Kraken: Inherently Parallel Transformers For Efficient Multi-Device Inference
–Neural Information Processing Systems
Parallelism introduces collective communication that is both expensive and represents a phase when hardware resources are underuti-lized.
Neural Information Processing Systems
Oct-9-2025, 18:38:24 GMT
- Country:
- Asia > Middle East
- Jordan (0.04)
- Europe
- France (0.04)
- Italy > Calabria
- Catanzaro Province > Catanzaro (0.04)
- North America
- Canada > Ontario
- Toronto (0.04)
- United States > California
- Santa Clara County > Santa Clara (0.04)
- Canada > Ontario
- Asia > Middle East
- Genre:
- Research Report > Experimental Study (0.93)
- Industry:
- Government (0.68)
- Technology: