Ashwin Kulkarni Email & Phone Number

Cupertino, California, United States

GPU Performance Modeling

Austin, Texas, United States

SEG-GPU Silicon Validation team- Designed a standard flow to be used for validation of a GPU spec in all Apple SoCs- Automated filtering of workloads that can be run on every manufactured SoC to test its functionality - Carried out robust testing of this flow to ensure its stability on current and future generations of GPUs

Santa Clara, California, United States

DRAM Fault Injection and automating regression flows- Automated selection of fault injection nodes from a design of 1 million nodes to achieve more coverage with less simulations.- Performed fault injection in around 20,000 nodes and debugged RTL bugs in the DRAM block of a computer vision SoC.- Wrote python scripts to automate this entire flow which reduced debugging time from 30 days to 6 days.

Atlanta, Georgia, United States

ECE8803 - Cybersecurity of Drones taught by Prof. Saman Zonouz- Since this was a new course, I along with 3 TAs were given the task of creating the course assignments based on the course content- Used Ardupilot software for creating assignments that covered topics including GPS Spoofing attacks, PID Controller attacks, Logic Bomb attacks on an x86 binary- Conducted office hours every week for solving doubts in assignments

Atlanta, Georgia, United States

"Distributed GNN Accelerator" project under Prof. Callie Hao- Created a simulator to calculate network latency of communication between FPGAs- Automated the process of running network tests for many-many communications between 8 remote FPGAsby writing Python, Shell scripts- Exported the time taken and retry counts of network tests for varying Bandwidths and Data- Integrated the simulator into a Distributed GNN (Graph Neural Network) system for Hardware Accelerators

Mumbai, Maharashtra, India

Backend Software Development Engineer- Part of the "NeuPass" team developing a new super-app called "TataNeu"- Designed and developed a REST microservice for error ticket generation and resolving network errors- Reduced response time to 150ms (approx. 200% speed-up), and wrote unit test cases with upto 92% coverage- Built a second microservice in Kotlin, acting as a middleware to filter out invalid customer transactions- Processed a pool of approx. 100K transactions per day to filter out invalid transactions

Pune, Maharashtra, India

Designed and maintained COEP Boat Club's official website

Pune, Maharashtra, India

Backend Software Engineering- Developed a REST API using Python Flask for clustering of financial documents- Reduced average clustering time from 35 to 27 minutes (23%) by calculating optimum number of clusters- Implemented K-Means, DBSCAN, Agglomerative, Birch clustering algorithms using scikit-learn framework

Master Of Science - Ms, Electrical And Computer Engineering, 3.9 / 4.0

Specialization: Computer Architecture

Bachelor Of Technology - Btech, Electrical Engineering, 9.24 / 10

Activities and Societies: COEP Boat ClubRank:- 2 / 87

Minor Degree, Computer Engineering, 9.5 / 10

12Th Std., Science, Percentage:- 83.84%