leonardt
Re:Recurrent Neural Networks Hardware Implementation on FPGA
on Monday November 23, 2015 @07:49PM (#10618430)
> Figure 8: The execution time is projected to decrease with the increase of number of LSTM cells running in parallel. This can lead to significant performance improvement.
They do say in the text that "Figure 8 shows the expected speed up, assuming the data throughput is high enough to handle the parallel processing" so take it with a grain of salt. There could and most likely will be (as there always is) other factors that prevent ideal scaling.
Re:Ask HN: I'm in SF for the weekend. What places can I tour that tout...
on Saturday March 07, 2015 @02:05PM (#9162715)
+1 to this, the Cal Academy of the Sciences has an awesome earthquake exhibit right now. Really gives a humbling sense of how insane those earthquakes in the past were.