Did you know AI is an ideal workload to run in an EUC environment? That’s what I cover in this VMUG EUC Day session, Virtual Desktops, GPUs, and Things You Didn’t Know You Could Do. We explore how AI instances benefit from virtual desktop environments and why organizations who are developing AI workloads should consider running their AI development environment with EUC functionality.
This session is part of the VMUG EUC day being offered today. You can find out more and check out all the great content at on the VMUG website. There will be many great presenters that are well versed in all areas of EUC. I recommend watching it live or in replays if your day is already full.
After EUC Day is over I’ll post the session recording and you can watch the session at your leisure. Should you have questions following the event you can use the contact page to reach out with them or any of the other methods I mentioned in the session.
For those wondering why you want to do AI in a EUC environment, it boils down to the ability to have a known powerful set of management tools for you workload and having the ability to deliver that environment both repeatably and quickly for AI developers. Additionally it makes it easier for developers to consume AI resources like they do any other environment. This is a win – win for almost all organizations and unlocks the synergies of AI and IT.
If you had the chance to join us live for this community session at VMware Explore for session CMTY2254LV, thank you so much. If you are checking it after the event thank you as well! Your support means the world to us. This page contains links to the videos, the session abstract and a bunch of information for the session we submitted. If you have any questions or need additional details about unlocking the magic of GenAI in VMware VCF, please contact me.
The abstract for this community session is:
The abstract for the session is: “Join us for an electrifying session where we decode the AI buzz and reveal its practical magic! Visionary leader Gina Rosenthal and renowned vExpert Tony Foster will guide you through the AI labyrinth without the jargon overload. Explore the inner workings of large language models (LLMs), demystify training and inferencing, and discover the secret sauce of retrieval-augmented generation (RAG). Learn how Gen-AI differs from traditional AI and how it can supercharge your business strategy, especially within the context of VMware VCF. We’ll delve into the intricacies of integrating Gen-AI seamlessly into your existing IT infrastructure, unlocking untapped potential. Get ready to leave this session with a data scientist’s swagger, poised to infuse AI brilliance into your organization! Don’t miss out!”
The session is available from VMware TV on YouTube under the live section:
This session really is about demystifying all the stuff that’s happening behind the scenes when it comes to AI. In short there is nothing magic or crazy happening, its just a computer program doing a big math problem. We hope you enjoyed the session or the replay!
This is the 2024 VMware Explore session that John and I presented on Quantum Computing. It was a well attended session for the last day and last time slot of the conference. I’m sharing the details of the session with those interested here.
Session Abstract: The advent of quantum computing heralds a new era of computational power, promising unprecedented capabilities that could revolutionize industries and scientific research. However, along with its potential benefits comes a looming specter of uncertainty and fear. In this presentation, we delve into the implications of quantum computing, exploring the potential negative disruptions it could bring. One of the primary concerns is quantum hacking, where conventional encryption methods become vulnerable to quantum algorithms. As companies race to adapt their security technologies to fend off these threats, it begs the question: Are we adequately prepared for the quantum era? Join us as we navigate through the risks and challenges posed by quantum computing and contemplate strategies to mitigate its potential adverse impacts.
If you would like to view the slides presented a PDF copy we are pleased to share them with you. Should you have any questions about the content presented please feel free to use my contact page to reach out to us.
You got approval to attend VMware Explore this year! That’s fantastic, I look forward to seeing you there. By now you have probably started to build your schedule in session builder. I have a few Explore sessions you might want at the top of your list. And some of you might be wondering what AI Qubits are.
As of today, they don’t exist, they aren’t real. They sound cool though, don’t they? I’ll be presenting on topics related to both qubits and AI this year at Explore. We’ll be breaking down topics to their atomic parts and spinning them to get the answers people are curious about.
I bet that means you’re curious about what all these sessions are.
There are four of them, and they range from introductory to deep dives and futuristic exploration of topics. The four sessions you will want to attend are:
Needless to say, you should be able to find me fairly easily Monday and Thursday (hint check the community space).
You are probably wondering about all these sessions and they’re making your head spin faster than a couple of entangled qubits.
Let’s start with the breakout session on quantum computing. John Arrasjid (VCDX001) and I are talking about the perceived perils of quantum computing. You may have even heard that quantum computing is going to break encryption sometime “really soon” and it’s going to be pandemonium. We’re going to demystify all this and talk about what it will take to achieve this, where we’re at currently, and what IT teams can start doing today to protect themselves from the big scary quantum computers whose qubits are smaller than the tip of a pin. Never have so many big companies and governments been scared of something so small, welcome to the future.
Speaking of the future let’s look at the session on unlocking the magic of Gen-AI. This isn’t a deep technical community session that will teach you to deploy an AI model on VMware VCF in under 30 minutes. That would be pretty cool, and I think it can be done, but that’s not what we are talking about in this session. Gina Rosenthal and I are going to break apart all the hype around AI and make it easier for folks to understand from how does AI and LLMs even work to what is RAG and why we need it. We also answer the question why you would want to run AI in a VCF virtual environment. This is going to be a fun session that allows almost anyone to sound good in a room and know what they’re talking about.
Now it’s time to dive to the depths of things you can do with AI in my VMware Code session on taking code and turning it into a vCenter plugin. That’s exactly what I will demonstrate in this 25-minute Code session. Here a few years back I created some simple but powerful scripts that most people know as “VDI by day and Compute by Night.” The idea is simple you have all these users on virtual desktops (AKA Omnissa Horizon) with Virtual GPUs (vGPUs). Well, most users only use those desktops for about 1/3 of the day the rest of the time those resources are idle. My scripts allow you to recover those unused GPU resources and run your AI or other GPU intensive workloads on the idle GPUs.
I’ve converted those scripts into python and a REST API, during my code session, we are going to take that API and use it as the base for a vSphere plugin. We’ll explore the underlying structure of a plugin and all the things that need to exist to create one. Then we will do exactly that. Take my code and turn it into a functional plugin allowing you to reclaim vGPU resources in your environment.
That brings me to the last session I’ll be presenting at Explore. I’ll show you how to install my resource recovery plugin in your own test/dev environment along with what you can do with it. It’s got some cool features not found in the scripts and it should make administering vGPU environments a lot nicer. I’d share some of the features, but they’re not all written yet, so you’ll have to wait for Explore.
As a bonus, you will also want to attend Jodi Shely and John Arrasjid’s session on Resilient Infrastructures: Self-Healing with VCF, Aria & VMware Ecosystem [VCFB1444LV]. I’ve had a preview of the deck for this session, and this is one of the holy grails of IT, self-healing environments. If you manage an enterprise IT environment you’ll want to be in the audience and see how the concept of resilient infrastructures can help you move beyond keeping the lights on.
That’s a lot of content on the buzz worthy topics of AI and quantum computing. If you are attending explore and are even thinking about AI or quantum computing these are some of the sessions, you cannot afford to miss. They will help provide clarity on how both are continuing to grow and transform the datacenter and improve operations.
I challenge you to plunge your hands into the quantum foam at VMworld this year, seize some qubits, and elevate your knowledge.
Thank you to everyone who was able to attend our session at VMware Explore in Las Vegas. The VMTN vBrownBag Tech Talk, VMTN3079LV, John Arrasjid and I presented on the Quantum Leap: The Ever‐Changing Virtual Space of Quantum Computing had great attendance. For those who want to watch it again or weren’t able to attend you can find both the slides and the recording below.
It’s important to reiterate that Quantum Computing is still an emerging technology and the capabilities it will provide in the future are still anyone’s guess. What we present in this session are just hypothesis about what could be possible. We have presented them here as a way to start the conversation and get folks thinking about what could be. The exciting time in the quantum fields are still to come.
Some great references on quantum computing can be found here:
Thank you to everyone who was able to join my session in person at VMware Explore in Las Vegas. I was honored to present the vBrownbag Tech Talk on “The SEO Strategies for My Community Blog That Gave Me Six-Pack Abs.” I am really excite that I had more that six people in the audience for one of the last sessions of the conference. In this post you can find my slides, the recording of the Tech Talk and some discussion of material that those in attendance asked following the session.
Below is the video recording of the session:
One of the questions following the presentation that came up was about AI generated content. The initial question that was asked was about the use of AI generated content for blogs. My answer for the foreseeable future is honesty is the best policy, you should disclose on both personal and commercial blogs that the certain content was AI generated.
Obviously, SEO systems are actively looking to determine if content is AI generated. They are or will be using the same (or better) tools as others to detect AI generated content. And if you aren’t upfront about it, they may eventually start dinging your SEO score. So you need to be honest about it, and the sooner the better.
In my personal opinion, there is nothing wrong with using AI generated content. If you don’t learn to do it someone else will. I haven’t done it on any of the blogs I write, as I’m waiting for the legal and business stuff to all shake out. Also, a lot of stuff I blog on wouldn’t be known by AI yet because its either a hair brained idea or it’s internal information.
Which brings up the last interesting point from the follow on discussions about my session. Everyone is trying to detect AI generated content for a lot of reasons. One is around not deceiving the reader, which we talked about above. Another major factor for businesses and government agencies is the disclosure of non-public information, be it outright inclusion or inclusion in the AI model. Both can be damaging and this is why so many vendors are starting to offer private models for organizations, so that corporate users can get the same ChatGPT experience but not leak sensitive information. Which, again is another good reason to disclose something is AI generated upfront.
Permanent link to this article: https://www.wondernerd.net/vmware-explore-vmtn3082lv-the-seo-strategies-for-my-community-blog-that-gave-me-six-pack-abs/
Thank you to everyone who joined me for VMTN3081LV at VMware Explore. It was great having the opportunity to present “AI on the Horizon: Delivering Virtualized AI Environments to Students.”
This session started as a submission as a full session, then was whittled down to a 26 minute vBrownbag Tech Talk, and finally squeezed in as 12ish minute vBrownbag Tech Talk. Which I’m sure you can guess means, each round more and more information got taken out. At some point I will record the full hour long version with a live demo, until then This is the goodness that is my vBrownbag.
You can watch the presentation here:
You can also download a copy of my slides for VMTN3081LV.
If you would like to find out more about this design, please use the contact me page.
Permanent link to this article: https://www.wondernerd.net/vmware-explore-vmtn3081lv-ai-on-the-horizon-delivering-virtualized-ai-environments-to-students/
One of the things I have wanted to do for a while is to post the abstracts I submit for various conferences. This is my first attempt at doing just that. These are the abstracts I can publish, that I submitted for NVIDIA GTC 2023 (spring). There is one abstract I can’t share with folks because it’s closely related to a project I’m working on. It’s also worth noting that none of these abstracts were accepted for the conference this year. However if you have a user group or a conference where you would like to hear about these topics please use the contact page and let me know.
Session Title: How to deliver Red Hat OpenShift with NVIDIA GPU capabilities on bare metal hyper-converged infrastructure
Session Abstract: Your organization uses Red Hat OpenShift for its workloads from containers to VMs. The only thing left to do… get your NVIDIA GPUs up and running in your OpenShift environment. Oh yeah, it’s on bare-metal, and you need to support multi-instance GPUs (MIG). This might sound daunting to accomplish. First you must install the NFD Operator, then the NVIDIA GPU operator, and finally you’ve got to configure MIG. If all that seems like a foreign language don’t worry, Tony and Praphul will guide you through the process step by step. They will demonstrate step by step how to bring NVIDIA GPUs into your OpenShift environment. They will even take it one step further and do it on a Dell PowerFlex hyper converged bare metal environment. If you’re using Red Hat OpenShift this is a session you won’t want to miss.
Session Title: Your MLOps desktop delivered virtually, the details behind the magic
Session Abstract: MLOps, it’s been a big buzzword for a few years. Everyone is trying to hire MLOps engineers from the financial institutions, to retailers, to the agriculture industry. That comes with some big issues for IT, like how do we get someone working from home a powerful MLOps capable work environment? And even more importantly how do we do it at scale? In this session join the WonderNerd for a deep dive into the magic of delivering Linux based virtual desktops for MLOps users. We will look at how to build Linux VDI environments using VMware vSphere and VMware Horizon. This session will go well beyond traditional VDI use cases, covering multi-instance GPUs (MIG) and designing virtual desktops with multiple vGPUs. We’ll then look at how these virtual desktops can be integrated with an existing MLOps strategy to scale as more engineers join the organization. We’ll even look at some examples of how all this can be easily automated as well as ways to reclaim resources at the end of a project.
