Estimated reading time: 4 minutes
If you weren't at the December NET+ AWS Tech Jam, you missed the electric atmosphere as Kevin Murakoshi, fresh from his first-ever in-person re:Invent experience, delivered a no-breathtaking tour through AWS's latest innovations. No slide deck can capture the enthusiasm that filled the virtual room as Kevin, working "literally until 9 o'clock this morning" to finalize his presentation, fired off announcement after announcement without missing a beat. This wasn't just information sharing – it was a masterclass in cloud excitement from someone who had witnessed the future firsthand.
New Foundation Models That Won't Break Your Budget
"We have to start here. We always start here," Kevin declared as he dove into what might be the most significant announcement for higher education: Amazon's new Nova family of foundation models. You could feel the collective energy shift as participants realized these models deliver performance comparable to leading commercial offerings but at dramatically lower costs – the kind of game-changer that budget-conscious institutions have been desperately waiting for.
The Nova family ranges from the lightweight Nova-Micro to the high-performance Nova Pro, with Kevin noting that Amazon's own benchmarking demonstrated they're "considerably less expensive than previous models at similar performance." As screenshots popped up faster than you could process them, Kevin's enthusiasm was contagious. This wasn't just another incremental AWS update – it was the cost advantage that could finally open doors for institutions to experiment with AI applications that previously seemed financially out of reach.
The .NET Tool Higher Ed Has Been Waiting For
You could practically hear the virtual gasps when Kevin revealed the new Q Developer .NET porting tool. As he raced through slide after slide, this announcement stopped everyone in their tracks. This wasn't just another feature – this was the answer to a universal campus pain point that's been draining budgets for years.
"Everyone has home-built .NET apps running on Windows servers because they were cheap as little as licensing was inexpensive. But you're running Windows and usually SQL running Windows servers just to run a .NET app. And that license gets expensive," Kevin explained, his excitement palpable even through Zoom.
The chat erupted as participants immediately recognized what this meant. This tool can port applications from .NET Windows to .NET Core that runs on Linux containers – potentially eliminating thousands in licensing costs while simplifying infrastructure. You could almost feel the collective wheels turning as everyone mentally inventoried their legacy Windows applications. For institutions with aging Windows applications, this capability alone justifies staying connected to the community for implementation guidance.
Cost Control Solutions That Finance Will Love
The finance-focused announcements revealed AWS has been listening closely to higher education's unique challenges. The introduction of Scale Serverless V2 with scale-to-zero capability means you'll no longer pay for database instances when they're not being used. This addresses a long-standing criticism, with Kevin emphasizing: "People have long criticized AWS for saying, 'Well, serverless is great, but the minimum cost isn't 0.'"
For institutions with cyclical workloads tied to academic calendars, this feature could dramatically reduce costs during low-usage periods. The tradeoff? A 15-second warm-up period that most applications can easily tolerate.
Another finance breakthrough came with the Account Invoice Configuration capability. Finally addressing the perennial challenge of multiple payers across departments, this feature allows specifying different cost centers for separate invoices without requiring multiple AWS Organizations.
Research Computing Enablers
Several announcements directly addressed research computing challenges. The new S3 Tables capability automates the management of tabular data, eliminating the complex data management overhead that research teams face with incremental data loads.
For data-intensive workloads, the SRD protocol enhancements now support GPU Direct and GPU NVMe storage to achieve an astounding 1,200 gigabit speeds between GPUs on different nodes. While Kevin smiled and simply called this "ludicrous connectivity speed," the implications for AI and simulation workloads are profound.
The community reaction to the S3 Browser announcement was immediate and enthusiastic. This capability finally provides a user-friendly interface for researchers to access their data without requiring AWS console access – addressing a long-standing friction point when migrating from traditional file storage systems to S3.
What's Next? You Decide
In a breathless finish that somehow maintained the same high-energy tempo he started with, Kevin concluded with what felt like the 100th announcement of the session. As Bob Flynn opened the exit poll, the chat erupted with questions and excited comments. This wasn't just passive information consumption – it was a community actively digesting ideas together.
If you missed this session, you not only missed the comprehensive, rapid-fire overview of re:Invent's most impactful announcements but also the warmth of a community that celebrates each other's wins. The flurry of links shared in the chat, the spontaneous questions that sparked mini-discussions – these are the moments that transform isolated IT work into a connected community journey.
The good news? You can relive it by watching the recording and there's always next month's Tech Jam to join this vibrant knowledge-sharing community. But as anyone who attended would tell you – being there live as Kevin raced through "one slide for every announcement" without slowing down once is an experience that simply can't be replicated in a blog post.
Be sure to check out the other blog posts we've written. As always, feel free to send any feedback to tmanik[at]internet2[dot]edu.