Orbital Industries

Today, we’re pleased to announce we’ve raised $50 million in Series B funding led by @pluralplatform This round also includes existing and previous investors NVentures (NVIDIA), @radicalvcfund, @CompoundVC and @FlyVC. The raise scales our AI engine for the physical economy and builds next-generation data centre hardware. Orbital Industries builds hardware from the atoms up, addressing the physical constraints of AI infrastructure. Through @OrbitalITGlobal, our commercial brand for data centre infrastructure, our first products tackle bottlenecks in cooling and deployment, with a non-toxic fluid for next-generation GPUs and modular infrastructure that drastically cuts build times. Read more about the round: orbitalindustries.com/news/press/orb…

We’re excited to announce Orbital Industries is working with @nvidia to deliver DSX AI Factory Infrastructure. Through our data center division @OrbitalITGlobal , we are implementing DSX infrastructure at scale via modular systems. Manufactured in controlled environments and deployed rapidly, modular AI factories accelerate the path from energy input to AI output up to six times faster than traditional approaches. Within the wider DSX ecosystem, Orbital IT is focused on the power, efficiency, and delivery-speed challenges of scaling AI factories for the next generation of machine intelligence, enabling partners to maximize tokens, power, and profit. Our CEO Jonathan Godwin shared more on why the next wave of AI depends on a new approach to infrastructure: "By working with NVIDIA on DSX, we are building the foundation for a new model of AI data center deployment - one that is faster, repeatable and designed for the realities of global demand. By using our AI platform, we remove errors in the design phase, slashing lead times and ensuring our modular data centers operate at peak thermodynamic efficiency from day one." This announcement comes ahead of London Tech Week 2026, where the infrastructure constraints of AI, from deployment speed to power density, will take centre stage. Our COO Daniel Miodovnik joins a panel at London Tech Week on the Deep Tech Stage tomorrow at 15:30pm on ‘The Limits of AI Data Centres and What Comes Next.’ Read more about the partnership: it.orbitalindustries.com/news/press/orb…

Really excited to get this out! Remarkable to me how quickly this field continues to progress. And I think it’s going to accelerate even faster thanks to the combination with LLMs. Please try them out and let us know how you find them.

Excited to announce more capital for @Orbital_Ind to execute on the massive near-term vision of enabling data centers and ambitious long-term vision of building products centered around novel materials derived from our SOTA family of models. We’ve been fortunate to get to partner with Orbital in every round of the company’s existence thus far and it is now our largest position in our second fund. Congrats to the entire team.

Exclusive: Orbital Industries, startup using AI to discover exotic new materials, raises $50 million Series B funding round fortune.com/2026/05/28/exc…

The London based startup raised €43 million in a Series B led by Plural with NVIDIA's venture arm also writing a cheque. Building a data centre usually takes 18 to 24 months,sometimes up to 3 years and Orbital Industries does it in 180 days. @Orbital_Ind does not just build data centres faster,It simulates materials at the atomic level before anything physical is made like design, test, optimise all virtually. The AI data centre market sits at $344 billion today. It is projected to hit $2 trillion by 2032. The bottleneck is no longer computing power. It is heat, energy and how fast you can build the physical infrastructure to house that power. Orbital is going after that bottleneck directly. One modular unit,factory built,dropped into site. 180 days from purchase order to live capacity. The founder Jonathan Godwin spent 5 years at DeepMind working on AI for science and materials design. He left to apply that work to a problem that is becoming urgent fast,the physical world cannot keep up with AI progress.

1/ I'm really excited to share that Plural is leading @Orbital_Ind's $50m Series B round Orbital is using AI to discover and design novel compounds from the atoms up, then engineering the products that exploit them directly.

