Hi, my name is Saandiya KPS Mohan
I’m a Computer Engineering student passionate about
AI hardware & fintech systems.

Developed the Tensor Compute Accelerator (TCA) featuring both GEMM and convolution controllers. Integrated with a weight-stationary systolic array and 32-bank scratchpad memory, implementing im2col-based convolution and optimized data flow for high efficiency. Designed, simulated, and verified modules with RTL, testbenches, and performance evaluation.

Designed and implemented a single-precision IEEE-754 FPU for RISC-V, including adder, subtractor, and multiplier modules. Contributed to SystemVerilog RTL design, testbench development, and verification using QuestaSim to ensure correctness, precision, and performance.

Developed a reaction-time Whack-A-Mole game using an STM32F091 microcontroller, 32 x 32 LED matrix, and keypad. Implemented multiple difficulty levels, scoring system, and persistent high-score storage using I2C EEPROM. Designed modular firmware in C to control game logic, display, and input handling.

Designed and implemented a multicore RISC-V processor with caches to run single- and dual-threaded programs. Built incrementally from single-cycle and pipelined designs, adding cache hierarchy, shared memory bus, and thread-level parallelism. Evaluated throughput, instruction latency, and FPGA resource utilization, demonstrating improved scalability and performance.

Designed and tested a USB Full-Speed Bulk-Transfer Endpoint as an AHB-Lite SoC module. Implemented 16-entry, 16-bit data-width circular buffers for OUT/IN transactions with full/empty flags and read/write enable logic. Developed a testbench to simulate token transactions and verify data correctness, ensuring reliable USB protocol handling.

Researched and developed an AI-driven NPC simulation in Unreal Engine. Designed NPCs capable of recognizing and interacting with everyday objects through the Smart Objects framework, creating realistic, context-aware behaviors. Built high-fidelity simulations of the Purdue University campus to support architectural safety and accessibility research. Findings on sensory stimuli in gaming platforms were presented at the Spring Undergraduate Research Conference 2024.