Krishna Sankisa
Member Of Technical Staff at AMD
Hyderabad, Telangana, India
Summary
🤩
Rockstar🎓
Top SchoolKrishna Sankisa is a GPU compiler engineer with 8 years of experience building high-performance ML and graph compiler backends, currently a Member of Technical Staff at AMD. He specializes in MLIR, LLVM and AMDGPU backend work—contributing to the widely used llvm-project by refactoring LDS lowering and integrating address sanitizer instrumentation for memory accesses. Before AMD he developed neural network graph runtimes and scheduling for Blaize accelerators and implemented secure C++ TrustZone apps at Qualcomm, giving him a rare blend of compiler, runtime and systems security experience. Based in Hyderabad, he combines deep low-level compiler expertise with practical hardware-aware optimizations that bridge research-grade IRs and production device targets.
8 years of coding experience
7 years of employment as a software developer
BITS Pilani, Birla Institute of Technology and Science
English, Telugu
Github Skills (11)
asa10
llvm10
gpu-programming10
c-language10
amdgpu10
cprogramming-language10
addresssanitizer10
compiler-development10
management9
low-level9
memory-management9
Programming languages (4)
Github contributions (5)
llvm/llvm-project
Nov 2021 - Apr 2025
The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:77 reviews, 43 PRs, 29 pushes in 3 years 5 months
Contributions summary:Krishna primarily focused on improving the AMDGPU backend of the LLVM compiler project. Their work involved refactoring and moving utility methods related to local data store (LDS) operations within the AMDGPU target, specifically modifying the `AMDGPUAsanInstrumentation.h/.cpp` and `AMDGPUSwLowerLDS.cpp` files. The user also updated the `removeFnAttrFromReachable` function to accept an array of function attributes for removal and incorporated address sanitizer (asan) instrumentation for memory accesses within the lowered LDS code. Furthermore, they added and enabled a pass to lower LDS accesses to global memory and handle address space casts.
compilerstechnologiesclangsubmittoolchain
skc7/llvm-project
Sep 2023 - Mar 2025
The LLVM Project is a collection of modular and reusable compiler and toolchain technologies.
Contributions:223 pushes, 24 branches in 1 year 6 months
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.