
Imagine being right in the heart of the tech world where the next big AI breakthrough is just being revealed! That’s exactly what we at Aixia experienced during Nvidia’s GTC conference in San Jose – often referred to as the “AI Super Bowl” or even the “Woodstock of AI”.
Highlights from the conference
The atmosphere was electric as Jensen Huang, Nvidia’s visionary CEO, took the stage in front of thousands of excited attendees. His presentation revealed, among many other things, three exciting innovations that will shape the AI landscape in the coming years:
Blackwell Ultra – More powerful calculations
The new generation GPU architecture is planned to be launched in the second half of 2025. The most impressive is the new rack system(GB300 NVL72) that connects 72 GPUs with 36 Grace CPUs for enormous computing power. Simply put, the platform delivers about 1.5 times higher performance than its predecessor Hopper. What does this mean in practice? Well, data centers running AI services can become up to 50 times more profitable! For us, and not least our customers, who develop AI solutions, this opens doors to more advanced projects that were previously far too time-consuming or expensive.
NVIDIA Dynamo – the heart of the AI factory
The third and perhaps most exciting innovation was Dynamo –
a new open framework to streamline AI in production. Think of it as
the “AI factory operating system” – software that ensures AI models
can run faster and cheaper.
Jensen Huang demonstrated how Dynamo can double
the performance of large language models through smart management of computational resources.
This is absolutely critical to making AI accessible to more businesses and
use cases.

The future is exciting – and closer than you think
One of the most thought-provoking aspects was the discussion around ‘reasoning AI’ – systems that can analyze complex problems and act autonomously. With the power of the new GPU clusters, we envision AI that can:
- Analyze industrial processes in real time
- Controlling robots that make their own decisions
- Simulate scenarios to find optimal solutions
At Aixia, this inspires us to think bigger about our industrial automation projects – perhaps by training AI systems to control machines and vehicles in factories, using advanced simulation as a basis.
From vision to reality
We left GTC 2025 full of energy and ideas. New technologies open up opportunities not only to do things faster – but to do completely new things that were previously impossible.
With the new tools, the question is no longer whether something is possible to optimize with AI, but how fast we can do it. And that’s a thought that drives us forward every day.
Want to know more about how these new technologies can help your business? Get in touch with us at Aixia – we’re happy to share our insights and help you navigate the AI landscape!
