Data Driver

Blog archive

Everything Anyone Could Possibly Want to Know About .NET 7 Performance Improvements

Here, I'll save you hours of time with a quick TL;DR: .NET 7 is fast.

But I come here not so much to talk about that, but the sheer size of the blog post announcing that.

Published today, Aug. 31, the post by Microsoft's Stephen Toub, "Performance Improvements in .NET 7," weighs in at more than 76,000 words, according to Microsoft Word. It must have taken me a couple minutes just to copy the thing to get a word count. I didn't time it; I just resolutely kept by finger down on the mouse button as I gritted my teeth and grimly came to terms with how long it was going to take to copy the text, scrolling through page after page after page, determined to finish the monumental, seemingly Sisyphean task. I thought Trump might be re-elected before I finished and could go out to vote. (Yes, later I realized I could have saved some time by clicking on the first word, holding the Shift key down and then clicking on the last word -- but I'm accustomed to the click-and-scroll method, which usually takes a second or two.)

I did time how long it took just to paste the text into Word for a word count: a good 16 seconds. To paste text.

"Whew! That was a lot. Congrats on getting through it all," said Toub toward the end. As if someone actually did. If you could read three words per second, it would take 6 to 7 hours to read (someone check my math on that -- I'm a journalist).

"Hi Stephen! Thank you for your great articles. I really enjoy reading this article -- although it will take me some time πŸ™‚," said the first comment to the post.

"I swear there articles are getting longer every year but I love it. Didn't finish it all yet but great work!" read another comment.

All joking aside (for now), the dev team seems to have done a great job increasing performance in .NET 7, which will debut in November.

Toub exhaustively detailed improvements in 23 areas, ranging from Setup and JIT to Registry and Analyzers.

Toub's Tome TOC
Toub's Tome TOC (source: Microsoft).

He also improved on my TL;DR above (which I hadn't seen when I penned mine).

TL;DR: .NET 7 is fast. Really fast. A thousand performance-impacting PRs went into runtime and core libraries this release, never mind all the improvements in ASP.NET Core and Windows Forms and Entity Framework and beyond. It's the fastest .NET ever. If your manager asks you why your project should upgrade to .NET 7, you can say 'in addition to all the new functionality in the release, .NET 7 is super fast.'

As for what's next, Toub said: "The next step is on you. Download the latest .NET 7 bits and take them for a spin. Upgrade your apps. Write and share your own benchmarks. Provide feedback, positive and critical. Find something you think can be better? Open an issue, or better yet, submit a PR with the fix. We're excited to work with you to polish .NET 7 to be the best .NET release yet; meanwhile, we're getting going on .NET 8 πŸ™‚"

He might this very second be writing the .NET 8 performance improvement post in order to finish on time.

Posted by David Ramel on 08/31/2022


comments powered by Disqus

Featured

  • Agents Now Conduct 'Deep Research' in Azure AI Foundry Limited Preview

    Microsoft has brought OpenAI's Deep Research model to Azure AI Foundry, giving developers API and SDK access to autonomous research agents that gather, analyze, and report on web-scale data. Now in public preview, the capability powers enterprise workflows with reasoning-grade intelligence and programmable orchestration.

  • Linear Regression Using JavaScript

    Dr. James McCaffrey presents a complete end-to-end demonstration of linear regression using JavaScript. Linear regression is the simplest machine learning technique to predict a single numeric value, and a good way to establish baseline results for comparison with other more sophisticated regression techniques.

  • Creating Simple Chat Bots with Microsoft Fabric Datastores

    At Visual Studio Live! San Diego, Ginger Grant of Desert Isle Group will lead a practical, demo-driven session on how to build simple yet powerful chatbots using Microsoft Fabric lakehouses and warehouses. Attendees will learn how to use AI skills and grounding techniques to enable conversational data access -- quickly and cost-effectively. Ideal for developers ready to extend analytics with conversational interfaces.

  • Prompt Engineering? See VS Code Team's System Prompts for Copilot Chat, Now Open Source

    "Explore the codebase and learn how agent mode is implemented, what context is sent to LLMs, and how we engineer our prompts."

  • VS Code Goes Transparent as Open-Source AI Editor

    Microsoft has open sourced the GitHub Copilot Chat extension, marking a major step in turning VS Code into an open-source AI editor focused on transparency, collaboration, and community-driven development.

Subscribe on YouTube