Google launches DiffusionGemma with 4x faster text generation

INVESTING.COMJun 10, 4:59 PM UTC

Key insights

  • Google's release of DiffusionGemma, an AI model capable of significantly faster text generation, highlights advancements in AI technology. This development, in collaboration with NVIDIA, could spur innovation and competition within the AI and semiconductor sectors. While not a direct market mover, it signals ongoing technological progress that may indirectly benefit companies involved in AI development and hardware, potentially influencing investor sentiment towards tech stocks.
Google launches DiffusionGemma with 4x faster text generation

Investing.com - Google (NASDAQ:GOOGL) released DiffusionGemma today, an experimental open model that generates text up to four times faster than traditional language models on dedicated GPUs. The 26 billion parameter Mixture of Experts model uses text diffusion to generate entire blocks of text simultaneously rather than processing tokens sequentially.

The model activates only 3.8 billion parameters during inference and fits within 18GB VRAM limits of high-end consumer GPUs when quantized. DiffusionGemma achieves over 1,000 tokens per second on a single NVIDIA (NASDAQ:NVDA) H100 GPU and over 700 tokens per second on NVIDIA GeForce RTX 5090 hardware.

DiffusionGemma generates 256 tokens in parallel with each forward pass, allowing bi-directional attention where every token can attend to all others. The model iteratively refines its own output to fix mistakes in real-time, though Google states its overall output quality is lower than standard Gemma 4 models.

Google released the model under an Apache 2.0 license on Hugging Face. The company designed DiffusionGemma for researchers and developers exploring speed-critical, interactive local workflows such as in-line editing, rapid iteration, and generating non-linear text structures.

The model works with MLX, vLLM with Red Hat integration, Hugging Face Transformers, Unsloth, and NVIDIA NeMo. Google worked with NVIDIA to optimize performance across the hardware stack, including GeForce RTX 5090 and 4090 GPUs for consumer setups and Hopper and Blackwell systems using NVFP4 kernels for enterprise deployments.

This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.

ProPicks AI evaluates GOOGL alongside thousands of other companies every month using 100+ financial metrics. Using powerful AI to generate exciting stock ideas, it looks beyond popularity to assess fundamentals, momentum, and valuation. The AI has no bias—it simply identifies which stocks offer the best risk-reward based on current data with notable past winners that include Super Micro Computer (+185%) and AppLovin (+157%). Want to know if GOOGL is currently featured in any ProPicks AI strategies, or if there are better opportunities in the same space?

Continue reading on INVESTING.COM

Related Articles