Home Business Analyst BA Agile Coach Introduction to NVIDIA AI Enterprise

Introduction to NVIDIA AI Enterprise

94
1

Artificial intelligence is no doubt one of the hottest topics in enterprise tech at the moment but one thing is preventing many companies from taking AI into use it’s super difficult nvidia is the world’s leading ai solutions provider and they’ve now taken up on this very challenge they want to

Democratize ai for all companies across all industries with their nvidia ai enterprise platform let’s talk about that hello my name is markus lenin i’m your enterprise tech enthusiast and this is where we discuss all about enterprise tech from independent point of view i’m happy to see you stopped by before we

Get started i want to thank the sponsors of today’s video here at paccard enterprise and nvidia so let’s take a quick look at the hardware and software stack that’s needed to run ai first we need capable hardware to run everything we can use physical servers on-prem or

Some public cloud services next we need a container platform of some sort because let’s face it virtual machines cannot offer the flexibility that’s needed with ai not even to mention bare metal last we need a load of ai software you can design configure and install all of

The above by yourself i emphasize it is possible but to successfully do so requires a lot of knowledge testing trial and error resources time and money and you can still not be sure you’ve chosen the best route to success also what happens when there’s a problem you have dozens of support phone numbers

To call and no one takes an overall comprehensive responsibility of the stack this is exactly where nvidia ai enterprise taps in let’s take a closer look first if you want to run your ai workloads on-prem you can choose servers from the list of nvidia certified servers that are tested and qualified by

Nvidia to run even the most demanding ai workloads as expected if you prefer to use public cloud nvidia has recommendations about suitable services from all major hyperscalers too next you will want to use some sophisticated container orchestration platform at the moment nvidia ai enterprise supports red hat openshift

Via vertanzu and different flavors of Kubernetes support for more container platforms are regularly added what comes to the software bit nvidia has a broad selection of ai tools frameworks and applications that nvidia has bundled and pre-configured for you to hit the ground running for instance there’s multi-instance gpus that allows for

Sharing one physical gpu with multiple Containers virtual machines and users for data preparation there’s nvidia rapids for training nvidia tau toolkit PyTorch and TensorFlow as for inferencing tensor rt and to deploy your ai project at scale nvidia triton inference server as a big juicy cherry

On a cake all of the above comes with one single support phone number to call in case of an issue this is a huge advantage especially with as complex environments as ai i repeat myself if you know what you’re doing you can technically stitch all of the above

Together by yourself but for most of us trusting our demanding and valuable ai projects with nvidia is definitely the way to go i’ve discussed before how hpe esmerald can help you hit the ground running with your ai project but even with the full force of hpe esmeral there

Are still some manual steps left to do for the customer and this is the gap that the nvidia ai enterprise can fail together with hpe esmeral nvidia ai enterprise can provide a complete and holistic ai solution from hardware to production lifecycle management if you want to take this to the next level you

Can get all of the above hpe esmeral nvidia ai enterprise and all the hardware via hpe green lake in which case you will get everything as a service hardware will be owned by hpe all installations configurations scaling and maintenance is taken care of by hpe nvidia and their partner networks all

You have to do is focus on using the ai environment and pay only for what you use that was all from me this time if you liked the video hit the like button and if you want more of these videos subscribe until the next videos
Artificial intelligence is no doubt one of the hottest topics in enterprise tech at the moment but one thing is preventing many companies from taking AI into use: it’s very DIFFICULT to implement!

NVIDIA is the worlds leading AI solutions provider and they’ve now taken up on this very challenge: they want to democratize AI for all companies across all industries with their NVIDIA AI Enterprise platform. Let’s talk about it!

*** DISCLAIMER! This video is sponsored by HPE and NVIDIA ***

======================
Table of Contents
======================
0:00 Introduction
0:46 General AI requirements
1:44 Introducing NVIDIA AI Enterprise
3:22 NVIDIA AI Enterprise + HPE Ezmeral
3:45 Even easier with HPE GreenLake

======================
Relevant resources
======================
Solving Real Business Problems with HPE Ezmeral:

======================
Music
======================
Epidemic Sound

======================
Business Enquiries
======================
Would you like me to review/introduce your cool data center solutions?
Contact me: markus[at]techenthusiast.com

