gpu   9455

« earlier    

BlazingDB | Distributed GPU SQL Engine on RAPIDS
"BlazingDB helps enterprise organizations operationalize their data lakes. We offer fast, simple, scalable GPU-powered SQL engine that runs directly off of files available inside enterprise data lakes. Query tens to hundreds of terabytes in seconds with our state of the art SQL analytics engine."

Similar to MapD? https://www.omnisci.com/
sql  gpu 
yesterday by tobym
Crypto crash leads to inventory pile-up at Nvidia, sales slaughtered • The Register
The top line slide for the quarter was "driven primarily" by a 45 per cent drop in gaming revenue to $954m, "weaker than expectations heading into the quarter", said CFO Colette Kress on an earnings call with financial analysts.

She said: "Post crypto inventory of GPUs in the channel caused us to reduce shipments in order to allow excess channel inventory to sell through." The level of stock is predicted to "normalise" in the current Q1 of fiscal '20. Nvidia made a big gamble on crypto miners snaffling up fat GPUs and it simply hasn't paid off.

The second reason given from the gaming sales collapse was "deteriorating macro economic conditions, particularly in China, impacted consumer demand for our GPUs". Again, this is another challenge voiced by Apple.

The third and final excuse was that sales of the high-end next-gen architecture, the GeForce RTX 2080 and 2070, came in lower than planned.
nvidia  mercato  problemi  gpu  bitcoin  criptovalute  videogame  spunti 
5 days ago by nicoladagostino
Video: Nvidia support was abandoned in macOS Mojave, and here's why
The first Mac to include an Nvidia graphics processor was released in 2001, but Apple was still using chips made by ATI, the company that was eventually bought out by AMD in 2008.

In 2004, the Apple Cinema Display was delayed, reportedly because of Nvidia's inability to produce the required graphics card, the GeForce 6800 Ultra DDL.

Then in 2008, Apple's MacBook Pro shipped with Nvidia graphics chips that revolutionized the MacBook by taking over the functions of the Northbridge and Southbridge controllers alongside actual graphics rendering. Because of it, Intel filed a lawsuit against Nvidia, making things a bit complicated for Apple.


Nvidia processor in a 2008 MacBook Pro


Not only that, but Apple had to admit that some 2008 MacBook Pros had faulty Nvidia processors, which led to a class-action lawsuit for Nvidia and lost profits for Apple due to MacBook Pro repairs.

Around the same time, the iPhone transformed the mobile computing market and meant phones now needed GPUs, and Apple decided to go with Samsung instead. At this time, Nvidia believed that its own patents also applied to mobile GPUs, so they filed patent infringement suits against Qualcomm and Samsung, trying to get them and possibly Apple to pay license fees.

In 2016, Apple said no to putting Nvidia processors in the 15-inch MacBook Pro and instead went with AMD, publicly because of performance per watt issues.


Nvidia competitor AMD recently launched the Radeon 7, a high-performance 7-nanometer GPU that reportedly has drivers on the way for macOS Mojave


And now, in 2019, there aren't any functional drivers for modern cards in Mojave at all. In October of 2018, Nvidia issued a public statement stating that Apple fully controls drivers for macOS and that they can't release a driver unless it's approved by Apple.
nvidia  storia  sda  soa  spunti  apple  mac  macintosh  gpu  amd 
5 days ago by nicoladagostino
MacVidCards.com - Home
GPU for vintage Mac Pros. Large selection.
GPU  video  video_processors 
8 days ago by tomjlowe
Introducing AresDB: Uber’s GPU-Powered Open Source, Real-time Analytics Engine | Uber Engineering Blog
The most common problem that real-time analytics solves at Uber is how to compute time series aggregates, calculations that give us insight into the user experience so we can improve our services accordingly. With these computations, we can request metrics by specific dimensions (such as day, hour, city ID, and trip status) over a time range on arbitrarily filtered (or sometimes joined) data. Over the years, Uber has deployed multiple solutions to solve this problem in different ways.
gpu  database  uber  mpp  time  series  erick 
9 days ago by bikesandbooks

« earlier    

related tags

2018  2019  2080  7nm  abstract.algebra  acceleration  aclu  advice  agriculture  ai  akitio  alienware  amazon  amd  analytics  android  apple  application  approximate  array  article  artificial  artificialintelligence  as  aws  azure  benchmark  benchmarks  bias  bitcoin  blazingsql  blender  blog  boost  browser  c++  c  cache  camera  candidate  cg  check  checkup  china  chrome  chromebook  clojure  cloud  code  coders  commercial  comparison  compilation  compile  compilers  compute_shaders  computer-science  computergraphics  computervision  containers  cool  cooling  counterexample  course  courses  cpu  crack  criptovalute  cryptocurrency  cryptography  cuda  culling  d3d  data-parallelism  data-science  data  database  datastore  db  deep-learning-compiler  deep-learning  deep  deep_learning  deeplearning  dell  dev  development  diagnostic  diagnostics  docker  dynamics  ecs  eff  efi  egpu  erick  external  f#  face  facialrecognition  fastai  fauxtomation  flask  fluid  fluids  for  free  freeware  friend  fsm  functional-programming  functional  futhark  g  gamedev  gamer  gaming  gcp  giordano  glsl  google  gpgpu  grait  graph-processing  graph-visualization  graphics  grin  gui  hack  hardware  hash  hdd  health  history  hosting  household.it  howto  html  hypervisor  ibm  image  images  intelligence  iommu  ios  js  julia  jupyter  lab  labor  laptop  law  learning  library  license  lightweight  linux  login  low-level  mac  macbook.pro  macbook  machine-learning  machine  machinelearning  macintosh  macos  marco  mechanicalturk  mem  memory  mercato  mesh  metal  microsoft  mimblewimble  miner  mit  ml  monitoring  moore's  moore  mpc  mpp  nearest  neighbor  neuralnets  neuralnetworks  ngc  node  notebook  nouveau  numeric  numerical-programming  numericalcomputing  nvidia's  nvidia-docker  nvidia-p4  nvidia  nvram  opencl  opengl  opensource  optimal  optimisation  optimization  optimizing-compilation  os  outsourcing  overclocking  packet  paper  parallel  parallelism  parsing  passthrough  password  path_tracing  patternrecognition  pbr  pci  performance  physics  pixelbook  platform  plt  pocl  police  postgresql  practical  prediction  privacy  problemi  processing  profiler  profiling  programming-language  programming  project  proxmox  python  pytorch  quantlib  r  radeon  rainbow  raspberrypi  raspi  raykurzweil  realtime  recommendation  regualation  rekognition  renderer  rendering  research  resource  response  rtx  salt  science  scientificcomputing  sda  search  security  sensors  series  server  service  sha  shaders  shading  simd  similarity  simulation  smartphone  soa  socialmedia  software  spark  spunti  sql  ssd  storia  surveillance  sysadmin  table  technology  tensorflow  theory  threads  thunderbolt  thunderbolt2  time  tips  to  tools  toread  tpu  tricks  tutorial  uber  ubuntu  ui  unity  unveils  vc4cl  vertex  video  video_processors  videogame  videos  vii  virtualization  visualcommonsense  vulkan  weapon  web  webgl  webgpu  windows  work 

Copy this bookmark:



description:


tags: