Federico Busato
@fbusato.bsky.social
Democratize CUDA optimizations @NVIDIA
CUDA Core Compute Libraries (CCCL)
Former lead of the Sparse Linear Algebra team | Opinions are my own.
CUDA Core Compute Libraries (CCCL)
Former lead of the Sparse Linear Algebra team | Opinions are my own.
Reposted by Federico Busato
AI data center buildouts already rival the Manhattan Project in scale, but there’s little public info about them.
So we spent the last few months reading legal permits, staring at satellite images, and scouring news sources.
Here’s what you need to know. 🧵
So we spent the last few months reading legal permits, staring at satellite images, and scouring news sources.
Here’s what you need to know. 🧵
November 10, 2025 at 6:03 PM
AI data center buildouts already rival the Manhattan Project in scale, but there’s little public info about them.
So we spent the last few months reading legal permits, staring at satellite images, and scouring news sources.
Here’s what you need to know. 🧵
So we spent the last few months reading legal permits, staring at satellite images, and scouring news sources.
Here’s what you need to know. 🧵
Reposted by Federico Busato
🆕 on the #Cpp Release Radar!
🎉 NVIDIA/cccl: v3.1.0
👉 Download and details here:
🔗 https://github.com/NVIDIA/cccl/releases/tag/v3.1.0
🎉 NVIDIA/cccl: v3.1.0
👉 Download and details here:
🔗 https://github.com/NVIDIA/cccl/releases/tag/v3.1.0
October 14, 2025 at 10:44 PM
🆕 on the #Cpp Release Radar!
🎉 NVIDIA/cccl: v3.1.0
👉 Download and details here:
🔗 https://github.com/NVIDIA/cccl/releases/tag/v3.1.0
🎉 NVIDIA/cccl: v3.1.0
👉 Download and details here:
🔗 https://github.com/NVIDIA/cccl/releases/tag/v3.1.0
🆕 on the #Cpp Release Radar!
🎉 NVIDIA/cccl: v3.0.3
👉 Download and details here:
🔗 https://github.com/NVIDIA/cccl/releases/tag/v3.0.3
🎉 NVIDIA/cccl: v3.0.3
👉 Download and details here:
🔗 https://github.com/NVIDIA/cccl/releases/tag/v3.0.3
October 7, 2025 at 5:07 AM
Reposted by Federico Busato
U.S. is once again the world's frontrunner in private AI investment in 2024, leaving other countries behind with a whopping $109.1B 🤖
Pradeep Jayakody breaks down the data at: https://www.voronoiapp.com/category/AI-Private-Investment-A-Global-Breakdown-in-2024-6637
Pradeep Jayakody breaks down the data at: https://www.voronoiapp.com/category/AI-Private-Investment-A-Global-Breakdown-in-2024-6637
September 22, 2025 at 5:03 PM
U.S. is once again the world's frontrunner in private AI investment in 2024, leaving other countries behind with a whopping $109.1B 🤖
Pradeep Jayakody breaks down the data at: https://www.voronoiapp.com/category/AI-Private-Investment-A-Global-Breakdown-in-2024-6637
Pradeep Jayakody breaks down the data at: https://www.voronoiapp.com/category/AI-Private-Investment-A-Global-Breakdown-in-2024-6637
Reposted by Federico Busato
Preprint:
Highly efficient protein structure prediction on NVIDIA RTX Blackwell and Grace-Hopper
nvda.ws/4n4xzz9
Visit the NVIDIA Digital Biology Labs website to find more information like this:
t.co/R9ufEZrGEA
Highly efficient protein structure prediction on NVIDIA RTX Blackwell and Grace-Hopper
nvda.ws/4n4xzz9
Visit the NVIDIA Digital Biology Labs website to find more information like this:
t.co/R9ufEZrGEA
nvda.ws
September 16, 2025 at 1:59 PM
Preprint:
Highly efficient protein structure prediction on NVIDIA RTX Blackwell and Grace-Hopper
nvda.ws/4n4xzz9
Visit the NVIDIA Digital Biology Labs website to find more information like this:
t.co/R9ufEZrGEA
Highly efficient protein structure prediction on NVIDIA RTX Blackwell and Grace-Hopper
nvda.ws/4n4xzz9
Visit the NVIDIA Digital Biology Labs website to find more information like this:
t.co/R9ufEZrGEA
Reposted by Federico Busato
Being a professor is so weird. You are constantly having to act like you know things.
