Posts by Thiago Crepaldi Productionizing TurboQuant on AMD GPUs for KV-Cache-Bound LLM Inference 11 June 2026 Inesh Chakrabarti* , David Limpus* , Aditi Ghai Rana* , Bowen Bao , Spandan Tiwari , Thiago Crepaldi , Ashish Sirasao English Applications & models AI/ML LLM Performance Memory *Equal contributions. Read more ...