Responsibilities & Opportunities

• Develop simulation features to enable rapid architectural exploration and performance evaluation
• Run simulation models for new products and analyze results to characterize the product and workloads
• Investigate cutting-edge, high-performance server CPU core and SoC architecture design, providing vital data support for crucial decision-making processes
• Develop strategies for software/hardware co-optimization features and lead the integration of software and hardware components for the next-generation processor
• Analyze and extract the distinctive features of real-world scenario workloads, delivering essential insights to the internal SoC Team
• Develop performance models for upcoming server CPUs and perform architectural studies and performance projections for applications of interest like databases, distributed storage, and engines for large-scale data analytics
• Work closely with System Architects and System Software Engineers to evaluate and understand performance characteristics
• Build, integrate, enhance, and document simulator components with new HW features
• Bring full SW stack up on DGX Simulator
• Improve performance, fix bugs across user and kernel stack, and automate execution flow

Required Qualifications

• Ph.D. in Computer Science, Computer Engineering, or related field
• 5+ years of experience in virtualization and HW simulation/emulation technologies
• Profound background in workload modeling and CPU architecture
• Strong understanding of system architecture, hardware design, hypervisors, and HW emulators (e.g., Qemu, KVM, VDK, SIMICs)
• Experience in developing and using performance simulators (e.g., GEM5 O3 model, Sniper) and dynamic binary instrumentation infrastructures (e.g., QEMU, DynamoRIO, x86 PIN)
• Experience in performance projection and architectural exploration using SoC simulators
• Experience in system modeling, GPU software development, parallel programming, computer architecture, machine learning, and software engineering
• Analytical and performance analysis skills
• Excellent collaboration and interpersonal skills
• Fluent in both verbal and written English

Preferred Qualifications

• Experience in HW and SW stack bring-up using simulators, emulators, etc.
• Experience in hardware interfaces (e.g., PCIe, SPI, I3C) with Linux boot solutions on x86 and ARM class platforms
• Experience in simulator development for NPU/GPU
• Experience in scalable end-to-end simulations
• Knowledge of NPU/GPU architecture

About us

MangoBoost, Inc.

• MangoBoost is a start-up company that develops groundbreaking new data processing units (DPUs). MangoBoost’s novel DPUs can be customized and flexibly integrated into server systems to address myriad system overheads and make other devices (such as SSDs, GPUs, NICs, etc.) smarter. As a result, MangoBoost DPU offers a dramatic boost in system efficiency, performance, scalability, and cost-effectiveness. MangoBoost leverages ~10 years of prior world-leading research done in academia, which has produced extensive full-system patent-pending prototypes published at premier conferences in computer architecture and systems areas (for example ISCA and OSDI). MangoBoost has received a $10M seed fund in 2022 and a $55M Series A fund in 2023 and is actively growing in the US and Korea.
• MangoBoost is comprised of the world's top computer system developers. The founder is currently one of the world's leading computer system architects, and other members are also experts with outstanding careers in various aspects of computer system development. Most team members have conducted research and worked at top-tier universities and companies worldwide. Since 2014, our team has been developing the world's top-level DPU, and we are the only start-up capable of designing a DPU that can compete immediately with giants like Amazon and NVIDIA.