Dilip Puri

Firmware Engineer

Bengaluru, Karnataka, India5 yrs 10 mos experience

Highly StableAI ML Practitioner

Key Highlights

Expert in developing firmware for AI accelerators and GPUs.
Strong background in low-level system programming with C and C++.
Passionate about AI, HPC, and network firmware innovation.

Stackforce AI infers this person is a Firmware Engineer specializing in Embedded Systems and AI Accelerator technologies.

Contact

Skills

Core Skills

Firmware DevelopmentEmbedded SystemsGpu Software DevelopmentPerformance Optimization

Other Skills

Embedded SoftwareCC++AI AcceleratorsGPUsOpenCLLevel-ZeroGPU DriversGTestJSONMatlabResearchPythonGitNetworking

About

I am a Firmware Engineer at Intel, I have hands-on expertise in developing mission-critical firmware for AI accelerators and GPUs. My work includes designing and implementing schedulers, bare-metal OS, and efficient processing of hardware commands, while also optimizing descriptors for dedicated hardware systems. I’ve worked extensively with ARC/ARM processors, DMA and memory controllers, low-level register programming, as well as firmware for networking components like NICs. My core strength lies in low-level system programming, primarily using C and C++, with a focus on delivering optimized, high-performance solutions for complex hardware challenges. I’m passionate about pushing the boundaries of AI, HPC and Network firmware innovation, while building scalable and efficient embedded systems.

Experience

5 yrs 10 mos

Total Experience

5 yrs 10 mos

Average Tenure

5 yrs 10 mos

Current Experience

Intel corporation

3 roles

Firmware Engineer

Promoted

Feb 2022 – Present · 4 yrs 3 mos

Project: Jaguar Shores AI GPU (Current)
Currently developing event-based bare-metal firmware for Jaguar Shores AI GPU to control UALink Network Hardware (UAL Bridge, UAL Ctrl, SerDes)
Implementing firmware solutions for internal memory management and SerDes control on next-generation AI GPU architecture
Project: Gaudi AI Accelerator (Gen 2 & 3)
Developed end-to-end bare-metal firmware for Gaudi AI Accelerators (Gen 2 & 3) controlling runtime execution of MME, DMA, and TPC hardware blocks workloads
Implemented network-side command processing functionalities for AI Accelerators ensuring seamless hardware software integration
Added critical performance features and optimizations of hardware descriptors to firmware achieving highly performant and efficient system operation
Designed comprehensive testing framework using GTest with multiple test cases for firmware validation and reliability

Embedded SystemsEmbedded SoftwareCC++Firmware DevelopmentAI Accelerators+1

GPU Software Development Engineer

Jun 2020 – Jan 2022 · 1 yr 7 mos

Worked as GPU Software Development Engineer in AXG Super-Compute Linux Debug team
Debugged graphic applications in C/C++ and multiple issues related to OpenCL, Level-Zero and Compilers for PC GPU drivers contributing to performance optimizations and feature enhancements

CC++OpenCLLevel-ZeroGPU DriversGPU Software Development+1

Graduate Intern

Jun 2019 – May 2020 · 11 mos

Enhanced NNP-Training accelerator profiler tool with additional features
Implemented JSON-based configuration and conducted testing using GTest framework

GTestJSON