AI Chip Architectures - Jacob Peake
jepeake.com · 9,562 words · saved by 1 readers
A look at AI Chip Architectures. NVIDIA, AMD, TPUs, Trainium, Groq, Cerebras.
At the 2018 International Symposium on Computer Architecture, John Hennessy and David Patterson delivered their Turing Lecture: "A New Golden Age for Computer Architecture". In the 1980s, when Hennessy and Patterson did their Turing Award-winning research, single-threaded CPU performance grew 52% a year. By 2018, with the end of Moore's Law and Dennard Scaling, the rate was 3%. There was a need for domain-specific architectures (DSAs). Their worked example was Google's TPU v1, already in production: 29× the throughput of a CPU on neural-network inference, at 80× better energy efficiency.…
saved by
related reading
- AI Chip Architecturesjacobpeake.com
- Taalas is what Etched should have been.zach.be
- TPU Inference Externalization Full Steam Ahead - InferenceXnewsletter.semianalysis.com
- An Interview with MatX CEO Reiner Pope About LLM Chipschipstrat.com
- OpenAI Jalapeño: Better Than Nvidia Blackwellnewsletter.semianalysis.com
- The Inference Shift – Stratechery by Ben Thompsonstratechery.com
- The Short Case for Nvidia Stock | YouTube Transcript Optimizeryoutubetranscriptoptimizer.com
- TPU Deep Divehenryhmko.github.io
- The Best GPUs for Deep Learning in 2023 — An In-depth Analysistimdettmers.com
- How To Scale Your Modeljax-ml.github.io
- Touching the Elephant - TPUs | Consider the Bulldogconsiderthebulldog.com
- Reiner Pope of MatX on accelerating AI with transformer-optimized chipscheekypint.substack.com