SiMa.ai Launches Palette 3.0 to Run More AI Models on One Chip
SiMa.ai has introduced Palette 3.0, a software update designed to run multiple artificial intelligence models simultaneously on a single chip, targeting one of the biggest challenges in physical AI: delivering advanced reasoning and real-time decision-making without exceeding the power and memory limits of drones, robots, and industrial machines.
Announced on November 17, 2026, the release builds on SiMa.ai's Agentic Palette platform introduced in June. It enables large language models (LLMs), vision-language models (VLMs), and other AI workloads to operate concurrently on the company's Modalix machine learning system-on-chip (MLSoC).
The company claims the update can increase model token-generation throughput by up to 3.5 times through speculative decoding, a technique that accelerates AI inference without requiring a corresponding increase in computing hardware.
The development comes as physical AI moves beyond recognizing objects and processing visual information toward systems capable of interpreting their surroundings, reasoning about situations, and taking actions autonomously.
For manufacturers...
Copyright of this story solely belongs to www.ciol.com. To see the full text click HERE