Usually, located at /usr/local/cuda/bin
$ nvprof python train_mnist.py
I prefer to use --print-gpu-trace.
importPackage(Packages.com.sk89q.worldedit.blocks); | |
importPackage(Packages.com.sk89q.worldedit); | |
//joins up blocks in the dimension dim | |
function squash(blockarray, dim, arraysize){ | |
var i = new Array(); | |
var dims = new Array(); | |
for (var y = 0; y <= arraysize*2; y++){ |
importPackage(Packages.com.sk89q.worldedit.blocks); | |
importPackage(Packages.com.sk89q.worldedit); | |
//joins up blocks in the dimension dim | |
function squash(blockarray, dim, sx, sy, sz, arraysize){ | |
var i = new Array(); | |
var dims = new Array(); | |
for (var y = 0; y <= sy; y++){ |
// points = number of points (minimum 3) | |
// outer = radius to outer points | |
// inner = radius to inner points | |
module Star(points, outer, inner) { | |
// polar to cartesian: radius/angle to x/y | |
function x(r, a) = r * cos(a); | |
function y(r, a) = r * sin(a); | |
// angular width of each pie slice of the star |
#!/bin/bash | |
# install CUDA Toolkit v9.0 | |
# instructions from https://developer.nvidia.com/cuda-downloads (linux -> x86_64 -> Ubuntu -> 16.04 -> deb) | |
CUDA_REPO_PKG="cuda-repo-ubuntu1604-9-0-local_9.0.176-1_amd64-deb" | |
wget https://developer.nvidia.com/compute/cuda/9.0/Prod/local_installers/${CUDA_REPO_PKG} | |
sudo dpkg -i ${CUDA_REPO_PKG} | |
sudo apt-key adv --fetch-keys http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/7fa2af80.pub | |
sudo apt-get update | |
sudo apt-get -y install cuda-9-0 |
nice infographics on https://www.reddit.com/r/MechanicalKeyboards/comments/jihbhi/custom_mechanical_keyboard_infographic_v30_is_now/
Cherry MX Brown - to make it similar to the Sun Type 7 keyboard I was using previously (excellent keyboard however too wide).
Fn key - the "house/home" key