Products
Taalas HC1 Technology Demonstrator
- Runs Llama 3.1 8B model
- TSMC 6nm | 815mm2 | 53B Transistor
- 2.5 kW Server
Instantaneous Inference
HC1 demonstrates the power of Taalas hardcore model silicon technology, delivering over 17k tokens per second per user on Llama 3.1 8B model.
Source: Model Llama 3.1 8B, Nvidia Baseline (H200), B200 measured by Taalas | Groq, Sambanova, Cerebras performance from Artificial Analysis | Taalas Performance run by Taalas labs | Input sequence length 1k/1k