Permanent link to this article: https://www.wondernerd.net/2023-gtc-submission-abstracts-i-just-had-to-share/
You’re in the right place if you are looking for resources from VMworld Code Session 2778 – Talk Nerdy to Me, Using Python to Create VMs with vGPUs for AI Workloads. Thank you to everyone who was able to join the session live. For those who weren’t able to attend the live session, you can catch the replay of it on VMworld.com.
Below you will find the slide deck of material I had for the session, it includes the code snipits I used during the presentation. You can also find the fully functional blocks of code up on GitHub in my VMworld21 repo, you’ll just need to update credentials/vcenter.
Click to download the CODE2778 Talk Nerdy to Me Slide Deck
This is a short 30 minute session and there’s a lot to cover when working with GPUs and VMs. Hopefully this session covers the material you need to get started in automating the creation of VMs with GPUs for AI/ML/DL/HPC workloads. Because everyone’s situation is different it’s very difficult to create a script that sorta fits for everyone.
UPDATE: The VMware {code} team posted the videos up to you tube you can watch my session below.
The Code
The way I’ve set the table for this session is starting out with a connection to the vCenter (I know it seems simple). From there we capture the GPUs in a host as well as what vGPU profiles those GPUs/hosts support. Next we add a vGPU to a VM programmatically and then remove it. Once we’ve worked through those basics we bring it all together and create a VM with a vGPU.
Now lets break down some of those parts as to why they are important. First of all it’s important to know that we have GPUs in the hosts we are working with. By using the first script to detect GPUs in the host we can do a safety check on our hosts to make sure we are able to do what we want. There are also many other things we can do with that information.
Next when creating an AI VM, you need to know what profiles are available to consume in that VM. It’s better to do this programmatically than to manually assign profiles. This is important especially when you have multiple GPUs in your environment and you need the right one for the right VM. Malformed VMs won’t start and that leads to service problems and user dissatisfaction. This is why knowing the vGPU profiles are important.
Now we get into the the actual meat of this, we add a vGPU to an existing VM. This may be all some folks need to do, they may already have VMs created and just need a vGPU profile to it. It’s really the foundational process to get a vGPU into a VM weather its at creation or later on.
Next we perform some clean up of the VM. This isn’t really necessary for VMs, it’s just helpful to know how to do it. This set of code modifies the VM and removes the last vGPU in the VM. If you want to be a bit more specific about which vGPU you want to remove from the VM, you’ll need to be a bit more specific in the for loop to pick the one you want to remove.
Lastly we bring this together and create a VM programmatically. This script is just a slight modification of the pyVmomi Community Samples to create a VM. You’ll see we took a bit of the code from the add a vGPU to a VM and pasted into the create a VM code sampe. That adds the vGPU to the VM when we create it. Now we just run the code using a quick script.
All of that fills up a 30 minute session really quickly. Especially with questions.
What to do with this?
Those bits of code are all you need to start creating VMs for AI/ML workloads. You need to incorporate them into your automation processes.
I didn’t deploy an OS or install the vGPU drivers during this session, that really outside the virtual hardware side of things, its the same reason I don’t deploy an app in this session. Additionally I’ve found that AI admins get really particular how these things get setup and would get very cranky about an outsider like me saying how it should be done.
A few other things you should know. First you need to use a vGPU profile that consumes the entire GPU if you want to have multiple vGPUs on a VM. Partial vGPU profiles will not work. So in my demo with the P4 GPU I need to use the P4-8Q or P4-8C profile for it to be able to power up. The other part of that is the profiles have to be the same for all vGPUs in the VM, so if you have an P4-8C and a T4-8C those can’t be on the same VM. They all have to be the same.
Hopefully this content is helpful content. If you have questions, or want to see more drop a comment at the end of the blog.
Now for some fun stuff
I’d like to thank the VMware {code} team for all their help promoting this session. They did a fantastic job with the promo video for this. You get to see me in my cape and WonderNerd glasses.
https://twitter.com/i/status/1443266825123340288
I did another version of a video tease for this. It was decided it may be a bit risque and is almost not safe for work. If you would like to watch it send me a DM.
I hope you enjoyed VMworld!
May your servers keep running and your data center always be chilled.
This is a quick post for my VMworld 2021 vBrownbag session VMTN2835. This session covers updates to the VDI by day compute by night scripts that I’ve been writing and updating for a couple of years. You can download all of the modules needed for this from my GitHub repo. The session is airs at 2:15 PM Central Time on October 5th.
You can download the slide deck here or by clicking on he picture above. The demo of how the scripts work is part of the session. To watch the session, please sign up for VMworld and you can catch it on replay if you missed the live session.
UPDATE: A little while after VMworld the vBrownbag team posts the videos up to their YouTube channel. You can now watch the video below.
Some may ask why this collection of modules isn’t bundled up as a single set of scripts. I’ve done it this way because they can be used independently of the rest of the scripts. For example you may want to gather profile information about the GPUs in your hosts, or you may want to determine the carrying capacity of your hosts. Both of these can be done done with the scripts independently and don’t require the entire VDI by day set of scripts.
There are a several changes to the script that I talk about in the session. To quickly recap those for you, they are:
Support for the NVIDIA Ampere line of GPUs – A100, A40, and A10, the A16 will follow shortly after GA
A lot of new parameter updates to be compatible with vSphere 7.0 U2
The view commands no longer support multiple hosts states
The default host states have been changed to “Connected” only
Work has been done to change from using the get-VMhost command
Looking at that list you can probably guess the next two items that will be changing soon for these scripts. First we will be changing to a get-View command to speed things up, especially for larger environments. The other one is to support the A16 GPU when it GAs. Though it is worth noting that, it’s not a good candidate for VDI by day compute by night because it’s optimized more for VDI than AI/ML. This is not unlike the A100 & A30 which can’t be used to drive graphics in VDI. The optimum card for most will be the A40.
For the rest of whats been updated be sure to catch VMTN2835. I hope you have a great VMworld, any feedback on these modules is greatly appreciated.
May your servers keep running and your data center always be chilled.
Privacy & Cookies: This site uses cookies. By continuing to use this website, you agree to their use.
To find out more, including how to control cookies, see here:
Cookie Policy
Statements
The fine print: The views expressed herein are solely those of the author and contributing data feeds, these do not reflect or represent the views of employers, associates, associated business entities, groups, organizations, or any other entity (listed or otherwise). Further more some material on this website is dynamically generated and the author does not control its content and is not liable for said content. Content is property of its respective owners. All rights reserved. All material is provided as is, where is, with no expressed or implied warranty of fitness, usability, or merchantability.
In the interest of full disclosure, I am an employee of Dell EMC. The views expressed herein do not represent the views of Dell Technologies or its subsidiary companies.
![VMware Explore Session Title Slide: [CMTY2254LV] Unlocking the Magic of Gen-AI in VMware VCF: Where Dreams Meet Reality](https://i0.wp.com/www.wondernerd.net/wp-content/uploads/2024/09/image.png?ssl=1)
![Catch these sessions at VMware Explore:
• Quantum Computing: What Are You Scared Of? [INVB1447LV] on Thursday, August 29th at 10:15 AM
• Unlocking the Magic of Gen-AI in VMware VCF: Where Dreams Meet Reality [CMTY2254LV] on Thursday, Aug 29th at 9AM
• Unlocking vGPU Power: From Code to vCenter Plugin [CODE2229LV] on Monday, August 26th, at 11AM
• Empowering Virtual Environments: The vGPU Resource Scheduler Plug-in [CMTY2241LV] on Monday, August 26th at 1PM](https://i0.wp.com/www.wondernerd.net/wp-content/uploads/2024/07/Explore_Catch_Sessions.png?ssl=1)
![Quantum Computing: What Are You Scared Of? [INVB1447LV] on Thursday, August 29th at 10:15 AM](https://i0.wp.com/www.wondernerd.net/wp-content/uploads/2024/07/CODE2229LV_Promo_Card-1.png?ssl=1)
![The SEO Strategies for My Community Blog That Gave Me Six-Pack Abs [VMTN3082LV] Cover slide for VMware Explore 2023 vBrownBag Tech Talk.](https://i0.wp.com/www.wondernerd.net/wp-content/uploads/2023/08/VMTN3082LV-SEO_Stratagies-Cover.png?resize=750%2C422&ssl=1)