An INSANELY cool UK AI startup founded by an ex-DeepMinder has just announced a $50m Series B. This is the coolest, most under-the-radar UK AI company I've come across. It designs, engineers and manufacturers physical infrastructure, using AI to accelerate how new technologies are discovered and brought to market. Its first use case is AI data centres where it has already had a breakthrough with cooling fluid and signed a multiple year deal with AWS. The team behind it are absolutely CRACKED: > Jonathan Godwin (@jgodwin_ai): AI researcher who spent five years at DeepMind working on AI for science, engineering and advanced materials design > James Gin-Pollock (@gin_james): A repeat AI founder who previously sold a company to Shutterstock > Daniel Miodovnik (@Dmiodovnik): Worked across finance, government AI and advisory work to the UN Today the company - @Orbital_Ind, has announced a $50m Series B led by @soundboy of @pluralpl. VERY COOL

Today, Orbital Industries launches @OrbitalITGlobal, our new commercial brand for AI data center infrastructure. Read the post below to find out more.

Keep an eye out for our upcoming OrbMol-v2 release next week. In the meantime, explore our Orb models here: github.com/orbital-materi…

Recently we announced our role in the release of @NVIDIA's ALCHEMI Toolkit. Here's what it looks like in action running highly efficient batched MD simulations with OrbMol-v2 on a single GPU within our agentic operating system. OrbMol-v2 is the upcoming, latest addition to our Orb family of models and it enables state-of-the-art speed and accuracy for simulating organic chemical systems. We use it to discover new breakthrough materials, such as our novel two-phase coolants for the next generation of NVIDIA and AMD GPUs.

Link to the full paper: pubs.rsc.org/en/Content/Art…

Orb-v3 is open source. If you're using it in your research, we'd like to hear about it. Reply to this thread. github.com/orbital-materi…

Orb-v3 ranked first in a new @digital_rsc benchmark of leading foundational machine learned interatomic potentials. It achieved the lowest MAE (0.198 eV) on non-outlier predictions across all models evaluated, and classified good vs bad ionic conductors with 84.84% accuracy at a 500 meV screening threshold. Why this matters: migration barriers determine how fast ions move through a material. Ionic conductivity is what separates viable battery candidates from non-viable ones. Calculating migration barriers accurately with DFT-NEB is the gold standard, but it's too slow for high-throughput screening across thousands of candidates. The study benchmarked the leading foundational machine learned interatomic potentials against DFT-NEB migration barriers across 574 battery material structures, spanning layered oxides, spinels, olivines, perovskites, and NaSICONs. Orb-v3 was the top performer across this diversity of chemistries.

At this week’s AWS Summit London, Arthur Hussey shared how we built our agentic operating system. This is an integrated AI platform that unites domain-specific AI across materials science, hardware design, and manufacturing. The foundation of how every Orbital product gets built. Arthur also walked through how we train and maintain it at scale using Multi-Instance GPUs on AWS HyperPod. MIG partitions a single GPU into isolated instances, each with its own memory and compute, so multiple inference jobs run in parallel on one accelerator without interference. Higher utilisation, lower cost per experiment. Orbital were the first in the world to run MIG on AWS HyperPod, which was only possible thanks to a custom integration we built. This drove AWS to complete an official global integration for MIG, now available to every HyperPod customer. Thanks to the AWS team and everyone who came to Arthur's session.

And here's a link to the ALCHEMI Toolkit: nvidia.github.io/nvalchemi-tool…

Keep an eye out for our upcoming OrbMol-v2 release In the meantime, explore our Orb models here: github.com/orbital-materi…

We're proud to be part of the beta release of @nvidia's ALCHEMI Toolkit. This toolkit is a unified environment for building custom atomistic simulation workflows, built on the same work that powered speedups to Orb-v3, the latest version of our simulation model Orb. Recap from March’s Orb-v3 update: Up to 1.7x faster inference on large systems (up to 100,000 atoms) and up to 33x faster inference on batched small systems with TorchSim support. Our forthcoming OrbMol-v2 model will leverage components from the toolkit including: → PME electrostatics for periodic Coulomb interactions → The MTK integrator for batched constant-pressure molecular dynamics The result: materially lower inference time.

We're hiring into our Machine Learning team here at Orbital! Here are the open roles: ➡️ Engineering Manager ➡️ Tech Lead Manager ➡️ Senior Machine Learning Engineer Follow the link to apply: jobs.ashbyhq.com/orbitalindustr…