binarybana

Using the below four author profiles as a guide, help me write my own technical blog profile for the NVIDIA tech blog.

```
About Andrew Kerr
Andrew is a Senior GPU Compute Architect at NVIDIA. He joined NVIDIA's Compute Architecture group in 2012 after finishing his Ph.D. at Georgia Institute of Technology. Lately, Andrew's technical focus has been to design and implement abstractions for linear algebra on GPUs to facilitate programmability as performance continues to scale.
View all posts by Andrew Kerr
Duane MerrillAbout Duane Merrill
Duane Merrill is a Senior Research Scientist with NVIDIA Research.  His principal research interests are parallel algorithm and programming model design. In particular, his work has focused on problems involving sparse, irregular, and cooperative computation, notably parallel sorting and graph algorithms. He is the author of CUB, a library of “collective” software primitives to simplify CUDA kernel construction, performance tuning, and maintenance. He received his Ph. D. from

binarybana

import tvm 
from tvm import relay
from tvm.contrib import graph_runtime

import numpy as np

shapes = []
for sort in [1000, 10000, 100000, 1000000]:
    for non_sort in [1, 10, 100]:
        for axis in [0, 1, 2]:

binarybana

Keybase proof

I hereby claim:

• I am binarybana on github.
• I am jmknight (https://keybase.io/jmknight) on keybase.
• I have a public key ASAfhwLxKMGSO7MGvJqN0H3b8edS8W1IvVK05MDnY2YrvAo

To claim this, I am signing this object:

binarybana

{
  "AWSTemplateFormatVersion": "2010-09-09",
  "Description": "Kubernetes 1.0.4 on EC2 powered by CoreOS 766.3.0 (stable)",
  "Mappings": {
      "RegionMap": {
          "eu-central-1" : {
              "AMI" : "ami-bececaa3"
          },
          "ap-northeast-1" : {
              "AMI" : "ami-f2338ff2"

binarybana

using PyPlot
using Base.Test

function minops{T<:Integer}(n::T)
  queue = (T,T)[]
  push!(queue, (n,0))
  visited = Set{T}()
  pops,evals = 0,0
  while true
    pops += 1

binarybana

data = rand(0:1000,48,18000)

function test()
        reload("uniquename2.jl")
        normfac = vec(mapslices(x->quantile(x,0.75), data, 2))
        cls = MPM.mpm_classifier(rand(0:100,4,3), rand(0:100,4,3))
        MPM.sample(cls,1)
end

for i=1:100

binarybana

jason@toxicvm:~/test3$ ~/src/julia/julia testcase.jl
1
2
Warning: replacing module MPM
3
Warning: replacing module MPM
4
Warning: replacing module MPM
5

binarybana

julia> include(joinpath(Pkg.dir(), "OBC/test/testcase.jl"))
(1.8508481979370117,0.0008389949798583984)

julia> include(joinpath(Pkg.dir(), "OBC/test/testcase.jl"))
Warning: replacing module OBC
Warning: replacing module NSum
Warning: replacing module MPM
Warning: Method definition sum(Float64,) in module MPM at /home/jason/.julia/v0.4/OBC/src/mpm.jl:141 overwritten in module MPM at /home/jason/.julia/v0.4/OBC/src/mpm.jl:141.
Warning: replacing module SAMC

binarybana

using Mamba

data = (Symbol=>Any)[:x1 => [10, 12, 14, 9],
    :x2 => [10,11,12,8]]

N = length(data[:x2])
data[:N] = N

d=10.0

binarybana

using StatsBase

randsubseq2(A::AbstractVector, m::Integer) = randsubseq2!(Array(eltype(A),m), A)

function randsubseq2!(S::AbstractVector, A::AbstractArray)
    m = length(S)
    m == 0 && return S # needed for correctness: code below requires m > 0
    n = length(A)
    m == n && return copy!(S, A)
    0 <= m <= n || throw(DomainError())