ARM-based CPU design for AI inference and cloud servers
Ampere designs custom ARM CPUs optimized for energy-efficient AI inference and cloud workloads. The stack reveals a company deep in silicon and systems validation: SystemVerilog + UVM for design verification, Cadence + Siemens for EDA, and emerging AI software layers (PyTorch, vLLM, SGLang, llama.cpp). Hiring velocity is accelerating with 65 roles posted in the last 30 days—heavily weighted toward principal engineers and interns—suggesting both deep technical hiring and workforce scaling around next-generation silicon bring-up.
Notable leadership hires: Director, Procurement & Strategic Sourcing
Ampere is a semiconductor design company founded in 2017, headquartered in Santa Clara, designing CPUs and related silicon IP for cloud infrastructure and edge AI. The company focuses on high-performance, power-efficient compute architectures built on ARM instruction sets. Current project work spans physical design (PCB/package guidelines), post-silicon validation and test yield optimization, power management circuits, and open-source JTAG/debug tooling for their custom ARMv9 SoC. Manufacturing and supply-chain hiring (6 roles + Director-level procurement position) indicates movement toward foundry partnerships and volume production. The engineering org (79 roles) dominates headcount, with targeted growth in data and finance—typical of a company scaling from design through manufacturing ramp.
Ampere designs 64-bit ARMv9 SoCs optimized for energy-efficient AI inference and cloud servers. Current projects include the Ampere One Aurora CPU and next-generation power management and interconnect IP.
Ampere's stack includes Cadence and Siemens for EDA, SystemVerilog and UVM for verification, PyTorch and vLLM for AI software modeling, and custom open-source JTAG and debug tooling for their ARMv9 architecture.
Other companies in the same industry, closest in size