• R5SEBA – An N-way out-of-order RISC-V processor with speculative execution, branch recovery, and associative caches
• ConvoLite – A CUDA-accelerated deep learning engine with shared memory tiling and pipelined convolution
• μTracker – A microarchitectural profiling framework with integrated testbenches and RTL debug automation

📞 Live walkthroughs available by request.

Most of my work is private to protect intellectual property and maintain academic integrity.

I do not share sensitive source code publicly, especially coursework or competitive designs. However, I'm happy to:

• Share proof of authorship and commit history
• Walk through projects live over a verified call
• Discuss design, test, and debug methodology

🛡️ Contact me for verified access or demonstrations.

🔬 Architecture & RTL Design

• Superscalar, out-of-order, speculative execution pipelines
• Tomasulo's algorithm, register renaming, CDB, PRF, ROB
• Associative and non-blocking cache design (write-back, write-allocate)
• RTL: SystemVerilog, Verilog, Chisel (exploring)

🛠️ Hardware Design & Verification

• Synthesis, P&R, timing closure, static timing analysis
• ASIC & FPGA workflows — Vivado, Quartus, ModelSim, Verdi
• Formal verification via JasperGold, plus UVM-style validation
• CMOS and VLSI design techniques

⚡ High-Performance Computing & Parallelism

• CUDA (shared memory tiling, coalesced access, INT8 quantization)
• OpenMP, multithreading, low-latency pipelines
• Real-time image/video processing on FPGAs (30 FPS @ high-res)

🧠 Software & Simulation

• C/C++, Python, Bash, MATLAB, TCL, Perl
• Embedded systems (Arduino, STM32, custom SoCs)
• Web development (HTML/CSS/JS, Flask, Node, Firebase)
• Linux internals, kernel mods, custom drivers, networking stacks
• Database design, file system tweaking, cross-arch simulation

🧰 Tools & Platforms

• Cadence Virtuoso, Synopsys Verdi, Vivado, Quartus, SLURM, Makefiles
• Git, GitHub, GitOps workflows, CI/CD
• OS-native scripting (Linux, macOS, Windows PowerShell)

🧩 AI/ML Hardware Optimization

• ResNet-50 inference acceleration with TensorRT
• INT8 quantization, CUDA tensor cores
• 4x+ speedup over CPU baselines for real-time inference

🧑‍🔬 Meta-Skills

• 🧠 Extreme learning agility — I pick up new stacks like I’ve been doing them for years
• 🔧 Builder’s instinct — from fixing engines to optimizing memory subsystems
• 💡 Hyper-resourceful — nothing gets blocked, everything gets solved
• 🔥 Full-stack to full-system — I don't need a framework. I am the framework

📧 Email: [email protected] · [email protected]

If you see it like a video game, where the daily and weekly "bonuses" are less bugs and less trouble, it becomes much easier to make even one small improvement daily which compounds over time!