Reddit deep-dive: next-gen AI via memristors, spiking nets, and 1% sparse routing
Robert__Sinclair · reddit · 2026-10-24
A Reddit essay sketches a next-generation AI built on four pillars: compute-in-memory with memristors to kill the Von Neumann bottleneck, event-driven spiking neural networks (Loihi-style), extreme sparsity with only 1-2% of parameters active per token, and bio-inspired local learning rules like STDP for continuous adaptation. The vision: watt-scale on-device AI instead of megawatt data centers. Personal speculation, not a paper.
More from Infra
- Jhaddix: learn to host and tune your own local AI models before it's banned — AIFlow_ML · 2026-10-24
- DeepSeek's 80% gross margin secret is its org structure, not just tech — zephyr_z9 · 2026-10-24
- Kong gateways now carry 500T tokens, growing 140x YoY toward a quadrillion — AccBalanced · 2026-10-24
- AI credit risk hits new highs as Oracle 5Y CDS spikes to ~300 — AccBalanced · 2026-10-24
- Nebius is hoarding GPUs to sell at scarcity prices — funded by a $3.5B 364-day bridge loan — tengyanAI · 2026-10-24
- Spark-X2.6-4B GGUF quantized model trends on Hugging Face under Apache 2.0 — XHToken · 2026-10-24