Real Name: Matthew Frank. My background is in parallel computer architecture. I was on a team that built one of the first many-core processors. I've written runtime systems, operating systems and parallelizing compilers. I taught digital design, computer architecture and operating systems classes at a large land-grant university in the Midwestern United States for a few years. I've developed programming tools and libraries for multi-core and GPGPU. I'm currently a GPGPU architect at NVidia, focused on performance tuning the architecture for Deep Learning.
Member for 8 years
14 profile views
Last seen Jan 3 '20 at 15:33
- Computer Science 16.6k 16.6k 11 gold badge3838 silver badges7777 bronze badges
- Stack Overflow 3.1k 3.1k 11 gold badge1616 silver badges2222 bronze badges
- Signal Processing 1.4k 1.4k 1111 silver badges1818 bronze badges
- Unix & Linux 721 721 88 silver badges88 bronze badges
- Electrical Engineering 569 569 33 silver badges1212 bronze badges
- View network profile
Top network posts
- 86 Why does a processor have 32 registers?
- 52 Time complexity of a compiler
- 44 How do I use a boost condition variable to wait for a thread to complete processing?
- 39 What is making my cursor 'randomly' disappear when using gnome-teminal?
- 34 why do I need std::condition_variable?
- 29 What does the processor do while waiting for a main memory fetch
- 28 Contrasting Peterson’s and Dekker’s algorithms
- View more network posts →