Skip to content

Instantly share code, notes, and snippets.

Avatar
🐰
Deep in the rabbit hole

Carsten Bauer carstenbauer

🐰
Deep in the rabbit hole
View GitHub Profile
View SIMDjl_performance.jl
julia> function vector_dot(B, C)
a = zero(eltype(B))
for i in eachindex(B,C)
@inbounds a += B[i] * C[i]
end
return a
end
vector_dot (generic function with 1 method)
julia> function vector_dot_macro(B, C)
View readqcd.jl
using DelimitedFiles
using LinearAlgebra
using Test
function readqcd(fname::AbstractString; verify=true)
open(fname, "r") do f
header_str = readuntil(f, "END_HEADER\n")
header_mat = readdlm(IOBuffer(header_str), '=')
header = Dict(strip(header_mat[i, 1]) => header_mat[i, 2] for i in 2:size(header_mat, 1))
@carstenbauer
carstenbauer / nvlink_bench.jl
Last active Feb 8, 2022
Nvidia NVLink Bandwidth Measurement with Julia
View nvlink_bench.jl
@carstenbauer
carstenbauer / Manifest.toml
Created Jan 27, 2022
Attempt to measure core-to-core latency in Julia
View Manifest.toml
# This file is machine-generated - editing it directly is not advised
julia_version = "1.7.1"
manifest_format = "2.0"
[[deps.ArgTools]]
uuid = "0dad84c5-d112-42e6-8d28-ef12dabb789f"
[[deps.Artifacts]]
uuid = "56f22d72-fd6d-98f1-02f0-08ddc0907c33"
View gist:909003a87bcdd2aef228591728394972
{
"cells": [
{
"cell_type": "markdown",
"id": "202a91a0-8ac9-4d47-a628-85786980cbb2",
"metadata": {},
"source": [
"# CPU\n",
"\n",
"## Single-threaded sparse mat-vec product"
View gist:33a5b57de78cce6a3cecef88d4e2c5e7
{
"cells": [
{
"cell_type": "markdown",
"id": "202a91a0-8ac9-4d47-a628-85786980cbb2",
"metadata": {},
"source": [
"# CPU\n",
"\n",
"## Single-threaded sparse mat-vec product"
@carstenbauer
carstenbauer / bench.jl
Created Aug 9, 2021
SAXPY benchmark CUDA.jl (broadcasting)
View bench.jl
print("Loading modules...")
using BenchmarkTools
using CUDA
using DataFrames
using CSV
using Printf
println("done!")
const a = 3.1415f0
@carstenbauer
carstenbauer / gpu_saxpy_comparison.jl
Last active Jul 26, 2021
CUDA broadcasting vs kernel
View gpu_saxpy_comparison.jl
# ==============================================================================
# Example program for CUDA Julia tutorial
# Written by: Carsten Bauer
# ==============================================================================
using BenchmarkTools
using CUDA
using Printf
function saxpy_gpu_kernel!(z,a,x,y)
i = (blockIdx().x - 1) * blockDim().x + threadIdx().x
View gist:fd6cc660a4aad707ba087919a2d0eed0
(itensors) pkg> add ITensors
Updating registry at `/p/project/chku27/hku273/.julia/registries/General`
┌ Error: curl_easy_setopt: 48
└ @ Downloads.Curl /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.6/Downloads/src/Curl/utils.jl:36
┌ Error: curl_easy_setopt: 48
└ @ Downloads.Curl /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.6/Downloads/src/Curl/utils.jl:36
┌ Error: curl_easy_setopt: 48
└ @ Downloads.Curl /buildworker/worker/package_linux64/build/usr/share/julia/stdlib/v1.6/Downloads/src/Curl/utils.jl:36
Resolving package versions...
┌ Error: curl_easy_setopt: 48
View meshgrid.jl
"""
meshgrid(xvec) = meshgrid(xvec, xvec)
Produces a 2D meshgrid `X,X` by repeating xvec in y-dimension and xvec in x-dimension.
"""
meshgrid(v::AbstractVector{T}) where T<:Number = meshgrid(v, v)
"""
meshgrid(xvec, yvec)
Produces a 2D meshgrid `X,Y` by repeating xvec in y-dimension and yvec in x-dimension.
"""