% Monte Carlo estimate of pi from 20 BILLION samples, drawn in 2000 batches.
% (Pasting into MATLAB? Cut batches to ~50 -- generating this many doubles
% takes CPUs minutes, which is rather the point.)
% The draws, the comparison and the mean all fuse into one reduction pass,
% so each batch's 10 million points are generated, tested and counted
% without ever touching GPU memory.
n = 10000000;
batches = 2000;
s = zeros(1, 1);
tic;
for k = 1:batches
  s = s + mean(((2*rand(n,1) - 1).^2 + (2*rand(n,1) - 1).^2) <= 1);
end
t = toc;
fprintf('pi ~ %.6f  (%.1f billion samples, %.0f million samples/sec)\n', ...
  4*s/batches, batches*n/1e9, batches*n/1e6/t);