Jul 2021 – Jan 2023 · 1 yr 6 mos

• Research Lead for ultra-low power graphics pipelines for future Meta AR and VR devices proposing novel pixel generation pipelines and augmentations into product teams for 2D and 3D rendering. Leading a silicon R&D team for research and advanced development to tackle relevant problems, including highly efficient 2D and 3D rendering for AR/VR in constrained and distributed form-factors

Oct 2019 – Jul 2021 · 1 yr 9 mos

• Graphics Architect at Facebook Reality Labs working on novel rendering pipelines for future AR/VR products. Exploring ultra-low-power graphics pipelines for 2D and 3D content

Mar 2018 – Oct 2019 · 1 yr 7 mos · San Jose

• Lead GPU Fixed-Function Architect, involved in planning, architecture, design, and execution of Samsung's in-house GPU IP. Owner for state management for new graphics APIs like Vulkan

Sep 2015 – Feb 2018 · 2 yrs 5 mos · Mountain View, CA

• As part of the Advanced Processor Lab's architecture team, defining the architecture for a low-power mobile GPU built from scratch. Research and development of novel technologies like variable-rate shading, tiling, hidden surface removal (~20% overdraw reduction), compression (3-4x footprint compression for RGBA8 textures, ~25% texture bandwidth reduction. >2x footprint and bandwidth reduction for vertex attributes), resource allocation (Up to 20% reduction in warp lifetimes with out-of-order register allocation)

Apr 2014 – Aug 2015 · 1 yr 4 mos · Mountain View, CA

• As part of the Advanced Processor Lab, contributing to R&D related to novel hardware and system design for mobile processors

Jun 2012 – Aug 2012 · 2 mos · Austin, Texas Metropolitan Area

• As part of the Heterogeneous System Architecture (HSA) Workloads group, fluid simulation on heterogeneous architectures including AMD APUs showing >2x performance uplift

May 2011 – Aug 2011 · 3 mos · Santa Clara, CA

• As part of the Parallel Computing Lab, optimized hybrid fluid simulation methods for parallel architectures (published at ACM SIGGRAPH Asia 2012), demonstrating improved scaling and performance uplift on high-throughput compute devices like Xeon Phi

May 2010 – Aug 2010 · 3 mos · Portland, Oregon Metropolitan Area

• As part of the 3D Graphics Technology group, performance optimization of fluid simulation using OpenCL 0.98-1.0 on Intel CPUs

May 2009 – Aug 2009 · 3 mos · Los Angeles Metropolitan Area

• Worked towards improving audio-visual fidelity and realism of virtual agents, and contributed to the testing architecture for artists to verify animations. Studied the applicability of various body gestures to speech portions depending on the state of the virtual agent.

Aug 2008 – Mar 2014 · 5 yrs 7 mos · Computer Science Department, SItterson Hall

• As part of the GAMMA group, developed novel flow simulation algorithms, enabling simulation of granular media like sand, crowds and stampedes, and efficient large-scale fluid simulations. Included efficient and scalable implementations for high-performance computing platforms like wide-vector CPUs (including Xeon Phi) and GPUs.
• PhD Thesis work and 10+ research publications in leading conferences and journals.