# Samudra 2: 0.25 degree global ML ocean simulation

**URL:** <https://forum.access-hive.org.au/t/samudra-2-0-25-degree-global-ml-ocean-simulation/6583>\
**Category:** Machine Learning for Climate and Weather\
**Tags:** machine-learning, ocean\
**Created:** [2 July 2026 00:23 UTC](https://forum.access-hive.org.au/t/samudra-2-0-25-degree-global-ml-ocean-simulation/6583 "2026-07-02T00:23:13Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Aidan](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/aidan/32/42_2.png) [@Aidan](https://forum.access-hive.org.au/u/Aidan)\
**Post date:** [2 July 2026 00:23 UTC](https://forum.access-hive.org.au/t/samudra-2-0-25-degree-global-ml-ocean-simulation/6583/1 "2026-07-02T00:23:13Z")

</div>

[https://medium.com/@lz1955/samudra-2-a-fast-cheap-ai-ocean-model-now-at-the-scale-that-matters-b37883c62d51](https://medium.com/@lz1955/samudra-2-a-fast-cheap-ai-ocean-model-now-at-the-scale-that-matters-b37883c62d51)

> M²LInES’ neural ocean emulator now runs multi-year simulations at eddy-permitting resolution on a single GPU, turning a supercomputer-scale job into one you can run hundreds of times over. That opens the door to faster, cheaper ocean information for shipping, fisheries, energy, insurance, and seasonal forecasting.

 ![Global image](https://us1.discourse-cdn.com/flex020/uploads/access1/original/2X/0/04763b399d8d2ef261dceb877871975e075ecd4d.webp)

 ![Western Boundary Current](https://us1.discourse-cdn.com/flex020/uploads/access1/original/2X/1/133f23ea0c843af85457992feea52217379303f8.webp)

---

<div class="post-metadata">

**Author:** ![Tennessee](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/tennessee/32/2245_2.png) [@Tennessee](https://forum.access-hive.org.au/u/Tennessee)\
**Post date:** [2 July 2026 04:23 UTC](https://forum.access-hive.org.au/t/samudra-2-0-25-degree-global-ml-ocean-simulation/6583/2 "2026-07-02T04:23:19Z")

</div>

The data visualisation is amazing! The model also looks really interesting, thanks for sharing it.

---

<div class="post-metadata">

**Author:** ![navidcy](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/navidcy/32/101_2.png) [@navidcy](https://forum.access-hive.org.au/u/navidcy)\
**Post date:** [2 July 2026 23:19 UTC](https://forum.access-hive.org.au/t/samudra-2-0-25-degree-global-ml-ocean-simulation/6583/3 "2026-07-02T23:19:54Z")

</div>

Exciting and intriguing!

cc @taimoorsohail, @rmholmes, @PSpence

---

<div class="post-metadata">

**Author:** ![taimoorsohail](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.access-hive.org.au/taimoorsohail/32/201_2.png) [@taimoorsohail](https://forum.access-hive.org.au/u/taimoorsohail)\
**Post date:** [3 July 2026 03:46 UTC](https://forum.access-hive.org.au/t/samudra-2-0-25-degree-global-ml-ocean-simulation/6583/4 "2026-07-03T03:46:13Z")

</div>

Wow, very cool! I wonder if they overcame their previous hump of long-term projection drift outside the training window, or if this is just a higher resolution version of Samudra v1.