Personal Information
Unternehmen/Arbeitsplatz
San Francisco Bay Area United States
Beruf
Research Fellow at Stanford University
Branche
Education
Webseite
http://www.cs.utexas.edu/~ardavan/
Info
My research is in High performance computing and Computer architecture.
My PhD dissertation title is "Algorithm/Architecture Co-Design of Low Power and High Performance Linear Algebra Compute Fabrics"
I was the sole PhD student that initiated this project which also won the National Science Foundation Grant.
The research includes:
Cycle accurate micro-architectural simulator for low power high performance linear algebra accelerators
Power and performance model analysis of modern GPUs and CPUs for linear algebra applications
State of the art Micro architectural level research on Floating Point Multiply Accumulate (MAC) Units
Specialties: C/C++, P-threads, MPI, Verilog, VHDL, Visual...
Tags
parallelism
accelerator
matrix multiplication
locality
custom architecture
gemm
asic
linear aglebra
memory
fft
energy efficiency
dark silicon
computer hardware
Mehr anzeigen
Präsentationen
(2)Gefällt mir
(1)Personal Information
Unternehmen/Arbeitsplatz
San Francisco Bay Area United States
Beruf
Research Fellow at Stanford University
Branche
Education
Webseite
http://www.cs.utexas.edu/~ardavan/
Info
My research is in High performance computing and Computer architecture.
My PhD dissertation title is "Algorithm/Architecture Co-Design of Low Power and High Performance Linear Algebra Compute Fabrics"
I was the sole PhD student that initiated this project which also won the National Science Foundation Grant.
The research includes:
Cycle accurate micro-architectural simulator for low power high performance linear algebra accelerators
Power and performance model analysis of modern GPUs and CPUs for linear algebra applications
State of the art Micro architectural level research on Floating Point Multiply Accumulate (MAC) Units
Specialties: C/C++, P-threads, MPI, Verilog, VHDL, Visual...
Tags
parallelism
accelerator
matrix multiplication
locality
custom architecture
gemm
asic
linear aglebra
memory
fft
energy efficiency
dark silicon
computer hardware
Mehr anzeigen