Good Morning! NVIDIA introduced a new platform called NVIDIA Inference Microservices at GTC 2024 to help developers quickly deploy AI applications across NVIDIA GPUs. GitHub announced the public beta of its code scanning autofix tool that leverages CodeQL and Copilot AI to identify and fix vulnerabilities in code repositories. Epic Games plans to launch its own app store for iOS and Android devices later this year with a 12% revenue share for developers after the first 6 months, reigniting the debate around app store fees.

— Forrest Knight & Meghanadh Vasireddy

What Developers Need to Know from NVIDIA GTC

So GTC 2024 has come to an end, and as I’m writing this I’m on my way home from attending in-person. And let me tell you, it was an amazing experience! I chatted with many experts in the domain and attended sessions to specifically learn more about AI-assisted developer tools (for accelerated computing).

However, my biggest takeaway is this: tech professionals are to Jensen Huang as teenage girls are to Taylor Swift. He compared his keynote’s setup to a concert (which is a comparable comparison), but little did I realize he was truly the rockstar with the crazed fans mobbing him for selfies and signatures to go along with it.

Okay, but in all reality, my real biggest takeaway has to be NVIDIA’s new NVIDIA Inference Microservices (NIM) platform. It enables developers to swiftly create and deploy AI applications across its CUDA GPU base, facilitating data processing and AI model customization. NIM offers pre-built containers and APIs for efficient AI deployment in various domains, significantly reducing setup time from weeks to minutes. This approach allows for secure, scalable, and rapid development of generative AI applications within a company's existing infrastructure.

It’s essentially a set of tools that enable developers to deploy containerized AI applications quickly on NVIDIA GPUs. It has pre-built containers and standardized APIs to simplify the integration of AI capabilities into various domains, such as healthcare for patient data analysis and finance for real-time market predictions, ensuring both security and efficiency in managing proprietary data and scaling AI operations.

— Forrest Knight

GitHub's Code Scanning Autofix Goes tp Public Beta

GitHub has announced the public beta release of its code scanning autofix tool, designed to help developers quickly identify and remediate vulnerabilities within their code repositories.

The autofix tool leverages two key components:

• CodeQL: A semantic code analysis engine that scans the codebase and identifies potential security vulnerabilities.

• Copilot AI: GitHub's AI-powered coding assistant, which generates fix suggestions and explanations based on the vulnerabilities detected by CodeQL.

Currently, the tool supports JavaScript, TypeScript, Java, and Python. It can suggest fixes for over two-thirds of identified vulnerabilities across single or multiple files and dependencies. Support for C# and Go is planned for the future.

Under the hood, autofix uses private instances of large language models like GPT-4 through internal Copilot APIs to analyze code snippets and produce fixes/descriptions. However, these snippets are not used for further training of the models.

Limitations: While highly capable, the autofix tool may occasionally face limitations, such as failing to resolve a vulnerability or introducing new issues. It could also generate syntax errors, insert code in incorrect locations, or potentially risk issues with dependency changes or version compatibility.

Read More Here

Epic to launch iOS and Android store with controversial 12% cut

After years of legal battles with Apple and Google over their "monopolistic" app store policies, Epic Games is now set to launch its own app marketplace for iOS and Android devices later this year. This move has been enabled by the European Union's new Digital Markets Act.

The Competitive Offering:

• Developer-friendly 88/12 revenue share model, significantly better than Apple's 70/30 and Google's similar commissions

• 100% of earnings for developers in the first 6 months

The Perceived Hypocrisy: Critics are calling out Epic for implementing the very same reduced commission structure (12% after 6 months) that it vehemently opposed from Apple and Google, labeling their app store fees as anti-competitive.

The Epic Games Store will feature Epic's own titles like Fortnite alongside a curated selection of third-party games. Developers can choose to either:

• Use Epic's payment processing system and abide by the 88/12 revenue share

• Leverage their own payment systems and retain 100% of sales

With developers increasingly seeking more equitable distribution terms amid heightened regulatory attention on big tech's app store practices, the stakes are high for Epic to position its offering as a truly differentiated, developer-friendly alternative.

Read More Here

🔥 More Notes

• Introducing GNOME 46, “Kathmandu”: The GNOME project has released version 46, code-named "Kathmandu," which introduces several new features such as global search in the Files app, support for Microsoft OneDrive, remote desktop connection capabilities, and various improvements to accessibility, performance, and core apps. This release also includes enhancements for developers working with GNOME technologies.

• JetBrains is offering a new version of TeamCity designed for small and medium businesses: JetBrains has announced the beta for its new upcoming CI/CD service, TeamCity Pipelines. The new platform allows development teams to automate things like integrating changes into an application, testing them, and then finally delivering the updates to users.

• Android 15 gets its second developer preview with updates : The satellite communication functionality has been updated with new UI elements to provide a better user experience. Apps can also now use ServiceState.isUsingNonTerrestrialNetwork() to detect when a device is connected via satellite, and there is also now support for sending and receiving messages through SMS/MMS or RCS applications over satellite.

Youtube Spotlight

Node.js: The Documentary | An origin story

Back in 2008, most people thought of JavaScript as just a client-side language. But when Google's V8 appeared, young developer Ryan Dahl made the connection between non-blocking servers, V8, and JavaScript. It was by combining these key elements that he was able to create the now hugely popular Node.js.

Was this forwarded to you? Sign Up Here