Over the past year, generative AI has transformed the way people live, work and play, enhancing everything from writing and content creation to gaming, learning and productivity. PC enthusiasts and developers are leading the charge in pushing the boundaries of this groundbreaking technology. Countless times, industry-defining technological breakthroughs have been invented in one place—a garage. This week marks the start of the RTX AI Garage series, which will offer routine content for developers and enthusiasts looking to learn more about NVIDIA NIM microservices and AI Blueprints, and how to build AI agents, creative workflow, digital human, productivity apps and more on AI PCs. Welcome to the RTX AI Garage.

This first installment spotlights announcements made earlier this week at CES, including new AI foundation models available on NVIDIA RTX AI PCs that take digital humans, content creation, productivity and development to the next level. These models—offered as NVIDIA NIM microservices—are powered by new GeForce RTX 50 Series GPUs. Built on the NVIDIA Blackwell architecture, RTX 50 Series GPUs deliver up to 3,352 trillion AI operations per second of performance, 32 GB of VRAM and feature FP4 compute, doubling AI inference performance and enabling generative AI to run locally with a smaller memory footprint.

In October, we reported that Asahi Linux had made some pretty substantial progress in getting Linux working on Apple Silicon, with a custom GPU driver counting itself as the first OpenGL 3.0 compliant graphics driver for Apple Silicon. Now, according to a recent Fedora Magazine post, Asahi Linux now has a Fedora 41 Remix. The Fedora Asahi 41 Remix is still based on that same Asahi version from October 10, but it incorporates the myriad advancements from Fedora 41 and an improved Vulkan 1.4-conformant driver.

By default, Fedora Asahi Remix 41 ships with Plasma 6.2, although a GNOME 47 version is also available. However, despite using a Wayland-first dekstop environment as its default, Fedora Asahi 41 Remix will still be X11-first, for compatibility reasons. The Asahi team plans on getting Wayland working eventually, but there are technical hurdles to overcome before that is possible. Fedora Asahi Remix also ships by default with the improved DNF 5 package manager and the massively improved GIMP 3.0 image editor pre-installed as part of Plasma 6.2. Having a Fedora Remix for Asahi also provides a familiar experience for gamers on Apple Silicon Macs to get Windows games up and running via a mix of tools like Valve's Proton dxvk, FEX emulator, and vk3d-proton—check out our previous coverage of Asahi to find out more about which games are currently working. There are still a handful of hiccups when it comes to running Linux on Apple Silicon, including hardware incompatibilities, like a lack of Thunderbolt, microphone, Touch ID, and USB-C Display support.

NVIDIA announced today it is opening its first Vietnam research and development center, signaling its confidence in the country's bright artificial intelligence future. The company is collaborating with the Vietnamese government to establish its new Vietnam Research and Development Center focused on AI. NVIDIA will use the R&D center to focus on software development, capitalizing on the country's strong talent pool of STEM engineers, and to engage industry leaders, startups, government agencies, universities and students to accelerate the adoption of AI.

"We are delighted to open NVIDIA's R&D center to accelerate Vietnam's AI journey," said Jensen Huang, founder and CEO of NVIDIA. "With our expertise in AI development, we will partner with a vibrant ecosystem of researchers, startups and enterprise organizations to build incredible AI right here in Vietnam."

At AWS re:Invent, Amazon Web Services, Inc. (AWS), an Amazon.com, Inc. company, today announced the general availability of AWS Trainium2-powered Amazon Elastic Compute Cloud (Amazon EC2) instances, introduced new Trn2 UltraServers, enabling customers to train and deploy today's latest AI models as well as future large language models (LLM) and foundation models (FM) with exceptional levels of performance and cost efficiency, and unveiled next-generation Trainium3 chips.

"Trainium2 is purpose built to support the largest, most cutting-edge generative AI workloads, for both training and inference, and to deliver the best price performance on AWS," said David Brown, vice president of Compute and Networking at AWS. "With models approaching trillions of parameters, we understand customers also need a novel approach to train and run these massive workloads. New Trn2 UltraServers offer the fastest training and inference performance on AWS and help organizations of all sizes to train and deploy the world's largest models faster and at a lower cost."

NVIDIA today announced major advancements to Universal Scene Description, or OpenUSD, that will expand adoption of the universal 3D data interchange framework to robotics, industrial design and engineering, and accelerate developers' abilities to build highly accurate virtual worlds for the next evolution of AI.

Through new OpenUSD-based generative AI and NVIDIA-accelerated development frameworks built on the NVIDIA Omniverse platform, more industries can now develop applications for visualizing industrial design and engineering projects, and for simulating environments to build the next wave of physical AI and robots.

