Ruben's EUC Industry Bento Box - December 2023

Industry News & Insights

1. ControlUp Community Radio with Legendary Former CEO, Citrix, Mark Templeton

   In this podcast, Douglas Brown, VP of Community, has a captivating conversation with Mark Templeton, former CEO of Citrix and current Chairman at WorkSpot and board of Directors at Nutanix. They delve into the history of virtual workplaces, highlighting Citrix's pioneering role during Mark's tenure and its influence on the industry. The discussion covers key moments like Citrix's early days, the Microsoft partnership, the introduction of Seamless Windows, and the evolving IT strategies, especially post-COVID. They also explore future trends influenced by major industry moves, such as acquisitions by Citrix and VMware, and innovative concepts like the Nirvana Phone by Chris Fleck.

2. Humane's Ai Pin

   Launched amidst much anticipation, Humane's Ai Pin is a compact device designed for "ambient computing." Attaching magnetically to a lapel, it features an onboard camera, microphone, depth, and motion sensors, and is powered by a Qualcomm chip, processing data with AI. This screenless device offers a unique interaction method, including a touchpad, gesture recognition, and voice control, communicating via a “personic speaker” or Bluetooth headphones.

   The Ai Pin operates independently of smartphones, supported by a Humane-branded wireless network. Its innovative Laser Ink Display projects information, like incoming calls, onto the user's palm. The device boasts a two-piece design with a main computer and a magnetic, wirelessly charging battery booster for flexible wear and all-day usage. Running on the proprietary Cosmos operating system with the Ai Bus software framework, Ai Pin eliminates the need for traditional apps, connecting users directly to AI-driven services. Collaborations with Microsoft and OpenAI enhance its capabilities, allowing it to operate as a standalone device. User data management is streamlined through the Humane center platform, offering a privacy-secured portal for personalized service setup and access. Check out this awesome video for more details: https://vimeo.com/882968794

   I believe this device does fit well in one of the Black Mirror episodes, pity it isn't available in Europe yet 😊

   

3. Generative AI for Beginners - an Easy-to-Use 12-course Runway to AI

   Topics include Introduction to GenAI, Exploring and Comparing different Large Language Modules, using GenAI responsibly, Prompt engineering fundamentals, text generation applications, designing UX for AI applications and much more. All content can be downloaded/watched from their Github repository.

   

4. Amazon Makes Waves with a New... Thin Client?

   Yes, welcome to 2024 😊 The AWS WorkSpaces Thin Client, a device available for purchase directly through Amazon.com, is priced at $195 in the United States. This device offers the capability to connect to a single monitor and various USB peripherals such as a keyboard, mouse, headset, microphone, and more. Additionally, there is an optional USB 3.0 hub with an extra HDMI port, facilitating the connection to multiple screens. Opting for this hub increases the total cost to $280. Device management is conducted through the AWS Management Console, which incurs an additional charge of $6 per device monthly. At this point, the device can only be used with AWS Workspaces and Appstream. Gabe Knuth wrote a great article about it, certainly worth reading! Jeff Barr also created an excellent step-by-step guide.

   Some questions are unanswered: Would it be possible in the future to support other DaaS solutions such as Citrix, Frame, Microsoft, and VMware as well? We will see!

5. Microsoft Copilot Studio

   Unveiled at Microsoft Ignite 2023, Microsoft's Copilot Studio represents a significant advancement in the customization and creation of AI-driven solutions within the Microsoft 365 ecosystem. As a low-code tool, it facilitates the tailoring of Microsoft Copilot for specific enterprise scenarios, empowering users to easily build, test, and deploy both custom GPTs and standalone copilots. The platform offers comprehensive lifecycle management, encompassing building, deployment, analysis, and ongoing management within a unified web interface. Furthermore, Copilot Studio is integrated with Microsoft's broader AI and conversational technologies like Azure OpenAI Studio and Azure Cognitive Services, ensuring a robust and versatile development environment. Its availability is already established, with the integration with Copilot for Microsoft 365 accessible in public preview.

