Dmitry S.

Senior System Architect • Designing internal self-hosted AI automation and batch processing tools using fine-tuned open-source LLM [ Python, C++, Lua, Bash ] • Designing and implementing an Architectural Modeling System for automated mixed simulations (RTL sim, virtualized and native code execution) for IC architectural exploration and evaluation [ C, C++, SystemC, Verilog, SystemVerilog, Bash, TCL, SDL ] • Designing and implementing a PSS-alike verification framework for automated constraint-based test suite generation [ C, Bison, Flex ] • Designing a complex FPGA-based motherboard for hardware verification of ASICs - [ Altium ] • Designing and implementing HW and SW for a scalable, multi-channel, remotely-controlled power supply and power monitor for ASIC verification rack setups [ KiCAD, ALST, C, Python ] • Architectural contributions to various IPs • Writing architectural product specifications [ Dita, Oxygen, various other XML editors ] • Creating various automation tools and general tools for making the life easier [ C++, Qt, SQLite, Bash, etc ]

• Leading a HW development team, designing an AI-driven smart speaker [ Altium, Python ]

• RnD of a complete optical eyetracking solution; • development of an embedded computing platform based on Allwinner A13 SoC (1 GHz, 256 MB DDRII) from scratch (i.e., without a reference design); • RnD of INEUM’s first MCU board incorporating HDI PCB and STM32F4 MCU in WLCSP package. The board was eventually mass-produced; • designing, testing and implementing my own proprietary general-purpose ultra-small footprint RISC core ("CC0416") for a research of hyper-multicore machines; • development of proprietary multicore solutions (with our own board designs) based on CC0416 core: ◦ on Xilinx Virtex-4 (16 cores, 400 MHz clock); ◦ on four Xilinx Spartan-6 (4 х 256 cores, 200 MHz clock); ◦ 4096-core version (four aforementioned PCBs connected via a high-speed bus) was designed, but never manufactured; • development and deployment of a custom RTOS for CC0416, with support of multitasking, dynamic resource allocation and memory management in extensible multicore environment; • various solutions incorporating wide range of FPGAs and CPLDs: Spartan-3, Spartan-6, Virtex-4, Virtex-6, Kintex-7, CoolRunner-II, Mach XO2; • development of a fully articulated humanoid robot; designed and built in my lab from scratch (electronics, mechanics, power delivery, software); • development of proprietary software libraries: ◦ sound source vector approximation with two microphones for STM32-based platform; ◦ cognitive functions emulation library for our robotic research projects; ◦ voice recognition software based on my own offline recognition engine which fits into STM32F407 MCU; • custom 3D-printers development to suit our needs; • various research projects incorporating Deep Learning (using nVidia CUDA Toolkit, Caffe, Torch); • various nVidia Jetson TX-1 and TX-2 -based projects (mostly for computer vision applications)

• restoring old computing machines: ◦ vacuum tube-based “Ural-1” ◦ transistor-based “Razdan-3”