Skip to content
View JaiminKadiwal's full-sized avatar

Block or report JaiminKadiwal

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
JaiminKadiwal/README.md
Jaimin Kadiwal. RTL design, design verification, FPGA. A timing diagram with clk, valid and a data bus carrying RTL, verify, FPGA.

Hello, I'm a grad student at Johns Hopkins. I build digital hardware and then try to prove it wrong.

Most of what's here is SystemVerilog checked against a reference model, formal proofs and real hardware when I have it.

Right now I'm adding a controller to my systolic-array accelerator, and next I'm putting a UVM testbench on it. Below are the projects that are done and verified.

Project What it is Evidence
tensor-mac INT8 systolic-array matrix multiply Matches NumPy on 500 random tiles; formal proof that the accumulator never overflows
verified-rtl-blocks FIFO, skid buffer, round-robin arbiter 7,500+ random checks; formal proofs of no overflow and no starvation
fpga-keyword-spotting int8 neural net in VHDL on an Artix-7 Bit exact with Python on 128/128 vectors, on the board
risc8-rtl-to-gdsii 8-bit CPU from Verilog to 45 nm layout Six-corner sign-off at 7 ns with +1.24 ns slack
Radiation-hardened SRAM Reproduced a published 16T cell, then removed 3 transistors Read SNM +33%, hold SNM -64%

Tools: SystemVerilog, VHDL, Verilator, Icarus, cocotb, SymbiYosys, Yosys, Vivado, Cadence Genus / Innovus / Xcelium, Python, C++.

Pinned Loading

  1. tensor-mac tensor-mac Public

    INT8 systolic-array matrix-multiply accelerator in SystemVerilog. Output-stationary N x N PE grid, bit-exact with NumPy on 500 random tiles, accumulator overflow ruled out by an unbounded SymbiYosy…

    SystemVerilog

  2. risc8-rtl-to-gdsii risc8-rtl-to-gdsii Public

    8-bit multi-cycle RISC CPU taken from Verilog to signed-off layout on 45 nm: Genus synthesis, Innovus place and route, six-corner MMMC sign-off with OCV and SI. Closes at 7 ns with +1.24 ns slack; …

  3. verified-rtl-blocks verified-rtl-blocks Public

    Streaming RTL building blocks in SystemVerilog: FIFO, skid buffer, round-robin arbiter with ready/valid handshakes. FIFO and arbiter scoreboarded against Python golden models over 7,500+ constraine…

    SystemVerilog

  4. fpga-keyword-spotting fpga-keyword-spotting Public

    Keyword spotting with the neural network in FPGA fabric: an int8 MLP in VHDL on an Artix-7 (Cmod A7) classifies live MFCC features streamed over UART. Bit-exact with the Python fixed-point model on…

    VHDL

  5. HP13T-fault-tolerance-SRAM-cell HP13T-fault-tolerance-SRAM-cell Public

    HP13T: an experiment on a published radiation-hardened SRAM cell in 45 nm CMOS. Removed 3 transistors from HP16T to see what they do: access times unchanged, read SNM +33%, hold SNM -64%, and it st…