For the Geeks

1. NVIDIA LaunchPad

   NVIDIA's Launchpad provides free access to enterprise NVIDIA hardware and software through an internet browser. Users can experience the power of AI with end-to-end solutions through guided hands-on labs or as a development sandbox. Test, prototype, and deploy your own applications and models against the latest and greatest that NVIDIA has to offer. For instance, Audio Transcription, Annotate, and Adapt Medical Imaging Models with MONAI, Accelerated 3D Graphics with NVIDIA RTX Virtual Workstation, Bring Your Workload to NVIDIA L40S, AI Chatbot with Retrieval Augmented Generation and much more! Wouldn't it be great to run NVIDIA RTX Virtual Workstation on Azure, AWS, and GCP using Frame? - if you want to know how, just let us know so we can sign you up!

   Want to know more about Frame test-drive, trial and showcase? This 3-minute Demo video explains it all.

   

2. MSIX and App Attach Made Easy presented by Microsoft's Jim Moyle, Senior PM and Sushant Bansal, Product Manager.

   They show and tell MSIX App Attach enhancements, the importance of MSIX, the creation of packages, MSIX enhancements and answer common questions. As a bonus, they also talk about App-V to App Attach migration.

3. Inside Microsoft AI Innovations with Mark Russinovich

   At Microsoft Ignite 2023, Azure CTO Mark Russinovich presented Microsoft's latest advancements in AI and cloud technology. He emphasized sustainable and efficient data center developments, including the use of hydrogen fuel cells for backup power and advancements in server cooling technologies, particularly focusing on microfluidics cooling for better thermal management.

   Russinovich highlighted Microsoft's AI supercomputers, designed for OpenAI, and their high-performance capabilities facilitated by InfiniBand networks. He introduced Project Forge, a resource management system for AI workloads, featuring a global scheduler and a reliability system with transparent checkpointing for optimizing resource usage and handling failures.

   The presentation also covered innovations in AI model training and serving, including low-rank adaptive fine-tuning (LoRA) for efficient model training and a system for serving multiple AI models on GPUs simultaneously. Additionally, he demonstrated TypeChat for structured AI outputs and discussed AI Research's efforts in developing smaller, more efficient language models and techniques for model unlearning.

   The session concluded with a showcase of Azure Pong, demonstrating GPT-4's application in a gaming scenario, underscoring Microsoft's ongoing innovation in the realms of AI and cloud computing. As always a great technical session to watch!

   Ever seen 1,792 CPUs running in Windows Task Manager? Cool demo at 01:01:30 😊

   

4. Frame 3-min Demo Videos

   As the name suggests, these short videos present specific Frame-related technical topics such as "Preparing your Azure subscription for integration with Frame", "Behind the Scenes with Azure Resource Groups", "Configuring Microsoft Entra ID as a SAML IdP for Frame", "Installing and onboarding new applications," and many more.

   The first 14 videos are available to watch and much more to come! Subscribe and enjoy.

   

5. Microsoft is finally making custom chips — and they're all about AI

   Microsoft has unveiled its own custom AI and cloud hardware: the Azure Maia AI chip and the Azure Cobalt CPU, set to launch in 2024. The Maia chip, tailored for AI workloads like large language model training, is built on a 5-nanometer process with 105 billion transistors and supports sub 8-bit data types for efficient model training. The Cobalt CPU, a 128-core Arm-based processor, is optimized for Azure's general cloud services, boasting up to 40% improved performance over current commercial Arm servers. Both chips, integral to Microsoft's strategy to enhance AI cloud infrastructure, feature innovative designs like Maia's complete liquid cooling system for server efficiency. This development represents a strategic move to diversify Microsoft's cloud hardware, reducing reliance on external suppliers like Nvidia and enhancing Azure's AI capabilities. Microsoft emphasizes that these developments are not in competition with existing partners like Nvidia and AMD but are complementary, aiming to diversify supply chains and offer customers more infrastructure choices. This diversification is crucial, especially considering the heavy reliance on Nvidia for AI server chips currently.

   At the end it is great to have competition, it is healthy and good for customers and partners.

   

6. Bonus Video: "Fuzzy Feelings"

   Please, take a moment to watch this heartwarming short video. Happy Holidays and Merry Christmas!