AAEON, a leading designer and manufacturer of AI Edge solutions, has today confirmed that its Fanless Embedded AI System, the BOXER-8640AI, will support the newly released NVIDIA Jetpack 6.0 software development kit.

The NVIDIA Jetpack 6.0 SDK is equipped with a range of new features conducive to accelerating AI application development, such as Jetson Platform Services, a modular, API-driven suite of pre-built and cloud-native software services.

Intel's upcoming Falcon Shores GPU is shaping up to be a powerhouse for AI and high-performance computing (HPC) workloads, but it will also be an extreme power hog. The processor, combining Gaudi and Ponte Vecchio successors into a single GPU, is expected to consume an astonishing 1500 W of power - more than even Nvidia's beefy B200 accelerator, which draws 1000 W. This immense power consumption will require advanced cooling solutions to ensure the Falcon Shores GPU operates efficiently and safely. Intel's partners may turn to liquid cooling or even full immersion liquid cooling, a technology Intel has been promoting for power-hungry data center hardware. The high power draw is the cost of the Falcon Shores GPU's formidable performance promises. Intel claims it will deliver 5x higher performance per watt and 5x more memory capacity and bandwidth compared to its Ponte Vecchio products.

Intel may need to develop proprietary hardware modules or a new Open Accelerator Module (OAM) spec to support such extreme power levels, as the current OAM 2.0 tops out around 1000 W. Slated for release in 2025, the Falcon Shores GPU will be Intel's GPU IP based on its next-gen Xe graphics architecture. It aims to be a major player in the AI accelerator market, backed by Intel's robust oneAPI software development ecosystem. While the 1500 W power consumption is sure to raise eyebrows, Intel is betting that the Falcon Shores GPU's supposedly impressive performance will make it an enticing option for AI and HPC customers willing to invest in robust cooling infrastructure. The ultra-high-end accelerator market is heating up, and the HPC accelerator market needs a Ponte Vecchio successor.

Ben Skeggs, a lead maintainer of the open-source NVIDIA GPU driver for Linux kernel called Nouveau, has joined NVIDIA. Working as an open-source contributor for the Nouveau driver for more than a decade, Ben Skeggs has achieved the remarkable feat of working to support the NVIDIA GPU hardware on open-source drivers. Before joining NVIDIA, Ben Skeggs worked at Red Hat up until September 18th of 2023. At that date, he posted that he was resigning from Red Hat and stepping back from the Nouveau open-source driver development. However, this news today comes as a bit of an interesting development, as Ben Skeggs is going to NVIDIA, which has been reluctant in the past to support open-source drivers.

Now, he is able to continue working on the driver directly from NVIDIA. He posted a set of 156 patches to the driver, affecting tens of thousands of lines of code. And he signed it all off from the official NVIDIA work address. This signals a potential turn in NVIDIA's approach to open-source software development, where the company might pay more attention to the movement and potentially hire more developers to support these projects. Back in 2012, NVIDIA had a different stance on open-source development, infamously provoking the creator of the Linux kernel, Linus Torvalds, to issue some snide remarks to the company. Hopefully, better days are ahead for the OSS world of driver development and collaboration with tech giants.

According to ChinaScope, an effort is currently underway to strengthen Huawei's HarmonyOS platform's presence. The local government of Shenzhen has unveiled an ambitious program aimed at supercharging the development of native applications for the operating system. The "Shenzhen Action Plan for Supporting the Development of Native HarmonyOS Open Source Applications in 2024" outlines several key goals to foster a more robust and competitive ecosystem around HarmonyOS. One primary objective is for Shenzhen-based HarmonyOS apps to account for over 10% of China's total by the end of 2024. To facilitate this, the city plans to establish at least two specialized industrial parks dedicated to HarmonyOS software development across various application domains.

Furthermore, the initiative calls for over 1,000 software companies in Shenzhen to obtain HarmonyOS development talent qualifications, underscoring the city's commitment to cultivating a skilled workforce for the platform. Perhaps most impressively, the action plan encourages eligible companies to ramp up their outsourcing services for HarmonyOS app development, with a lofty target of reaching 500,000 HarmonyOS developers. This would represent a significant influx of developer talent focused on the platform if achieved. The Shenzhen government's push aligns with China's broader strategy to reduce reliance on foreign technologies and promote the adoption of domestic alternatives like HarmonyOS. While initially launched by Huawei as a workaround for U.S. sanctions, HarmonyOS has since expanded to power many devices, including smartphones, tablets, smartwatches, and TVs.

The latest version of AMD's open-source GPU compute stack, ROCm, is due for launch soon according to a Phoronix article—chief author, Michael Larabel, has been poring over Team Red's public GitHub repositories over the past couple of days. AMD ROCm version 6.0 was released last December—bringing official support for the AMD Instinct MI300A/MI300X, alongside PyTorch improvements, expanded AI libraries, and many other upgrades and optimizations. The v6.0 milestone placed Team Red in a more competitive position next to NVIDIA's very mature CUDA software layer. A mid-February 2024 update added support for Radeon PRO W7800 and RX 7900 GRE GPUs, as well as ONNX Runtime.

Larabel believes that "ROCm 6.1" is in for an imminent release, given his tracking of increased activity on publicly visible developer platforms: "For MIPOpen 3.1 with ROCm 6.1 there's been many additions including new solvers, an AI-based parameter prediction model for the conv_hip_igemm_group_fwd_xdlops solver, numerous fixes, and other updates. AMD MIGraphX will see an important update with ROCm 6.1. For the next ROCm release, MIGraphX 2.9 brings FP8 support, support for more operators, documentation examples for Whisper / Llama-2 / Stable Diffusion 2.1, new ONNX examples, BLAS auto-tuning for GEMMs, and initial code for MIGraphX running on Microsoft Windows." The change-logs/documentation updates also point to several HIPIFY for ROCm 6.1 improvements—including the addition of CUDA 12.3.2 support.

The story of the MSX platform, as it celebrates its 40th anniversary, is one of the games industries remarkable journeys that has left a powerful imprint on the world of technology and entertainment. Born in June 1983, the MSX, an acronym for "Machines with Software eXchangeability", emerged as a pioneering force that not only connected continents but also people through the magic of computing.

The birth of the MSX was a momentous event, an ambitious undertaking that materialized through the visionary partnership of Microsoft and ASCII Corporation. Their shared vision was clear: to create a standardized computing platform that would foster compatibility and ease of software development across a myriad of hardware manufacturers. What began as an idea soon evolved into a global phenomenon, revolutionizing the way people interacted with computers and forever altering the course of gaming history.

AMD today announced the AMD Versal Premium VP1902 adaptive system-on-chip (SoC), the world's largest adaptive SoC. The VP1902 adaptive SoC is an emulation-class, chiplet-based device designed to streamline the verification of increasingly complex semiconductor designs. Offering 2X the capacity over the prior generation, designers can confidently innovate and validate application-specific integrated circuits (ASICs) and SoC designs to help bring next generation technologies to market faster.

AI workloads are driving increased complexity in chipmaking, requiring next-generation solutions to develop the chips of tomorrow. FPGA-based emulation and prototyping provides the highest level of performance, allowing faster silicon verification and enabling developers to shift left in the design cycle and begin software development well before silicon tape-out. AMD, through Xilinx, brings over 17 years of leadership and six generations of the industry's highest capacity emulation devices, which have nearly doubled in capacity each generation.

Alphabet Incorporated's Google AI division has today revealed a planned update for its Bard conversational artificial intelligence chatbot. The experimental generative artificial intelligence software application will become capable of assisting people in the writing of computer code - the American multinational technology company hopes that Bard will be of great to help in the area of software development. Paige Bailey, a group product manager at Google Research has introduced the upcoming changes: "Since we launched Bard, our experiment that lets you collaborate with generative AI, coding has been one of the top requests we've received from our users. As a product lead in Google Research - and a passionate engineer who still programs every day - I'm excited that today we're updating Bard to include that capability."

The Bard chatbot was made available, on a trial basis, to users in the USA and UK last month. Google's AI team is reported to be under great pressure to advance the Bard chatbot into a suitably powerful state in order to compete with its closest rival - Microsoft Corporation. The Seattle-based giant has invested heavily into Open AI's industry leading ChatGPT application. Google's latest volley against its rivals shows that Bard's has become very sophisticated - so much so that the app is able to chew through a variety of programming languages. Bailey outlines these features in the company's latest blog: "Starting now, Bard can help with programming and software development tasks, including code generation, debugging and code explanation. We're launching these capabilities in more than 20 programming languages including C++, Go, Java, Javascript, Python and Typescript. And you can easily export Python code to Google Colab - no copy and paste required." Critics of AI-driven large language models have posited that the technology could potentially eliminate humans from the job market - it will be interesting to observe the coder community's reaction to Google marketing of Bard as a helpful tool in software development.