CRAM claims 100,000x+ speedup on Sudoku — 'I read the paper, it's Sudoku'
willdepue · x · 2026-11-17
BenHoov touts the CRAM architecture achieving a 100,000x–215,000x speedup over TRM on Sudoku, with a video showing how many independent CRAMs can train in the time one TRM takes. willdepue mocks the hype in greentext form: pressing on whether the 'massively more efficient architecture' is real, he finds the benchmark task is — just Sudoku, questioning whether such headline speedups rest on unfairly trivial tasks.
Related event: IBM's CRAM Energy Model Trains 1,000x Faster on Sudoku(3 posts)→
More from Research
- Jeremy Howard reimplements MIT's Recursive Language Models paper inside SolveIt, more concise than original — jeremyphoward · 2026-11-17
- SperidLabs' Iris-Z: One Unified Generative Model Beats Specialist Depth Models at Scale — speridlabs · 2026-11-17
- Vibe Checking the IRS: Auto-Formalizing the Tax Code in Lean to Catch Errors — SnooPeripherals5313 · 2026-11-17
- MIT distills Monet, Van Gogh and 14 other artists' entire oeuvres into single paintings at ECCV 2026 — alexbilz · 2026-11-17
- Topos Bio teams with NVIDIA cuPotential to boost molecular dynamics throughput up to 3.6x — AllThingsApx · 2026-11-17
- Infrathon Benchmark Tests LLM Agents on Real-World AI Infra Optimization; Claude Opus 5.5 Leads — RulinShao · 2026-11-17