Discover the enhanced capabilities of RAPIDS AI #cuDF which boosts Pandas performance up to 100x without code changes. Watch the on-demand #GTC24 session to learn how. nvda.ws/4cXeNFm

Meta Llama 3, the latest open #LLM from Meta — built with NVIDIA technology — is optimized to run on our GPUs from the cloud and data center to the edge and the desktop. bit.ly/3Un6CLb

NVIDIA Accelerates Inference on Meta Llama 3. A single NVIDIA H200 Tensor Core GPU generated about 3,000 tokens/second — enough to serve about 300 simultaneous users — in an initial test using the version of Llama 3 with 70 billion parameters. blogs.nvidia.com/blog/meta-llam…

Awesome to see RAPIDS AI 24.04 officially launches cuVS (the library to accelerate vector search on GPU), in addition to pandas 2.0 support, and significant expansions of NetworkX GPU acceleration! medium.com/rapids-ai/rapi…

If anyone going to C++Now is driving to Denver today or is stranded at the airport, send me a DM or email. Please retweet for visibility.

Learn how to achieve speed-of-(de)light by Revamping llm.c with the #CUDA C++ Core Libraries (CCCL) during this CUDA Mode talk. bit.ly/3JSDxBg

Hear Bradley Dice, Senior Software Engineer in GPU-Accelerated Data Analytics at NVIDIA, discuss the effort to hack 'import pandas' to boost its performance using cuDF, a GPU DataFrame library. Sign up for PyCon US's event #PyConUS to hear this talk live: nvda.ws/4dIbhyU.

Jensen Huang, Founder and CEO of NVIDIA joined Databricks Co-founder and CEO Ali Ghodsi for a fireside chat on innovation in the age of AI. "Whatever you do, just start," said Huang. "This is a fast-moving train. You don’t want to wait and observe an exponential trend."

Learn what it takes to do scans at the speed-of-light and how the #CUDA Core Compute Libraries (#CCCL) make it easy to achieve accelerated performance in this latest CUDA MODE talk. dy.si/uuUwyp

Encoding and Compression Guide for Parquet String Data Using RAPIDS AI NVIDIA developer.nvidia.com/blog/encoding-…

NEW feature: RAPIDS cuDF supports up to 2.1B rows of text data. Watch #pandas code with large strings get GPU-accelerated up to 30x with zero code changes. Try the notebook: nvda.ws/3LXTerT Read the blog: nvda.ws/3SKkDBD bit.ly/46GM5pq

Exciting to see GPU acceleration added to Polars! Process 100s of millions of rows of data in seconds on NVIDIA GPUs. 🙌 🔢 ⚡(via RAPIDS AI) bit.ly/3XCjttE

📣polars data GPU engine powered by RAPIDSai cuDF is now available in open beta. ⚡Accelerate Polars workflows up to 13x on NVIDIA GPUs. bit.ly/4dsrUgF

Join us at our #AIsummit for an in-depth session on "CUDA Key Features and Beyond" as NVIDIA's Stephen Jones, #CUDA Architect, dives into the latest innovations and future of #GPU computing. 📆 Oct 9, 2:00 PM - 2:40 PM EDT ➡️ nvda.ws/3YdcyIU 📍 Washington, DC

A true gem of Mother India! #RatanTata

Giving a talk at IIT Madras on October 22.

Join our monthly #CUDA Virtual Connect With Experts event to chat with NVIDIA developers about ETL, cuDF, and cuML, and ask live questions.🌟 📅 October 25, 10-11:30am PT. Learn more ➡️ bit.ly/4fgRo21

⚡Accelerate pandas workflows with RAPIDS cuDF. Learn how GPU acceleration and Unified Virtual Memory (UVM) enable faster, scalable data processing, even for datasets exceeding GPU memory. Tech blog ➡️ nvda.ws/4fYEuqn bit.ly/3OGnedr

🕔 Don’t let big data slow you down. Learn how to accelerate ⚡ Polars using lazy evaluation and GPU acceleration with NVIDIA cuDF. ➡️ nvda.ws/3YHQFkP

Extending Velox - GPU Acceleration with cuDF velox-lib.io/blog/extending…