Verification Engineer

Annapurna Labs Ltd. is an Israel‑based semiconductor design and systems‑engineering organization that operates as a subsidiary of Amazon and focuses on developing custom silicon and hardware systems for cloud infrastructure. Originally founded in Israel in 2011 as a privately held fabless semiconductor company, Annapurna Labs was acquired by Amazon in 2015 and subsequently became an engineering center within Amazon Web Services (AWS). The company’s stated role after acquisition has been to design and deliver purpose‑built processors, accelerators and system components that improve performance, security, cost efficiency and operational control for large‑scale cloud services. Company overview and scope of activity Annapurna Labs is organized around developing SoCs (systems on chip), platform controllers, and hardware offload engines specifically tailored for cloud datacenter workloads. Working within AWS, the Annapurna team concentrates on vertical integration of silicon, firmware, and accompanying system components so that Amazon can ship infrastructure products (server instances and storage/network accelerators) with improved price/performance, tighter security boundaries, and richer feature sets. The organization combines expertise in chip design, board‑level engineering, firmware, device drivers and systems software to move functionality out of general‑purpose CPUs and into dedicated hardware where this provides operational or efficiency advantages for large cloud fleets. Core business activities • Custom server CPU design: Annapurna Labs is the engineering origin of AWS’s internally developed Arm‑based CPU families. The team designs custom Arm‑architecture processors intended primarily for use in EC2 instances and other AWS services, optimizing core microarchitecture, memory subsystem, power management and IO integration for cloud workloads. • Hardware acceleration and offload: The group designs dedicated hardware components that offload networking, storage and virtualization tasks from host CPUs. These components enable higher throughput and lower latency for network and storage services while reducing host CPU utilization. • Secure virtualization and trusted compute: Annapurna engineers have developed secure, lightweight hypervisor and hardware isolation technologies that provide stronger separation between customer virtual machines and control plane functions, contributing to AWS offerings that emphasize tenant isolation and data protection. • System integration and productionization at scale: Beyond silicon, Annapurna Labs works on integrating chips into production server platforms, developing board and enclosure designs, creating firmware and driver stacks, and supporting hardware lifecycle and fleet operations for hyperscale deployments. Primary products and notable contributions After integration into AWS, the Annapurna team played central roles in several high‑visibility AWS hardware innovations: the AWS Nitro System and the AWS Graviton series of processors. The Nitro System is a collection of dedicated hardware and lightweight hypervisor components that offload virtualization, networking, and storage functions from guest instances and provide a foundation for the EC2 virtualization model used across many instance families. Nitro includes custom cards and controllers that implement acceleration and I/O functionality outside of the host CPU. The Graviton family—Arm‑based processors developed within AWS—owes its engineering origins to the Annapurna organization. These processors were introduced to provide cost‑effective, energy‑efficient compute for general purpose and specialized workloads in EC2. Graviton‑based instances have been used to deliver improved price/performance in many cloud workloads and have expanded AWS’s ability to tune CPU, memory and interconnect characteristics to specific instance classes. Operational role within Amazon and AWS As an AWS engineering center, Annapurna Labs contributes silicon and hardware platforms that are directly integrated into AWS products (EC2 instance types and their underlying server designs, storage and networking subsystems). The team’s engineering work is intended to give AWS greater control over the performance characteristics, security posture and total cost of ownership of its infrastructure while enabling feature innovations that are difficult to achieve using only off‑the‑shelf components. The group works closely with AWS product teams (compute, storage, networking and security) to align hardware capabilities to service requirements and customer needs. Market positioning and impact Annapurna’s integration into AWS represents a broader industry trend in which cloud providers invest in custom silicon and hardware systems to optimize for scale and to differentiate service offerings. By producing its own CPUs and offload devices, AWS (through Annapurna) can introduce instance types and acceleration features that emphasize energy efficiency, predictable performance, and enhanced isolation—factors important to hyperscale cloud economics and to enterprise customers seeking improved performance-per-dollar and security assurances. Employment, research and engineering culture Annapurna Labs has operated primarily as an R&D and engineering organization with expertise in chip microarchitecture, digital design, firmware, systems software and large‑scale production engineering. Located in Israel, the organization draws on talent from the country’s strong semiconductor, systems and software engineering communities and contributes to Amazon’s broader global hardware and systems engineering capabilities. Public presence and verification Public statements and product releases from Amazon and AWS identify Annapurna Labs as the internal engineering organization responsible for key AWS silicon and Nitro hardware capabilities. While Annapurna Labs operates as a part of Amazon’s private corporate structure, its engineering output is manifested publicly through AWS product announcements, instance families, and technical blogs describing the Nitro System and Graviton processors. These public materials document the company’s role as a silicon and systems engineering organization focused on cloud infrastructure hardware and the technical contributions it has delivered to AWS.