Ivan Yulaev

Software/Hardware Engineer
The Dini Group, La Jolla

Undergraduate Research Assistant
Center for Magnetic Recording Research
University of California, San Diego

Email: ivan at yulaev dot com
Work
Working at the Dini Group, Ivan has designed numerous multi-FPGA ASIC prototyping systems such as the DN7020K10. These systems typically feature many high-powered Xilinx or Altera FPGAs, a plephora of board clocking resources, high-speed serial links, and independent control and configuration via an on-board microprocessor. He has also designed algorithm accelerator cards such as the DNBFC_S12_PCIe, which feature an array of (relatively) low cost Xilinx Spartan or Altera Cyclone FPGAs acting as processors for High-Performance Computing applications. These cards can also be installed into a rack-mount server chassis, complete with a Linux or Windows PC acting as the host system, to build an all-in-one algorithm computing solution. Designing these systems can require a broad skillset, encompassing fields such as
  • Digital Logic Design with FPGAs and embedded microprocessors
  • Signal integrity
  • System Architecture
  • Mechanical Compatibility and Chassis Design
  • RTL and C programming for system test generation
  • FPGA simulation (Modelsim), synthesis, and place-and-route (for writing FPGA-based h/w designs)
  • Design Verification Testing
  • Design for Manufacturing
  • Regulatory Compliance (CE/FCC)
These board are brought to market as quickly as possible, often scheduled so as to be constrained by silicon availablity rather than project development time. Thus the designs are time as well as quality sensitive.


Research
Under Professor Eric Fullerton, Ivan has been involved in the development and simulation of nanoscale magnetic data storage devices, based on Magnetic Tunnel Junctions and Spin Torque Transfer driven Spin Valves. He has developed and used a macromagnetic simulation model to predict the behavior of these elements under externally applied fields and currents. Magnetic tunnel junction/spin valves nano-structures present a compelling alternative to traditional SRAM and embedded DRAM for implementation of fast and dense on-chip memories; Ivan's work focuses on optimizing the parameters of these devices for suitability in embedded applications.

Research Projects & Papers
Recent Talks
  • Spin Transfer Torque in Perpendicular Magnetic Anisotropy Spin Valves. Given at CMRR Research Review (Spring 2011) [Slides] [Notes]
  • Reduction of Critical Switching Current in Perpendicular Spin Valves. Given at UCSD Undergraduate Research Conference (Spring 2011) [Slides] [Notes]