PCIe Validation Engineer, SPIV team

Annapurna Labs designs the custom silicon and systems at the heart of AWS (Amazon Web Services). Here you'll move like a startup, build at the scale of the world's largest cloud, and push the limits of what hardware can do.

Annapurna Labs, as part of Amazon, is looking for a PCIe Validation Engineer.

As a PCIe Validation Engineer, you’ll validate PCIe interfaces of the newest generations on our new products across the full product lifecycle — from early silicon bring-up through manufacturing test integration. You’ll join a validation team responsible for PCIe link quality across all AWS server platforms, ensuring our custom Hosts and smart NICs perform flawlessly before reaching millions of customers.

In this role you will define and execute validation plans for PCIe interfaces across multiple ASIC generations and server platforms. You will verify silicon, firmware, and retimer functionality against design specifications, and drive complex system-level debugs spanning hardware, firmware, and software. You will work with PCIe IP vendors, retimer and switch vendors, and platform integration teams to ensure interoperability and link quality. You will set high standards for yourself and your work, continuously looking for ways to improve the quality and reliability of our products. We are transforming an industry, and we are looking for individuals who are ready for this challenge and want to reach beyond what is possible today.

Key job responsibilities

• Define and execute validation plans for PCIe interfaces on new ASIC generations, covering card-level and server-level link validation across multiple server platforms
• Validate interoperability with third-party PCIe components including, PCIe switches, and PHY IP from silicon vendors,
• Support manufacturing test integration — define and refine PCIe test criteria for contract manufacturers, analyze yield data, and drive feedback loops to improve product quality
• Investigate and root-cause PCIe link failures, error patterns, and performance anomalies at the system level — spanning NIC firmware, host BIOS/BMC, and kernel drivers
• Support RMA analysis — reproduce field failures on returned hardware, classify PCIe-related root causes, and feed findings back to design and firmware teams

About the team
Our SPIV team owns the performance and reliability of PCIe and Ethernet across every AWS server. We work end-to-end across the full product lifecycle, including:

• Pre-silicon IP vendor engagement
• post-silicon bring-up and interface optimization
• SerDes, firmware, and system-level debug
• Interface manufacturing test integration
• Fleet health monitoring and customer performance
• Interoperability with third-party components - retimer, switch, active cable, and optics

In partnership with chip design, Server design and Cloud health teams, you will validate new server platform generations; root-cause failures across hardware, firmware, and software boundaries and drive quality improvements that protect millions of customer workloads.
If you are a systems, hardware, or software engineer who thrives at the intersection of silicon, firmware, and server infrastructure — and wants your work to improve reliability across a global fleet — apply now.