======================
Follow me
======================
LinkedIn:
Twitter:
Website:
00:00 artificial intelligence is no doubt one
00:02 of the hottest topics in enterprise tech
00:04 at the moment but one thing is
00:05 preventing many companies from taking ai
00:07 into use it’s super difficult nvidia is
00:11 the world’s leading ai solutions
00:13 provider and they’ve now taken up on
00:15 this very challenge they want to
00:17 democratize ai for all companies across
00:20 all industries with their nvidia ai
00:22 enterprise platform let’s talk about
00:25 that
00:30 hello my name is markus lenin i’m your
00:32 enterprise tech enthusiast and this is
00:34 where we discuss all about enterprise
00:36 tech from independent point of view i’m
00:38 happy to see you stopped by before we
00:40 get started i want to thank the sponsors
00:42 of today’s video here at paccard
00:44 enterprise and nvidia so let’s take a
00:46 quick look at the hardware and software
00:48 stack that’s needed to run ai first we
00:51 need capable hardware to run everything
00:53 we can use physical servers on-prem or
00:55 some public cloud services next we need
00:58 a container platform of some sort
01:00 because let’s face it virtual machines
01:02 cannot offer the flexibility that’s
01:04 needed with ai not even to mention bare
01:06 metal
01:07 last we need a load of ai software you
01:11 can design configure and install all of
01:14 the above by yourself i emphasize it is
01:16 possible but to successfully do so
01:19 requires a lot of knowledge testing
01:22 trial and error resources time and money
01:25 and you can still
01:26 not be sure you’ve chosen the best route
01:28 to success
01:29 also what happens when there’s a problem
01:32 you have dozens of support phone numbers
01:34 to call and no one takes an overall
01:36 comprehensive responsibility of the
01:38 stack
01:39 this is exactly where nvidia ai
01:42 enterprise taps in let’s take a closer
01:44 look
01:45 first if you want to run your ai
01:46 workloads on-prem you can choose servers
01:48 from the list of nvidia certified
01:50 servers that are tested and qualified by
01:53 nvidia to run even the most demanding ai
01:55 workloads as expected if you prefer to
01:58 use public cloud nvidia has
02:00 recommendations about suitable services
02:02 from all major hyperscalers too
02:05 next you will want to use some
02:07 sophisticated container orchestration
02:09 platform at the moment nvidia ai
02:11 enterprise supports red hat openshift
02:14 via vertanzu and different flavors of
02:16 kubernetes support for more container
02:19 platforms are regularly added what comes
02:22 to the software bit nvidia has a broad
02:24 selection of ai tools frameworks and
02:26 applications that nvidia has bundled and
02:28 pre-configured for you to hit the ground
02:30 running for instance there’s
02:32 multi-instance gpus that allows for
02:34 sharing one physical gpu with multiple
02:37 containers virtual machines and users
02:40 for data preparation there’s nvidia
02:42 rapids for training nvidia tau toolkit
02:45 pytorch and tensorflow as for
02:47 inferencing tensor rt and to deploy your
02:51 ai project at scale nvidia triton
02:54 inference server as a big juicy cherry
02:57 on a cake all of the above comes with
03:00 one single support phone number to call
03:02 in case of an issue this is a huge
03:04 advantage especially with as complex
03:07 environments as ai i repeat myself if
03:10 you know what you’re doing you can
03:11 technically stitch all of the above
03:13 together by yourself but for most of us
03:16 trusting our demanding and valuable ai
03:19 projects with nvidia is definitely the
03:21 way to go i’ve discussed before how hpe
03:24 esmerald can help you hit the ground
03:25 running with your ai project but even
03:27 with the full force of hpe esmeral there
03:30 are still some manual steps left to do
03:32 for the customer and this is the gap
03:34 that the nvidia ai enterprise can fail
03:37 together with hpe esmeral nvidia ai
03:39 enterprise can provide a complete and
03:41 holistic ai solution from hardware to
03:44 production lifecycle management if you
03:46 want to take this to the next level you
03:48 can get all of the above hpe esmeral
03:51 nvidia ai enterprise and all the
03:53 hardware via hpe green lake in which
03:56 case you will get everything as a
03:57 service hardware will be owned by hpe
04:00 all installations configurations scaling
04:03 and maintenance is taken care of by hpe
04:06 nvidia and their partner networks all
04:09 you have to do is focus on using the ai
04:11 environment and pay only for what you
04:14 use that was all from me this time if
04:16 you liked the video hit the like button
04:18 and if you want more of these videos
04:20 subscribe until the next videos
04:23 [Music]

AISolutionsProvider

1 COMMENT

LEAVE A REPLY

Please enter your comment!
Please enter your name here