binga / Cloud Gpus
This repository contains information about Cloud GPU offerings for Machine Learning practitioners.
Stars: ✭ 395
Projects that are alternatives of or similar to Cloud Gpus
Gpu Physics Unity
Through this configuration, no per voxel data is transferred between the GPU and the CPU at runtime.
Stars: ✭ 342 (-13.42%)
Mutual labels: gpu
Cuda Api Wrappers
Thin C++-flavored wrappers for the CUDA Runtime API
Stars: ✭ 362 (-8.35%)
Mutual labels: gpu
Ilgpu
ILGPU JIT Compiler for high-performance .Net GPU programs
Stars: ✭ 374 (-5.32%)
Mutual labels: gpu
Curl
CURL: Contrastive Unsupervised Representation Learning for Sample-Efficient Reinforcement Learning
Stars: ✭ 346 (-12.41%)
Mutual labels: gpu
Kotlindl
High-level Deep Learning Framework written in Kotlin and inspired by Keras
Stars: ✭ 354 (-10.38%)
Mutual labels: gpu
Mapd Charting
Dimensional charting built to work natively with crossfilter rendered using d3.js
Stars: ✭ 368 (-6.84%)
Mutual labels: gpu
Arrayfire Python
Python bindings for ArrayFire: A general purpose GPU library.
Stars: ✭ 358 (-9.37%)
Mutual labels: gpu
Aparapi
The New Official Aparapi: a framework for executing native Java and Scala code on the GPU.
Stars: ✭ 352 (-10.89%)
Mutual labels: gpu
Realsr Ncnn Vulkan
RealSR super resolution implemented with ncnn library
Stars: ✭ 357 (-9.62%)
Mutual labels: gpu
Rendu
A simple realtime graphics playground for experimentations.
Stars: ✭ 343 (-13.16%)
Mutual labels: gpu
Hipsycl
Implementation of SYCL for CPUs, AMD GPUs, NVIDIA GPUs
Stars: ✭ 377 (-4.56%)
Mutual labels: gpu
Bayadera
High-performance Bayesian Data Analysis on the GPU in Clojure
Stars: ✭ 342 (-13.42%)
Mutual labels: gpu
Neogfx
Cross-platform GPU-oriented C++ application/game framework
Stars: ✭ 362 (-8.35%)
Mutual labels: gpu
Deep Learning From Scratch 3
『ゼロから作る Deep Learning ❸』(O'Reilly Japan, 2020)
Stars: ✭ 380 (-3.8%)
Mutual labels: gpu
Cloud GPU Vendors
This guide compares on-demand GPU vendors to help Machine Learning practitioners pick their preferred platform. Whether you're an enthusiast, newcomer or expert this guide will help you track updates and keep your costs low.
Cloud GPU Vendors
- "-" means not available
- "???" means maybe available
- GPU prices are per hour
- Storage prices are per GB/month
| GPU Vendor | K80 Price | V100 Price | Storage Price | Last Updated |
|---|---|---|---|---|
| Google Colaboratory (Free TPU***) | Free | [email protected] | 15GB max | 8-Sep-18 |
| Kaggle Kernels (Free TPU***) | Free | - | 8GB max | 8-Sep-18 |
| Salamander | $0.36 | $1.17 | $0.10 | 8-Sep-18 |
| Paperspace | $0.59 | $2.30 | $0.10 | 8-Sep-18 |
| Vast.AI | $0.64 | $2.30 | - | 19-Sep-18 |
| Oracle Cloud | - | $2.25 | $0.04 | 8-Sep-18 |
| Spell | $0.9 | $3.06 | - | 24-Oct-19 |
| Crestle | $0.59 | - | $0.42 | 8-Sep-18 |
| Microsoft Azure | $0.87 | $2.95 | $0.10 | 8-Sep-18 |
| LeaderGPU | - | $3.00 | $0.13 | 8-Sep-18 |
| Amazon Web Services (AWS) | $0.90 | $3.06 | $0.10 | 8-Sep-18 |
| DeepCognition | $0.90 | $3.06 | ??? | 8-Sep-18 |
| Google Cloud Platform (GCP) | $1.21 | $3.24 | $0.03 | 8-Sep-18 |
| FloydHub | $1.20 | $4.20 | 100GB included | 8-Sep-18 |
| IBM Cloud | $2.65 | - | ??? | 8-Sep-18 |
| Alibaba Cloud | - | ??? | ??? | 8-Sep-18 |
| Lambda Labs | - | ??? | ??? | 8-Sep-18 |
| Nvidia Cloud | ??? | ??? | ??? | 8-Sep-18 |
| Riseml | - | ??? | ??? | 8-Sep-18 |
| Valohai | ??? | ??? | ??? | 8-Sep-18 |
| VScaler | ??? | ??? | ??? | 8-Sep-18 |
| Hostkey | ??? | - | ??? | 6-Oct-20 |
- *** Colaboratory provides free access to Google's TPUs v2, and Kaggle has TPUs v3.
- @ - It's now possible to connect Google Colaboratory to a V100 by connecting it to a GCP instance.
Other GPUs
| GPU Vendor | P100 Price | GTX 1080 Price | Storage Price | Last Updated |
|---|---|---|---|---|
| Snark AI | $0.10 | - | ??? | 8-Sep-18 |
| Vast.AI | $1.65 | $0.10 | - | 8-Sep-18 |
| Oracle Cloud | $1.28 | - | $0.04 | 8-Sep-18 |
| Vector Dash | - | $0.51 | ??? | 8-Sep-18 |
| LeaderGPU | $1.50 | $0.75 | $0.13 | 8-Sep-18 |
| Paperspace | $1.72 | - | $0.10 | 8-Sep-18 |
| Microsoft Azure | $2.00 | - | $0.13 | 8-Sep-18 |
| Ovh | - | $1.15 | $0.09 | 8-Sep-18 |
| Neptune ML | $2.15 | - | $0.28 | 8-Sep-18 |
| Google Cloud Platform (GCP) | $2.22 | - | $0.03 | 8-Sep-18 |
| Alibaba Cloud | ??? | - | ??? | 8-Sep-18 |
| Exoscale | ??? | - | ??? | 8-Sep-18 |
| IBM Cloud | ??? | - | ??? | 8-Sep-18 |
| Hostkey | ??? | - | ??? | 6-Oct-20 |
Notes
-
Google Colaboratory, Kaggle Kernels & Crestle are only for Jupyter Notebooks and limited to 6 / 12 / 24 hour sessions
-
The prices for Google Cloud Platform assume you've attached the GPUs to a "n1-standard-16" instance.
-
250GB Paperspace volumes are $0.04 per GB; larger or smaller volumes cost more
Free Compute Credits
-
Google Cloud Platform gives $300 credits to anyone
-
AWS gives $110 credits to students, in partnership with GitHub
-
Azure gives $100 credits to students
-
Paperspace gives $15 credits to fast.ai students
-
Vast.AI, DeepCognition, Snark AI, Salamander & Crestle give approximately $1 credits to anyone
-
Hostkey give free trials to science projects and competitions
GPU Datasheets
Nvidia manufactures all the GPUs listed here
| GPU | TFlops (double / single precision) |
|---|---|
| K80 | 1.9 / 5.6 |
| V100 | 7.8 / 15.7 |
| P100 | 5.3 / 10.6 |
| GTX 1080 | 8.8 / 8.8 |
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].