Amazon's aggressive push with Graviton, Inferentia, and Trainium chips to achieve performance parity and energy efficiency. They explore how hyperscalers like Google, Microsoft, and AWS are developing semi-custom networking chips (e.g., Nitro) to reduce latency for AI workloads, while noting that automotive OEMs in China are increasingly licensing RISC-V IP for critical systems like ADAS and sensor fusion. The hosts analyze why these custom efforts remain under-reported and how consumer electronics marketing strategies may eventually drive automakers to brand their own silicon.

  • 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.

  • The Circuit · 39m

    Episode 20: Apple's Silicon Strategy

    Ben Bajarin and Jay Goldberg discuss Apple's strategy of designing its own System-on-Chips (SoCs) to achieve performance-per-watt advantages and control over the product roadmap. They analyze Apple's intense leverage over suppliers like TSMC and Broadcom, noting how Apple funds manufacturing and extracts technical knowledge to drive efficiency. The hosts also explore the limits of Apple's silicon ambitions, particularly regarding the difficulty of designing modems and RF components, and speculate on the custom chips required for upcoming Apple Vision Pro glasses.

  • The Amp Hour · 1h 05m

    Rigorous engineering stuff may be out the window

    Dave Jones and Chris Gammell discuss the decline of rigorous engineering in favor of rapid prototyping using SparkFun QuickBoards and JST connectors, while Jones details a frustrating teardown of an EcoFlow battery due to a poor-quality JST connector. They explore the broader implications of using off-the-shelf modules versus custom PCBs, touching on the trade-offs between flexibility and cost. The conversation shifts to industry news, including Skywater Electronics' TinyTapeout program for open-source silicon, the critical role of ASML in chip manufacturing, and US export controls on high-performance chips like NVIDIA GPUs. Finally, they review a Maxim Integrated USB Power Delivery controller, noting the industry's shift toward simpler, dedicated PD chips.

  • The Amp Hour · 1h 07m

    Downward Sloping Trace

    Dave Jones and Chris Gammell discuss the 'design deja vu' of independently creating nearly identical 1996 logic analyzer projects using Lattice Semiconductor CPLDs, and the technical nuances of timing vs. state analysis. They then detail the thermal design challenges of Jones's 'Micro Supply' project, explaining the decision to use a custom die-cast heat sink and plastic enclosure to achieve high volumetric efficiency. The conversation concludes with a critique of KiCad's via placement workflow and a recommendation for the book 'The Practicing Mind'.

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