Data centers don’t just need more power. They need more compute from every MW without compromising uptime.
Power, cooling, controls, equipment, workloads, and maintenance are tightly coupled but managed separately.
We’re building Edviro: a world model to operate them as one.
We just hit SOTA on the standard physics-ML benchmark (Navier-Stokes 2D turbulence):
Edviro research is building world models that understand and operate energy infrastructure. Buildings are our first deployment environment.
We started with the NS2D benchmark because fluid dynamics underpins many energy systems, including airflow and industrial processes.
The goal was to rethink fundamental architectural choices. Instead of asking a neural network to learn the entire PDE update, we analytically solve the linear component and train the network only on the nonlinear increment.
The results: we hit 0.0305 (± 0.000079) across three independent seeds. Applying symmetry-averaged inference further reduces error to 0.02969 at 2x inference costs. Against code released claims, we beat them by 36.9%.
If you’re working on operator learning, physics ML, controls, or operate large-scale energy systems, we’d love to talk.
Introducing Origami. chat
The world’s first AI that finds you new customers.
1000+ companies use Origami for their outbound.
RT + reply with your website and we’ll send you 5 of your perfect customers right now👇
if you’re a cracked high schooler listen up
we’re launching the beacon fellows program, a selective fellowship for students who wanna drive sustainability and transparency for taxpayer money in their schools using energy data.
details + apply here: https://t.co/0XWK2OZqIc
The cost of building has collapsed. The most important thing for founders is having the right community and enough resources to keep shipping without token spend becoming the bottleneck.
If you’re at UCSB and want to spend 8 weeks building, apply at https://t.co/vEEt3h7u4Z.
Built and open-sourced an internal AI email copilot for our startup in a day.
This is exactly why we’re running an 8-week builder sprint through https://t.co/vEEt3h7u4Z, backed by local VCs and alumni & with credits provided by @cursor_ai.
We needed an internal tool for managing leads with human-in-the-loop assurance. With @cursor_ai, it took just a day: lead management, templates, AI draft enhancement, reply suggestions, follow-up logic, and a dashboard to manage it (check it out at https://t.co/mxD4k6btiz)
@driaug_@useplunk just hacked together an @obsdmd plugin that can send Markdown files and manage campaigns using the Plunk API. I can send my clients newsletters in <5 seconds. Love your project!