Wednesday, October 11th 2017

NVIDIA Introduces RAPIDS Open-Source GPU-Acceleration Platform

NVIDIA today announced a GPU-acceleration platform for data science and machine learning, with broad adoption from industry leaders, that enables even the largest companies to analyze massive amounts of data and make accurate business predictions at unprecedented speed.

RAPIDS open-source software gives data scientists a giant performance boost as they address highly complex business challenges, such as predicting credit card fraud, forecasting retail inventory and understanding customer buying behavior. Reflecting the growing consensus about the GPU's importance in data analytics, an array of companies is supporting RAPIDS - from pioneers in the open-source community, such as Databricks and Anaconda, to tech leaders like Hewlett Packard Enterprise, IBM and Oracle.
Analysts estimate the server market for data science and machine learning at $20 billion annually, which - together with scientific analysis and deep learning - pushes up the value of the high performance computing market to approximately $36 billion.

"Data analytics and machine learning are the largest segments of the high performance computing market that have not been accelerated - until now," said Jensen Huang, founder and CEO of NVIDIA, who revealed RAPIDS in his keynote address at the GPU Technology Conference. "The world's largest industries run algorithms written by machine learning on a sea of servers to sense complex patterns in their market and environment, and make fast, accurate predictions that directly impact their bottom line.

"Building on CUDA and its global ecosystem, and working closely with the open-source community, we have created the RAPIDS GPU-acceleration platform. It integrates seamlessly into the world's most popular data science libraries and workflows to speed up machine learning. We are turbocharging machine learning like we have done with deep learning," he said.

RAPIDS offers a suite of open-source libraries for GPU-accelerated analytics, machine learning and, soon, data visualization. It has been developed over the past two years by NVIDIA engineers in close collaboration with key open-source contributors.

For the first time, it gives scientists the tools they need to run the entire data science pipeline on GPUs. Initial RAPIDS benchmarking, using the XGBoost machine learning algorithm for training on an NVIDIA DGX-2 system, shows 50x speedups compared with CPU-only systems. This allows data scientists to reduce typical training times from days to hours, or from hours to minutes, depending on the size of their dataset.

Close Collaboration with Open-Source Community
RAPIDS builds on popular open-source projects - including Apache Arrow, pandas and scikit-learn - by adding GPU acceleration to the most popular Python data science toolchain. To bring additional machine learning libraries and capabilities to RAPIDS, NVIDIA is collaborating with such open-source ecosystem contributors as Anaconda, BlazingDB, Databricks, Quansight and scikit-learn, as well as Wes McKinney, head of Ursa Labs and creator of Apache Arrow and pandas, the fastest-growing Python data science library.

"RAPIDS, a GPU-accelerated data science platform, is a next-generation computational ecosystem powered by Apache Arrow," McKinney said. "NVIDIA's collaboration with Ursa Labs will accelerate the pace of innovation in the core Arrow libraries and help bring about major performance boosts in analytics and feature engineering workloads."

To facilitate broad adoption, NVIDIA is integrating RAPIDS into Apache Spark, the leading open-source framework for analytics and data science.

"At Databricks, we are excited about RAPIDS' potential to accelerate Apache Spark workloads," said Matei Zaharia, co-founder and chief technologist of Databricks, and original creator of Apache Spark. "We have multiple ongoing projects to integrate Spark better with native accelerators, including Apache Arrow support and GPU scheduling with Project Hydrogen. We believe that RAPIDS is an exciting new opportunity to scale our customers' data science and AI workloads."

Broad Ecosystem Support and Adoption
Tech-leading enterprises across a broad range of industries are early adopters of NVIDIA's GPU-acceleration platform and RAPIDS.

"NVIDIA's GPU-acceleration platform with RAPIDS software has immensely improved how we use data - enabling the most complex models to run at scale and deliver even more accurate forecasting," said Jeremy King, executive vice president and chief technology officer at Walmart. "RAPIDS has its roots in deep collaboration between NVIDIA's and Walmart's engineers, and we plan to build on this relationship."

