AI news story

Sakana AI and NVIDIA Introduce TwELL with CUDA Kernels for 20.5% Inference and 21.9% Training Speedup in LLMs

Sakana AI and NVIDIA Researchers demonstrate that simple L1 regularization can induce over 99% sparsity in feedforward layers with negligible downstream performance impact, and translate that sparsity into real GPU throughput gains using new sparse d

  • Hardware
  • Source: MarkTechPost
  • Published: 2026-05-11
  • Signal score: 4
  • 21 sources

Editor's take

Researchers from Sakana AI and NVIDIA have demonstrated a method to achieve significant speedups in Large Language Model (LLM) inference and training by exploiting sparsity induced through L1 regularization.

This development is significant as it directly addresses the computational bottlenecks hindering wider LLM deployment. By achieving over 99% sparsity in feedforward layers with minimal performance degradation, the technique promises to lower the energy consumption and latency associated with running models like NVIDIA's own H100 GPUs, impacting cloud providers and enterprises seeking more efficient AI infrastructure.

The key question is how this sparse kernel implementation scales across different LLM architectures and hardware generations beyond the NVIDIA A100 and H100. Further investigation into its effectiveness with smaller, more specialized models, and its compatibility with other hardware accelerators, will be crucial in determining its broad applicability.

Signal score: 4

This event was corroborated by 21 independent sources. The signal score weighs cross-source corroboration, recency, source weight and topic salience. How we rank stories.

More Hardware stories

  1. Firebird Takes Its AI Factory Platform Global With a 2-Gigawatt Pipeline

    Unite.AI · 2026-08-08

    Firebird opened its first AI factory in Hrazdan, Armenia, on August 8, 2026, and used the ceremony to lay out the rest of the map: a second market in Kazakhstan with 125 megawatts

  2. NVIDIA AI Releases NOOA: An Object-Oriented Python Framework That Turns an AI Agent Into a Single Python Class

    MarkTechPost · 2026-08-07

    NVIDIA Labs has open-sourced NOOA (NVIDIA Object-Oriented Agents), a model-agnostic Python framework for building AI agents.

  3. Firebird Launches CIS Region’s Largest AI Factory in Armenia

    NVIDIA AI Blog · 2026-08-08

    The global buildout of AI infrastructure reached a new milestone today — Firebird, an emerging AI cloud, launched the CIS region’s largest AI factory in Armenia

  4. d-Matrix Buys Wallaroo to Orchestrate Inference Across Chips

    Unite.AI · 2026-08-03

    d-Matrix has acquired Wallaroo.ai, a maker of software for deploying and orchestrating AI inference, in a deal the Santa Clara chip company announced on August 3, 2026.

  5. ASML Supplier Zeiss Says It Can Handle Demand for Key AI Parts

    Bloomberg · 2026-08-03

    One of the critical suppliers in the semiconductor industry, Germany’s Zeiss Group, pushed back on investor concerns about bottlenecks in the AI supply chain and said it’s

  6. Thinking Machines Lab Releases Inkling-Small: A 276B Total, 12B Active Open Weights Multimodal MoE Model

    MarkTechPost · 2026-08-02

    Inkling-Small matches Inkling at a quarter the size, and its NVFP4 checkpoint runs on one NVIDIA B300 GPU