# Google Antigravity — Modernizing 40-year-old scientific code with Gemini 4 Argon

- Company: Google Antigravity (antigravity.google)
- Announced: 2026-11-03T08:00:00+00:00
- Category: not stated
- Coverage: not counted
- Announcement: no
- Group: routine
- Source: https://antigravity.google/blog/modernizing-40-year-old-scientific-code-with-gemini-4-argon
- Record: https://forck.live/items/31298-modernizing-40-year-old-scientific-code-with-gemini-4-argon
- Subject: Antigravity
- Models affected: Gemini 4 Argon

Google says its Antigravity agent, using Gemini 4 Argon, migrated the core of a 174,000-line Fortran coastal ocean model into memory-safe Rust in 59 minutes. The company picked FVCOM, a widely used coastal ocean simulation with 174,774 lines of Fortran 90 and C across 264 files, and gave the agent one prompt with four hard rules. Those rules required zero unsafe blocks, no heap allocations in hot loops, Rust files under 400 lines, and 95% branch coverage with zero numerical difference across 100 randomized seeds. The run had three stages: an 11-minute baseline capture that compiled the original Fortran with gfortran -O3, a 3-minute audit of all 264 files, and a 45-minute Rust port. During the port the agent scaffolded the Cargo workspace and rewrote the core solver into 13 Rust modules, each under 240 lines. Google says the agent caught six decades-old bugs, cut memory usage in half, and sped up simulations by 1.88x, and it is releasing the workflow as an open-source skill.

## Evidence

Verbatim from https://antigravity.google/blog/modernizing-40-year-old-scientific-code-with-gemini-4-argon:

> Using Gemini 4 Argon in Google Antigravity, we migrated the core of a 174,000-line Fortran coastal ocean model into 100% memory-safe Rust in 59 minutes. Along the way, the agent caught six decades-old bugs, cut memory usage in half, and sped up simulations by 1.88x. Today, we’re sharing how Antigravity ran the migration end to end and releasing the workflow as an open-source skill.
> 
> We built Google Antigravity for long-horizon engineering work where an agent needs to do much more than suggest the next few lines of code. Earlier this week, as part of the Gemini 4 Argon launch in Google Antigravity , we shared a short video from one of our favorite internal stress tests: pointing Antigravity at a 40-year-old Fortran codebase and letting it autonomously plan, rewrite, test, and optimize the simulation in Rust.
> 
> Gemini 4 Argon in Google Antigravity modernizes a 40-year-old Fortran ocean model into memory-safe Rust.
> 
> After we posted that clip on X and YouTube, we got a lot of questions from researchers and systems engineers asking how we kept the agent from breaking 40 years of validated math.

---

Record: https://forck.live/items/31298-modernizing-40-year-old-scientific-code-with-gemini-4-argon
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md