About
Toronto AI inference chip startup, now acquired by AMD. Taalas hardwires specific LLMs directly into SRAM-based silicon using structured ASIC technology, targeting inference teams priced out of GPU clusters. Its first chip, optimized for Llama 3.1 8B, hits over 16,000 tokens per second and claims 50x better energy efficiency than GPU equivalents.
Activity
2026
1 story so far in 2026
| Month | Stories |
|---|---|
| January | 0 |
| February | 0 |
| March | 0 |
| April | 0 |
| May | 0 |
| June | 0 |
| July | 0 |
| August | 1 |
| September | 0 |
| October | 0 |
| November | 0 |
Focus areas
1 story
