Intel and Altera are teaming up to build a bridge that carries sensor data from robots, medical machines, and defence systems into an Intel processor.
The aim is to help equipment makers connect and process that data. As machines gain more sensors, the data has to reach the host Intel chip’s memory quickly enough for the machine to act on it.
In an Intel newsroom release, the companies say they will combine Intel compute, connectivity, and software with Altera FPGAs (chips that stay reprogrammable after they’re built). Some applications need very high bandwidth or put input/output away from the compute board.
It’s a plan, not a product. Intel and Altera haven’t given a launch date, a price, or a product name, and they haven’t named any customers. What they describe is a set of modular reference solutions that equipment makers can build on, moving faster from prototype to deployment.
As sensors multiply at the edge and through Physical AI systems, our customers need a simpler way to turn real-world data into timely action.
Dan Rodriguez, corporate vice president and general manager, Edge Computing Group, Intel
The release names three markets. For healthcare, the companies are jointly developing a reference solution for the strict latency and real-time demands of advanced equipment. In aerospace and defence, they’re collaborating on reference solutions for next-generation mission systems, focusing on size, weight, and power for converged AI and signal-processing workloads.
For robotics, Intel and Altera plan to pair Intel compute, connectivity, and robotics software with Altera’s FPGA technology. The target is the sensing, timing, and real-time processing that intelligent machines need.
Developers will get access through Intel’s Open Edge Platform SDK and an Altera SDK, though neither company has said when.
Altera calls itself the largest pure-play FPGA provider. The Robot Report says Intel bought Altera for $16.7 billion in 2015 and later sold a 51% stake to Silver Lake.
Altera’s Farhad Shafai, head of business solutions and marketing, says the company’s FPGAs “stay programmable as sensors and AI models evolve.”
