meeting_minutes_with_jason
November 6:
Bryce:
Continue to work on Mandelbrot
Still reading Linux Device Drivers (LDD)
Creating hand-optimized Verilog accelerator
Victor:
Ahmed:
Check timing for PCIe clock used with accelerators
Automate generation of QSYS script / Legup PCIe wrapper
Look into how to specify to Legup scheduler an Fmax target
October 22:
Can the accelerators tell the dma controller to get system memory
Calculate processor only mandelbrot times for an iteration
Handcrafted Mandelbrot Verilog core
Mandelbrot working by 2 weeks
Data point of mandelbrot right now (for final presentation)
Time, area, comparison to software
October 15:
Get mandelbrot working on normal LegUp (Bryce to do next week)
Do gold check against LegUp mandelbrot
Figure out memory coherency
Figure out how to write directly to the processor, rather than memory
Ahmed to see how the accelerators can talk to pcie core, like NIOS does right now
meeting_minutes_with_jason.txt · Last modified: 2012/11/06 17:38 by bryce.long