NVIDIA RIVA 128 Graphics Accelerator Introduction | Generated by AI
The RIVA 128 (also known as NV3) was NVIDIA’s first major 3D graphics accelerator, released in August 1997. It marked NVIDIA’s breakthrough into the consumer graphics market, helping establish its reputation before the GeForce era.
Here’s a clear breakdown:
1. Name meaning: “RIVA” stands for Real-time Interactive Video and Animation. The “128” refers to its 128-bit graphics engine (not memory bus, but internal architecture).
2. Architecture and specs:
- Codename: NV3
- Manufacturing process: 350 nm
- Core clock: ~100 MHz
- Memory: 4 MB SGRAM
- Memory interface: 128-bit
- Bus: AGP 2× (also available in PCI versions)
- API support: Direct3D 5, OpenGL (mini ICD)
3. Capabilities: It was one of the first cards to combine 2D, 3D, and video acceleration on a single chip. It could handle bilinear filtering, perspective correction, and Z-buffering, which were advanced at the time.
4. Performance and competitors: At launch, it competed with cards like:
- 3Dfx Voodoo Graphics
- Rendition Vérité V2100
- ATI Rage Pro
- Matrox Mystique
In many benchmarks, the RIVA 128 outperformed early Voodoo cards in Direct3D titles but was weaker in Glide games (which were 3Dfx-optimized).
5. Legacy: It was followed by the RIVA TNT (NV4) in 1998, which introduced true 32-bit color and multi-texturing. The TNT line eventually evolved into the GeForce brand, starting with the GeForce 256 (NV10) in 1999.
In short: The RIVA 128 was NVIDIA’s first successful mainstream GPU, laying the foundation for the company’s dominance in PC graphics.
Would you like me to explain how its architecture differs from the later RIVA TNT or GeForce 256 (for example, what made those much faster)?