Additionally, some of the world's leading technology companies are supporting RAPIDS through new systems, data science platforms and software solutions:

"HPE is committed to advancing the way customers live and work. Artificial intelligence, analytics and machine learning technology can play a critical role in uncovering insights that can help customers achieve breakthrough results and improve the world we live in. HPE is unique in the market in that we provide complete AI and data analytics solutions from strategic advisory to purpose-built GPU accelerator technology, operational support and a strong partner ecosystem to tailor the right solution for each customer. We are excited to partner with NVIDIA on RAPIDS to accelerate the application of data science and machine learning to help our customers drive faster and more insightful outcomes."
- Antonio Neri, CEO, Hewlett Packard Enterprise

"IBM has built the world's leading platform for enterprise AI, regardless of deployment model. We look forward to extending our successful partnership with NVIDIA, leveraging RAPIDS to provide new machine learning tools for our clients."
- Arvind Krishna, senior vice president of Hybrid Cloud and director of IBM Research

"The compute world today requires powerful processing to handle complex workloads like data science and analytics - it's a job for NVIDIA GPUs. RAPIDS is accelerating the speed at which this processing and machine learning training can be done. We are excited to support this new suite of open-source software natively on Oracle Cloud Infrastructure and look forward to working with NVIDIA to support RAPIDS across our platform, including the Oracle Data Science Cloud, to further accelerate our customers' end to-end data science workflows. RAPIDS software runs seamlessly on the Oracle Cloud, allowing customers to support their HPC, AI and data science needs, all while taking advantage of the portfolio of GPU instances available on Oracle Cloud Infrastructure."
- Clay Magouyrk, senior vice president of Software Development, Oracle Cloud Infrastructure

Support from other leading innovators - including Cisco, Dell EMC, Lenovo, NERSC, NetApp, Pure Storage, SAP and SAS, as well as a wide range of data science pioneers - is appended to this press release.

Availability
Access to the RAPIDS open-source suite of libraries is immediately available at http://rapids.ai/, where the code is being released under the Apache license. Containerized versions of RAPIDS will be available this week on the NVIDIA GPU Cloud container registry.
Add your own comment

18 Comments on NVIDIA Introduces RAPIDS Open-Source GPU-Acceleration Platform

#1
Dammeron
nV and open-source? I think hell just froze solid...
Posted on Reply
#2
DeathtoGnomes
Dammeron said:
nV and open-source? I think hell just froze solid...
its like 2 words you never ever use in the same sentence.
Posted on Reply
#3
chaosmassive
1. this is terrific news
2. anything open-sources and nvidia never mix well
3. anything utilize hardware that exclusive only for nvidia is NOT open source,
although in the article said that nvidia 'work' with open sources community closely. (more like to 'convert them' to nvidia ecosystem, in my 2c).
Posted on Reply
#4
DeathtoGnomes
chaosmassive said:
3. anything utilize hardware that exclusive only for nvidia is NOT open source,
its all a smokescreen, appearances are the goal here.
Posted on Reply
#5
notb
There's a discussion going on where you can find links to statistics showing that Nvidia is a larger contributor to open source than you think (for starters: larger than AMD). :)
Posted on Reply
#6
SetsunaFZero
Open-Source you say hmmm, RAPIDS Homepage says "It relies on NVIDIA® CUDA® " :rolleyes:
Posted on Reply
#7
95Viper
Let's stay on topic, no insulting others and, definitely, no them versus us type of off topic jabs.
Thank You
Posted on Reply
#8
TheGuruStud
notb said:
There's a discussion going on where you can find links to statistics showing that Nvidia is a larger contributor to open source than you bigots think (for starters: larger than AMD). :)
You mean like when they said low level APIs are useless (talking about Mantle), but as soon as Khronos Group was formed/developing Vulkan, which was clear to be the new, universal, API, then they immediately jump on board? LOL
They're liars through and through. They were terrified of AMD getting an upper hand with it, so they had to join. And that's all you need to know about how useless DX is, too.
Posted on Reply
#9
Ferrum Master
DeathtoGnomes said:
its all a smokescreen, appearances are the goal here.
Jensen didn't wear his leather jacket too on Volvo presentation...
Posted on Reply
#10
Recus
chaosmassive said:
1. this is terrific news
2. anything open-sources and nvidia never mix well
3. anything utilize hardware that exclusive only for nvidia is NOT open source,
although in the article said that nvidia 'work' with open sources community closely. (more like to 'convert them' to nvidia ecosystem, in my 2c).
SetsunaFZero said:
Open-Source you say hmmm, RAPIDS Homepage says "It relies on NVIDIA® CUDA® " :rolleyes:
Radeon Open Compute Platform (ROCm), the Most Versatile Open Source Platform for GPU Computing

Posted on Reply
#11
prtskg
Recus said:
Radeon Open Compute Platform (ROCm), the Most Versatile Open Source Platform for GPU Computing


CUDA is hardware specific. ROCm softwares use C++ and OpenCL, right?
Posted on Reply
#12
John Naylor
chaosmassive said:

3. anything utilize hardware that exclusive only for nvidia is NOT open source,
The corrolary then would be that any open source app that only runs in Windows is not open source.
Posted on Reply
#13
renz496
chaosmassive said:
1. this is terrific news
2. anything open-sources and nvidia never mix well
3. anything utilize hardware that exclusive only for nvidia is NOT open source,
although in the article said that nvidia 'work' with open sources community closely. (more like to 'convert them' to nvidia ecosystem, in my 2c).
1) it's a good news no doubt.
2) is that so? while nvidia seems anti open source to many they might contribute many more to open source than some other that keep preaching about "open". some things might not work well (like their open source driver for deskstop GPU) but others like their open source driver support for tegra has been very good. why did you think nintendo switch was easily hack? part of the reason was the chip is very well known due to their effort to make the chip spec and documentation is very accessible to open source community.
3) did nvidia restrict this RAPID platform specifically to use nvidia GPU only? or else you can't do anything about it?

prtskg said:
And below those lines, we have this -
CUDA is hardware specific. ROCm softwares use C++ and OpenCL, right?
CUDA is not hardware specific. it is more optimized for nvidia hardware but it can run on any hardware. Qualcomm for example made their snapdragon chip to be able to run nvidia cuDNN.
Posted on Reply
#14
notb
prtskg said:
And below those lines, we have this -
CUDA is hardware specific. ROCm softwares use C++ and OpenCL, right?
But Nvidia card users can program using both CUDA and ROCm. And OpenCL. And other libraries as well.
We choose CUDA, because it's simply easier and more effective.
Posted on Reply
#15
prtskg
notb said:
But Nvidia card users can program using both CUDA and ROCm. And OpenCL. And other libraries as well.
We choose CUDA, because it's simply easier and more effective.
CUDA was released in 2007. There was no competition and since then it has the market firmly in its hold.

renz496 said:

CUDA is not hardware specific. it is more optimized for nvidia hardware but it can run on any hardware. Qualcomm for example made their snapdragon chip to be able to run nvidia cuDNN.
Can you give links for this? Is Qualcomm doing it like AMD did with Boltzmann Initiative?
Posted on Reply
#16
notb
prtskg said:
CUDA was released in 2007. There was no competition and since then it has the market firmly in its hold.
Not true. There were other languages you could use even way before CUDA was born. I've coded in Brook in high school, few years before CUDA was launched.
The difference was: everything before CUDA was cumbersome and with steep learning curve. In many ways these were typical open-source projects - lacking someone taking care of esthetics and friendliness.

Keep in mind what was happening in the period we're talking about. We're in mid 2000s. Yesterday, coding was really just for programmers and some scientists (mathematicians, physicists etc). But everything is changing: suddenly coding is becoming a basic skill for economists, analysts and so on.
Nvidia was the first to notice that GPGPU could be useful for someone other than hardcore low-level programmers. Simple as that.
10 years later none of the more modern APIs offer similar combination of effectiveness and simplicity. And IT industry is really flexible. During those 10 years NoSQL eat a big portion of RDBMS market, Python became more popular than C++ et cetera. Nothing even touched CUDA's dominance.

The cost? You need Nvidia hardware. It's negligible.
Posted on Reply
#17
prtskg
notb said:
Not true. There were other languages you could use even way before CUDA was born. I've coded in Brook in high school, few years before CUDA was launched.
The difference was: everything before CUDA was cumbersome and with steep learning curve. In many ways these were typical open-source projects - lacking someone taking care of esthetics and friendliness.

Keep in mind what was happening in the period we're talking about. We're in mid 2000s. Yesterday, coding was really just for programmers and some scientists (mathematicians, physicists etc). But everything is changing: suddenly coding is becoming a basic skill for economists, analysts and so on.
Nvidia was the first to notice that GPGPU could be useful for someone other than hardcore low-level programmers. Simple as that.
10 years later none of the more modern APIs offer similar combination of effectiveness and simplicity. And IT industry is really flexible. During those 10 years NoSQL eat a big portion of RDBMS market, Python became more popular than C++ et cetera. Nothing even touched CUDA's dominance.

The cost? You need Nvidia hardware. It's negligible.
Wasn't I saying the same thing? CUDA had no competition when it was released. Some programs here and there doesn't mean competition.
Posted on Reply
#18
medi01
Dammeron said:
nV and open-source? I think hell just froze solid...
Well, why, CUDA is Open Source didn't you know?

Just a marketing spin, yet another one, nothing to see here.
Posted on Reply
Add your own comment