A long while ago, I published a few posts describing my home cluster setup. It felt like a great achievement at the time, but truth to be told, the physical layout wasn't portable, and that caused me a lot of trouble during the last move...
Author: Mateusz Kaczanowski
# Parallella (part 13): Closing notes
The jury returned a unanimous guilty verdict. Guilty...
# Parallella (part 12): Tensorflow?
What can you do with Parallella? - that question still haunts me in the night. Even though we wrote the scheduler application before, I think we all agree that there must be a better way to utilize Epiphany chip. In the search for a better usecase, I found a Tensorflow port for microcontrollers. That got me thinking...
# Parallella (part 11): malloc
If I were to name one thing that makes modern programming languages great, I'd pick dynamic memory allocator (often interfaced via malloc/free functions). Imagine yourself keeping track of memory allocations, drowning in seemingly random addresses, and handling the fragmentation manually... Madness to say, at least.