Sahil Parmar

Software Engineer

Newark, California, United States13 yrs 1 mo experience

Highly Stable

Key Highlights

Over a decade of experience in software engineering.
Expert in compiler and runtime software stack innovations.
Passionate about AI and high-performance computing.

Stackforce AI infers this person is a highly skilled software engineer specializing in AI and high-performance computing.

Contact

Skills

Core Skills

CompilersLarge Language Models (llm)

Other Skills

AlgorithmsCC++CUDAComputer ArchitectureData StructuresDebuggingDevice DriversDomain Specific ArchitectureGNU DebuggerGPUJavaLinuxLinux KernelOpenGL

About

Hi there! I'm Sahil, a passionate software engineer with a knack for solving complex problems at the intersection of hardware and AI. I pioneer the development of compiler and runtime software stack innovations at Groq, driving optimizations for AI and large language models on our cutting-edge hardware. My work involves designing scalable systems, implementing fault-tolerant solutions, and creating high-performance tools to unlock the full potential of AI workloads. With over a decade of experience at companies like Groq and Nvidia, I’ve contributed to compilers, distributed systems, and graphics shaders. My research has been recognized at top conferences, and I’m passionate about building efficient, reliable systems that push the boundaries of machine learning. Let’s connect and see how we can shape the future of AI and high-performance computing!

Experience

Groq

Senior Staff Software Engineer

Aug 2019 – Present · 6 yrs 7 mos · United States · Hybrid

Large Language Models (LLM)Domain Specific ArchitectureCompilers

Nvidia

Senior System Software Engineer

Feb 2015 – Aug 2019 · 4 yrs 6 mos · United States · On-site

Worked in GLSL/Vulkan compiler team.

Compilers

Amazon

Software Development Intern

May 2014 – Aug 2014 · 3 mos · Seattle, Washington, United States · On-site

Worked in Quality Tools Team, to develop Fleet Owner Tools for monitoring and maintaining host machines, helping owners understand top users on their fleet and whether their resources are effectively used.

Ca technologies

Graduate Research Assistant

Oct 2013 – Jan 2014 · 3 mos · Stony Brook, New York, United States · On-site

Development on integration of Workload Automation with existing hadoop based solutions.

Nvidia

System Software Engineer

Jul 2011 – Jul 2013 · 2 yrs · Pune, Maharashtra, India · On-site

Worked in CUDA/PTX Compiler team.

Compilers

Clogeny

Software Engineer Intern

Jul 2010 – Jun 2011 · 11 mos · Pune Area, India · Hybrid

Worked on my final year bachelor's project 'aDICIOn', to develop a Self-learning Prototype for private cloud that exploits adaptive allocation of memory on a virtual machine to gracefully scale in the presence of varying workload.