1
2 Comments

NVIDIA Just Validated Our Niche Bet. Here Are the Numbers.

We build Ziva, an AI coding assistant for the Godot game engine. Last week, NVIDIA released a path tracing fork of Godot at GDC 2026. Full ray tracing. MIT licensed. On GitHub.

Here's why this matters to us specifically, and what it might mean if you're building for a niche market.

The bet we made

When we started building Ziva, Unity had 51% Steam market share. Godot had 5%. Every advisor and every VC said the same thing: build for Unity.

We picked Godot anyway because the growth curve mattered more than the snapshot. And the curve has been steep:

  • Godot games on Steam doubled yearly: 290 → 618 → 1,500 → 2,864
  • GMTK Game Jam 2025: Godot 40%, Unity 40%. Dead even.
  • GitHub: 102,000+ stars, one of the 50 most-starred repos

But there's a difference between "the numbers are good" and "NVIDIA builds rendering infrastructure for your platform." The second one is the signal that tells you the bet is paying off.

What NVIDIA actually did

They didn't release a plugin or a demo. They forked the entire engine, added production-grade path tracing using Vulkan, and put it on GitHub under MIT. The path tracer is GPU-agnostic (Vulkan, not proprietary CUDA). The denoiser is NVIDIA-only for now but AMD/Intel support is in progress.

Their stated plan: merge it back into mainline Godot via PR.

A $2.8 trillion company just committed engineering resources to making our platform better. For free. Under the same license we use.

What this changes for our business

1. Credibility with studios. The biggest objection we hear from potential customers at larger studios is "Godot isn't serious enough for us." NVIDIA investing in Godot's renderer answers that objection without us saying a word.

2. 3D market opens up. Most Ziva users build 2D games. Godot's 3D capabilities have been the main reason studios stayed on Unity or Unreal. Ray tracing narrows that gap significantly. More 3D Godot developers = more potential Ziva users.

3. AI + open source rendering. Because Godot's renderer is fully open, our AI can understand the rendering pipeline at the source code level. When ray tracing lands in mainline Godot, Ziva will be able to generate ray-traced materials and scenes from day one. Try doing that with Unreal's proprietary Nanite stack.

The niche market lesson

For founders considering a niche play: external validation compounds. When you pick a growing niche, you're not just betting on your own execution. You're betting that the ecosystem will attract investment from companies much larger than you.

We didn't make NVIDIA fork Godot. But we positioned ourselves so that when they did, it directly benefits our business. That's the advantage of picking a market with momentum.

The specific lesson: look for niches where the open source trajectory attracts corporate investment. NVIDIA's Godot fork follows the same pattern as IBM investing in Linux, Google investing in Kubernetes, or Meta investing in PyTorch. When big companies start contributing to your platform, your niche stops being niche.

What's next

We're already working on Ziva features that will support the new rendering capabilities when they land in mainline Godot. If you're building in the Godot ecosystem, now is probably the best time to start.

ziva.sh

posted to Icon for group Artificial Intelligence
Artificial Intelligence
on April 9, 2026
  1. 1

    This is a great breakdown—especially the point about external validation compounding over time. Betting on Godot early based on growth rather than market share looks like it’s really paying off now.

    The NVIDIA move feels like a huge credibility unlock, especially for studios that were hesitant about Godot’s 3D capabilities. Makes your positioning with Ziva even stronger going forward.

    Curious how you’re planning to capture the incoming wave of 3D developers?

    Prize pool just opened at $0. Your odds are the best right now.
    Quick one — do you work on ideas/projects like this?
    $19 entry → real competition
    🏆 Tokyo trip + hotel
    💰 Min $500
    Round is live 👉 tokyolore.com

  2. 1

    The IBM/Linux, Google/Kubernetes, Meta/PyTorch comparison is exactly the right frame for this. When a large company invests in a platform your tool is built on, you get a credibility transfer that would have taken years to earn independently.

    The niche lesson generalizes well beyond gaming: the best time to pick a niche is when the underlying platform is on an upward curve but before the big players have locked up the tooling ecosystem. You picked Godot when it was growing fast but still underserved. That gap is where durable businesses get built.

    Strong execution on the timing here. Curious whether you saw the NVIDIA partnership coming or if it was a genuine surprise validation.

Trending on Indie Hackers
I got my first $159 in sales after realizing I was building in silence User Avatar 52 comments I spent more time setting up cold email than actually selling. Here is what fixed it. User Avatar 41 comments Three Days Before Launch, I Let My Own Tool Tear Me Apart User Avatar 35 comments I got tired of rewriting the same content for 9 different platforms. So I built Repostify. User Avatar 29 comments A pattern I keep seeing in EdTech: traffic isn't usually the problem. User Avatar 23 comments I thought I was building a news visualization tool. Users thought it was a catch-up tool. User Avatar 20 comments