About the job
Oracle hardware platform development engineering is seeking a highly driven GPU/CPU Platform System Engineer at the Principal Engineer level. The GPU System Engineer will work within development engineering with a small team of talented engineers who lead the development and day-to-day engineering efforts for Oracle’s rapidly growing and successful Cloud AI platforms. You will participate in platform definition, platform development oversight as well as in house development, design reviews, system integration, performance testing and characterization. You will interact closely with third party GPU IC suppliers & partners as well as internal hardware and software development teams to help drive Oracle’s AI Cloud platform solution space. You will be a critical part of the team developing Oracle’s growing Cloud AI solutions. The team you will be joining has delivered the first and second generation of Oracle Cloud dedicated compute, AI platforms, and is working to build out the next generation of Cloud and Enterprise systems, with record breaking-performance, security, and world class quality, using the latest and greatest merchant silicon and technologies.
Responsibilities
Review and assessment of third-party merchant silicon used for AI Accelerator Modules.
Evaluation of system architecture and proposed implementation path analysis. You will participate in platform definition and analysis.
Provide platform development oversight for partners.
Work with in-house engineering functional experts on design and reviews.
Support and guide system integration, performance testing and characterization.
Support development program managers on technical assessments & planning.
Interact closely with third party GPU IC suppliers & partners as well as internal hardware, software development, quality assurance, cloud orchestration, hardware and software security experts, and Oracle manufacturing teams.
Document and specify design intent and design details where appropriate in collaboration with the appropriate engineering teams.
Participate in hardware platform security evaluations.
Guide partner internal Oracle teams on support needed to scale, monitor, and successfully deploy our products to the Cloud.
Assist Oracle Cloud and Support teams in the root-cause of potential hardware or software bugs through firsthand lab replication debug, remote debug, and calls with the appropriate teams supporting our deployed products.
Work with Oracle manufacturing teams to ensure that Oracle hardware is secure, robustly evaluated, performing at peak capabilities and well qualified for deployment to our Cloud customers.
Qualifications
Minimum
Technical hands-on experience with market leading GPU (or alternate AI platforms) from the hardware and platform development, test, and characterization perspectives.
Balance hardware performance priorities against power, cost, and cross-functional considerations. Be responsible for meeting hardware product performance and regulatory specifications, if applicable.
Solid knowledge of AI / GPU or/and AI/CPU platform architecture and their capabilities.
A strong understanding and experience running firmware and system diagnostics tools using BMC firmware, UEFI/ BIOS and Linux tools. Skilled in scripting to customize tests.
Solid working experience with GPU supplier test code as well as open-source AI test / characterization tools.
Experience with the architecture, design, and implementation of modern server platforms consisting of multiple architectures and vendors, including x86 and ARM server architectures.
Experience with hardware development at the system, board, and FPGA level.
Required experience with board level tools and ability to reviews hierarchical schematics, multilayer advance board layout, cross board interconnect and end-to-end connectivity analysis.
Strong communications skills and ability to clearly communicate complex technical issue across engineering disciplines as well as clearly and succinctly articulate issues for executives.
Demonstrated experience debugging and root-causing complex issues that may have a mix of hardware and software causes.
Experience with early stage bring-up and power-on, platform firmware debugging, prototype GPU & CPU complex and memory complex debugging.
An ability to isolate a problem to the source and the required creativity & expertise to devise timely and robust solutions.
Experience and understanding of the latest high-speed busses and interconnect used in modern Compute and AI platforms. Familiarity with their startup connectivity and operational robustness.
Preferred
10 years of experience with hardware, design and bring-up.
Comfortable with the use of hardware debuggers.
Experience in PCIe, DDR, Ethernet, USB, SPI, etc.
Experience with platform level security technologies present an advantage in the role.
Experience with power circuit design and signal integrity.