gpu   8853

« earlier    

[1806.07060] A model-driven approach for a new generation of adaptive libraries
Efficient high-performance libraries often expose multiple tunable parameters to provide highly optimized routines. These can range from simple loop unroll factors or vector sizes all the way to algorithmic changes, given that some implementations can be more suitable for certain devices by exploiting hardware characteristics such as local memories and vector units. Traditionally, such parameters and algorithmic choices are tuned and then hard-coded for a specific architecture and fo...
machine-learning  gpu  libraries 
yesterday by hschilling
What is Futhark? | Futhark - A High Performance Functional Array Language
Futhark is a small programming language designed to be compiled to efficient parallel code. It is a statically typed, data-parallel, and purely functional array language in the ML family, and comes with a heavily optimising ahead-of-time compiler that presently generates GPU code via OpenCL, although the language itself is hardware-agnostic. As a simple example, this function computes the average of an array of 64-bit floating-point numbers:
functional  gpu  data  apl 
2 days ago by cothrun
Running Rust on the GPU with Accel |
For the past month or so, I’ve been working on a follow-up to my series on Writing a Raytracer in Rust. This time around, I’ll be talking about writing a GPU-accelerated Path Tracer. As always, I’m writing it in Rust - including the GPU kernel code. Compiling Rust for GPUs at this point is difficult and error-prone, so I thought it would be good to start with some documentation on that aspect of the problem before diving into path tracing.
llvm  rust  gpu  cuda  windows 
5 days ago by ianweatherhogg
Paperspace: Cloud Machine Learning, AI, and effortless GPU infrastructure.
We are building out the world's most powerful GPU cloud. Accelerate your ML, AI, and data science workflow. Paperspace is trusted by some of the world's most respected companies. Learn more about how Paperspace can work for your business. via Pocket
cloud  gpu  infrastructure  machine  neural 
5 days ago by kintopp
WebGL2 powered geospatial visualization layers
webgl  geospatial  gpu  MIT  datavis 
6 days ago by sushidub
Experiments in GPU-based occlusion culling part 2: MultiDrawIndirect and mesh lodding – Interplay of Light
related to the skysaga digitialdragons talk.

This one goes into more detail about choosing LODs on the GPU as well.
culling  gpu 
8 days ago by MemoryStomp
an AMD initiative designed to enable developers to create ground-breaking PC games, computer generated imagery and GPU computing applications for great performance and lifelike experiences using no cost and open development tools and software.
gpu  gpgpu  hardware  computergraphics 
10 days ago by Z303

« earlier    

related tags

(proxmox  -  1st  2020  2d  2u  3d  43  [2014]  a  acceleration  accuracy  ai  alex  amd  and  api  apl  arm  array-computing  article  audio-generation  azure  benchmark  benchmarks  bi  book  build  building  by  c++  cg  claims  client  cloud  code  coding  community  comparch  compiler  compression  compute  computer  computergraphics  console  consumption  costs  cpu  cryptocurrency  cuda  culling  data  database  datavis  deep-learning  deep  development  distributed-computing  dmastiers  docker  docs  documentation  dotnet  double  download  downloads  driver  editing  editor  efficiency  egpu  energy  erosion  examples  faiss  floatingpoint  fonts  for  from  functional  fusion  futhark  gamedev  gaming  gcc  gentoo  geospatial  gigabyte  github  glass  google-cloud  gpgpu  gpu_passthrough  graphics  group  guide  gvt-g  hardware  high-performance-computing  honor  hsa  hsail  icon  iee754  ieee  image-editing  image  images  improve  in  infrastructure  inspiration  install  installation  intel  intro  introduction  iommu  java  keras  kernel  kvm  language  laptop  launch  learn  learning  lib  libraries  library  libvirt  linux  llvm  looking  lte  mac  machine-learning  machine  machine_learning  macos  mathematics  mbp  mediated  mini  mining  mit  ml  navi  neural  nn  noise  nuc  numbers  nvidia  oil  openacc  opencl  opencv  opengl  osx  ovmf  paas  papers  pass-through  pass  passthrough  pc  pcie  pcpart  performance  piperesearch  power  precision  processing  profiling  proglang  programming  ps5  py  python  pytorch  raytracing  razer  realtime  reduce  reference  rendering  repo:github  repository  rnn  rounding  rust  ryzen  server  setting  shader  simd  slot  soc  source  spark  speculated  speech-generation  spreadsheet  sse  sse2  staging  supercharge  supercomputing  talk  tech  technology  tensor-flow  tensorflow  terminal  the  theano  through  thunderbolt  ti  tinkerboard  to  tooling  toread  tpu  training  tuning  turbo  tutorials  tweet  ubuntu  uefi  unraid  up  usecase  vfio  vga  video  virtualization  vm  volta  vs  vulkan  webgl  williamson  windows  with  work  x86  xfce  you 

Copy this bookmark: