Why “ternary” is being discussed
The most recent Hacker News stories and comments contributing to this topic's mentions.
For now. One more ternary model type breakthrough, or MoE, engram thing (I don't fully understand those for the record, I just know they speed things up and use less VRAM) could…
by luxpir · Sep 21, 2026
Because everything is converging on a backlog of huge efficiency gains established in research, waiting to be combined. Looped transformers, a whole host of diffusion techniques…
by blurbleblurble · Sep 20, 2026
Line-Rate Post-Quantum Byzantine Consensus via L1D Balanced Ternary
by leadpiperl1 · Sep 20, 2026 · 2 points
Run a 27B reasoning model locally on a 16GB M2 Mac. Ferrox + ternary quantization delivers 5.9GB models without sacrificing much quality.
by antonellof · Sep 19, 2026
I think the 1.76 includes that. Plain ternary packed into bytes would be around 1.58.
by smallerize · Sep 18, 2026
How does that compare with the bf16 version? For my "Please recite Jabberwocky" test the bf16 almost passes but the ternary and even fp8 versions fail badly.
by raylad · Sep 18, 2026
Interest
Proportion of Hacker News items mentioning "ternary" over time.
Mentions
Total number of Hacker News items mentioning "ternary" over time.