Super Smash Bros. 64 Nintendo DS port reaches playable state

Super Smash Bros. 64 is running on a Nintendo DS. Not through streaming. Not through a standard Nintendo 64 emulator. This is a proper source port, built to make the original game logic work on hardware that was never designed to handle it. The project, called Smash64DS, has already reached a playable stage. Mario and Fox can fight on Dream Land in a timed match on real Nintendo DS hardware. It’s still rough, and performance has a long way to go, but the basic idea now works. That’s the part that matters.

Super Smash Bros. 64 is running on a Nintendo DS. Not through streaming. Not through a standard Nintendo 64 emulator. This is a proper source port, built to make the original game logic work on hardware that was never designed to handle it. The project, called Smash64DS, has already reached a playable stage. Mario and Fox can fight on Dream Land in a timed match on real Nintendo DS hardware. It’s still rough, and performance has a long way to go, but the basic idea now works. That’s the part that matters.

This isn’t a simple visual remake

At first glance, the Nintendo 64 and Nintendo DS might seem close enough for this kind of project. Both systems can display low-polygon 3D graphics. Both use cartridges. Both belong to roughly the same era of Nintendo hardware design. But a port like this isn’t just a matter of shrinking textures and reducing the resolution.

The original Super Smash Bros. code expects the Nintendo 64’s processor, memory layout, graphics system, controller input and audio hardware. The DS works differently in almost every one of those areas. So the developers have to keep the original game behaviour intact while replacing the parts that communicate directly with the old console.

That includes rendering, sound, input, timing and memory management. It’s a bit like moving a working engine into a completely different car. The engine might still run, but almost every connection around it has to be rebuilt.

The playable build is deliberately limited

The current version focuses on one specific match: Mario versus Fox on Dream Land, with a one-minute timer and no items. That might sound narrow. It is. But it also makes sense.

A controlled match gives the developers a consistent way to test frame rate, collision detection, animation, controls and game timing. When the same fighters, stage and rules appear every time, it’s easier to spot what’s slowing the game down or behaving incorrectly.

The project isn’t pretending to be a finished handheld release. Right now, it’s a technical build designed to prove that the core game can run on the DS at all. And it can.

Performance is the biggest problem

The port aims to run at 30 frames per second. To keep the original game logic moving at the correct speed, it processes two gameplay updates for every rendered frame. That sounds clever because it is. But it’s also expensive.

On real Nintendo DS hardware, the current build reportedly runs at around 13.5 to 15 frames per second during demanding moments. That’s well below the target, and it affects how responsive and fluid the match feels.

Fighting games are especially sensitive to timing. A small slowdown doesn’t just make the animation look rough. It can change movement, attacks, reactions and the rhythm of the match. So optimisation isn’t a finishing touch here. It’s the main job. The developers will need to find faster ways to handle graphics, effects, animation, collision checks and other calculations without changing how the game actually plays.

How the DS version is being built

Smash64DS uses reconstructed Super Smash Bros. 64 game code as its base, then adds new systems specifically for the Nintendo DS. The project separates the original gameplay logic from the hardware-specific work. That means the rules, character behaviour and match systems can stay close to the Nintendo 64 version, while the DS handles graphics, controls and sound through its own replacement code.

That separation is important. Without it, every change made for the handheld could risk breaking the game itself. By keeping the platform layer distinct, the developers can work on DS performance without constantly rewriting the underlying combat logic.

Building the port isn’t a one-click process, though. The current setup requires a Windows PC, Nintendo DS development tools, PowerShell, Python, CMake, GNU Make and Microsoft’s C++ toolchain. A build script checks the required software, verifies the user’s game data, extracts the necessary assets and produces a Nintendo DS executable. The resulting file is around 11.4 MB.

You still need your own game data

The project doesn’t include Nintendo’s commercial game files. Users have to provide a legally obtained North American version 1.0 dump of Super Smash Bros. 64. The build process then extracts the required textures, models, audio and other assets locally.

That approach keeps the public repository focused on original porting code rather than distributing copyrighted material. It also means this isn’t a ready-to-download game. You can’t simply grab a file and start playing. You need the correct game version, the development tools and enough technical knowledge to compile the project yourself.

It works, but it isn’t ready

Smash64DS is still an early technical project. There’s no polished public release. The frame rate is low, the available match setup is limited and major parts of the game still need work. Sound, visuals, gameplay coverage and stability all require more development.

Even so, the project has already passed an important milestone. A recognisable Super Smash Bros. 64 match now runs interactively on a Nintendo DS. That’s not a concept video or a mock-up. It’s working code on the actual handheld. The next challenge is clear: make it faster, more stable and less limited. That won’t be easy. But the hardest first step has already happened. The game runs.

Spread the love
error: