Lattice Semiconductor Sees AMI Combination Powering $3B AI Data Center Ambition

Lattice Semiconductor NASDAQ: LSCC CEO Ford Tamer and AMI Business Unit leader Sanjoy Maity outlined how the companies’ combination is intended to address growing management, control and security demands across AI data centers and physical AI systems.

Speaking at The Six Five Summit 2026, Tamer described Lattice as a provider of low-power programmable devices used alongside processors, networking equipment and sensors. The company’s small and mid-range FPGAs support functions including booting, power sequencing, security management, control, bridging, I/O expansion, sensor fusion and data processing, he said.

AMI, which Maity said has developed foundational firmware for four decades, provides firmware used to boot, manage and secure AI data center infrastructure. Its offerings also include telemetry, monitoring, error correction and a “single pane of glass” view intended to help administrators oversee complex data center environments.

Tamer said the companies have known each other for six years and serve similar customers, including hyperscalers, neocloud providers, server and networking OEMs, and silicon and systems partners. He characterized the businesses as complementary because Lattice operates close to hardware and sensors, while AMI provides low-level firmware for managing and securing those components.

“Together, we believe we’re going to provide secure management control solution, a turnkey, and speed up our customers’ time to market,” Tamer said. Maity said the combined organization aims to provide pre-validated solutions so customers can spend less time addressing integration issues and focus more on their own intellectual property and business development. He said the approach could reduce integration costs and accelerate product launches.

Tamer said AI infrastructure has undergone a rapid expansion in recent years, with 2025 centered on training workloads and 2026 seeing a greater shift toward inference and agentic AI. He said the shift has increased demand not only for GPUs but also for CPUs, storage and networking infrastructure. He also cited changes in networking architectures, liquid cooling and increased sensor deployment. According to Tamer, the number of sensors in a server has grown to as many as 1,200, monitoring functions ranging from fans and cooling distribution units to optical interconnects. That expansion has increased the use of low-power FPGAs from tens of devices per rack to hundreds of devices per rack, he said. Lattice’s devices are positioned near sensors and hardware components to provide deterministic, low-latency connectivity and parallel processing.