CPU
Central Processing Unit—the classic general-purpose computer chip (e.g., Intel) where only about 3% of the silicon activates per clock cycle, processing tasks largely sequentially.
As defined in The Man Who Built NVIDIA (with Stephen Witt)
Where it comes up · 9
- Arm: The Silicon Blueprint3 claims
- When Vectors Break Down: Graph-Based RAG for Dense Enterprise Knowledge - Sam Julien, Writer2 claims
- Databricks: From Data to Decisions1 claim
- AMD: How Chips Are Changing1 claim
- Early Stage Investing1 claim
- Full Tilt Investing1 claim
- The Man Who Built NVIDIA (with Stephen Witt)1 claim
- Dylan Patel - The Infinite Demand for Tokens, Claude Mythos, and Supply Constraints - [Invest Like the Best, EP.469]1 claim
and 1 more
Most important claims · 10 of 12Search all claims for “CPU” →
- GPUs use parallel/accelerated computing: on a classic Intel CPU only ~3% of the silicon activates per clock cycle, whereas on an NVIDIA chip 30-40% lights up, performing far more calculations per tick—harder to program but much faster and denser when done right.The Man Who Built NVIDIA (with Stephen Witt)
- Compute hardware has swung from general-purpose CPUs to GPUs to specialized ASICs and back toward specialized hardware, as Bitcoin mining acted as a petri dish for extreme demand for efficient compute that now resembles machine learning demand.Early Stage Investing
- Even in GPU-centric AI systems with eight or twelve GPU cards linked together, some number of CPUs is still required, and those CPUs are often ARM-based, so ARM benefits from the AI buildout without being in every accelerator.Arm: The Silicon Blueprint
- NVIDIA prefers to sell its own CPU, the Grace Hopper chip, which is ARM-based, showing that even the leading GPU vendor builds its systems on Arm architecture.Arm: The Silicon Blueprint
- Demand will shift from CPU-based data centers to GPU-based computing for crypto mining, CGI rendering, and AI, and the US is drastically behind other countries in building this capacity.Full Tilt Investing
- Bitcoin mining moved from CPUs to GPUs to ASICs, and once a specialized chip existed GPUs became unprofitable for mining; the same pattern will hold for transformers, where specialized ASICs will displace GPUs for inference.Gavin Uberti - Real-Time AI & The Future of AI Hardware - [Invest Like the Best, EP.356]
- Small, fast language models suitable for fine-tuning on CPU/small GPU inference are now readily available in the market, so teams building graph RAG systems today might not need to build specialized models like Writer did.When Vectors Break Down: Graph-Based RAG for Dense Enterprise Knowledge - Sam Julien, Writer
- CPUs are completely sold out because reinforcement learning environments run on CPUs and because AI-generated apps are deployed on cloud CPU instances, so CPU demand is skyrocketing alongside GPU demand.Dylan Patel - The Infinite Demand for Tokens, Claude Mythos, and Supply Constraints - [Invest Like the Best, EP.469]
- Databricks' model is fairly capital-light because most core data processing workloads run on CPUs rather than GPUs, so it avoids the GPU acquisition burden faced by AI-native companies, though model serving adds GPU costs of a much smaller order of magnitude.Databricks: From Data to Decisions
- Instead of using traditional graph databases for scaling, Writer built a specialized model (running on CPUs or smaller GPUs) to convert unstructured data into graph structures of nodes and edges, combined with context-aware chunking that preserves semantic relationships.When Vectors Break Down: Graph-Based RAG for Dense Enterprise Knowledge - Sam Julien, Writer