Abhikrant Sharma
Staff Software Engineer Manager at Qualcomm
Hyderabad, Sindh, United States
Summary
🤩
Rockstar🎓
Top SchoolAbhikrant Sharma is a Staff Software Engineer Manager with a decade of experience driving software design, integration, validation and product commercialization, primarily on Qualcomm Snapdragon platforms. He leads compiler and AI-inference efforts to optimize deep learning models across ARM, Adreno GPU and Hexagon DSP, combining pre-silicon emulation work with silicon bring-up and post-launch commercialization. At Qualcomm he progressed from DSP audio and vision acceleration to managing next-gen edge computing and AI compiler teams, delivering production-grade optimizations for on-device inference. An active contributor to the TVM deep learning compiler, he has implemented and optimized MatMulInteger16 and mixed-precision CUDA strategies, reflecting hands-on ML compiler expertise. He holds a Master’s in Software Engineering from BITS Pilani and brings uncommon end-to-end product lifecycle fluency—from emulation platforms to customer-facing support. Based in Hyderabad, he blends deep embedded-systems know-how with practical open-source contributions that bridge research compilers and real-world SoC constraints.
4 years of coding experience
10 years of employment as a software developer
BITS Pilani, Birla Institute of Technology and ScienceBachelor Of Technology Electronics and Communication, Bachelor Of Technology Electronics and Communication at Visvesvaraya National Institute of Technology10th Grade ICSE Board, 10th Grade ICSE Board at Sacred Heart High School,Dharamshala12th Grade CBSE Board, 12th Grade CBSE Board at DAV School Narwana
Github Skills (10)
compiler10
machine-learning10
deeplearning-ai10
compiler-compiler10
deep-learning10
onnx10
cuda9
tensor9
hexagons8
python8
Programming languages (2)
Github contributions (5)
apache/tvm
Nov 2021 - Jan 2023
Open deep learning compiler stack for cpu, gpu and specialized accelerators
Role in this project:
ML Engineer
ML EngineerContributions:36 reviews, 8 commits, 24 PRs in 1 year 2 months
Contributions summary:Abhikrant's commits primarily focused on integrating and optimizing the MatMulInteger16 contrib op for ONNX within the TVM deep learning compiler. Their contributions included implementing the op, addressing formatting and indentation issues, and enhancing CUDA batchmatmul strategies. The user also added related test cases and modified existing code to handle mixed precision. Furthermore, the user has contributed to several other areas, including the implementation of reshape and batch flatten slice operations for hexagon, using VTCM.
metalvulkancompilertensoropencl
abhikran-quic/tvm
Oct 2021 - Jul 2024
Open deep learning compiler stack for cpu, gpu and specialized accelerators
Contributions:136 pushes, 42 branches in 2 years 9 months
cpugpu-programminggpu-accelerationtvmdeep-learning
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.