September 11, 2025 at 7:18 PM
Being a professor is so weird. You are constantly having to act like you know things.
Reposted by Federico Busato
Are you just as excited as I am about deep learning, sparse matrices and tensors, sparse computations, structured and unstructured sparsity, sparse compilers, NVIDIA, and GPUs? If so, please note that my team at NVIDIA is hiring!
nvidia.wd5.myworkdayjobs.com/NVIDIAExtern...
nvidia.wd5.myworkdayjobs.com/NVIDIAExtern...
August 18, 2025 at 10:45 PM
Are you just as excited as I am about deep learning, sparse matrices and tensors, sparse computations, structured and unstructured sparsity, sparse compilers, NVIDIA, and GPUs? If so, please note that my team at NVIDIA is hiring!
nvidia.wd5.myworkdayjobs.com/NVIDIAExtern...
nvidia.wd5.myworkdayjobs.com/NVIDIAExtern...
Reposted by Federico Busato
⚡ Hackernews Top story: ACM Transitions to Full Open Access
Open Access Publication & ACM
Open Access Publication & ACM
www.acm.org
July 30, 2025 at 5:13 AM
⚡ Hackernews Top story: ACM Transitions to Full Open Access
Reposted by Federico Busato
🚨 New open-source tool: RefChecker
A tool that checks reference accuracy in academic papers.
✅ Parses PDFs, LaTeX, and text
✅ Catches broken DOIs, wrong authors, messed-up URLs
✅ Built with GitHub Copilot & Claude
🔗 github.com/markrussinov...
A tool that checks reference accuracy in academic papers.
✅ Parses PDFs, LaTeX, and text
✅ Catches broken DOIs, wrong authors, messed-up URLs
✅ Built with GitHub Copilot & Claude
🔗 github.com/markrussinov...
GitHub - markrussinovich/refchecker: A tool that validates academic paper references
A tool that validates academic paper references. Contribute to markrussinovich/refchecker development by creating an account on GitHub.
github.com
July 28, 2025 at 5:51 PM
🚨 New open-source tool: RefChecker
A tool that checks reference accuracy in academic papers.
✅ Parses PDFs, LaTeX, and text
✅ Catches broken DOIs, wrong authors, messed-up URLs
✅ Built with GitHub Copilot & Claude
🔗 github.com/markrussinov...
A tool that checks reference accuracy in academic papers.
✅ Parses PDFs, LaTeX, and text
✅ Catches broken DOIs, wrong authors, messed-up URLs
✅ Built with GitHub Copilot & Claude
🔗 github.com/markrussinov...
Reposted by Federico Busato
Is NVIDIA's GH200 the breakthrough for #HPC workloads, or is Grace merely a memory controller? In a Salishan Random Access talk, we analyze its complex memory system (with a hidden Easter egg!) and demonstrate its potential for revolutionary, large-scale science runs. 🎥🚀 #NVIDIA #AI
buff.ly/dUpJRs4
buff.ly/dUpJRs4
June 23, 2025 at 5:15 AM
Is NVIDIA's GH200 the breakthrough for #HPC workloads, or is Grace merely a memory controller? In a Salishan Random Access talk, we analyze its complex memory system (with a hidden Easter egg!) and demonstrate its potential for revolutionary, large-scale science runs. 🎥🚀 #NVIDIA #AI
buff.ly/dUpJRs4
buff.ly/dUpJRs4
Reposted by Federico Busato
May 14, 2025 at 11:14 PM
Reposted by Federico Busato
Native GPU (CUDA) programming in Python?
Nvidia releases CUTLASS 4.0, which ships with CuTe DSL, a programming language that is fully consistent with CuTe C++ in its programming model, APIs, abstraction level, and performance.
Nvidia releases CUTLASS 4.0, which ships with CuTe DSL, a programming language that is fully consistent with CuTe C++ in its programming model, APIs, abstraction level, and performance.
May 13, 2025 at 11:20 PM
Native GPU (CUDA) programming in Python?
Nvidia releases CUTLASS 4.0, which ships with CuTe DSL, a programming language that is fully consistent with CuTe C++ in its programming model, APIs, abstraction level, and performance.
Nvidia releases CUTLASS 4.0, which ships with CuTe DSL, a programming language that is fully consistent with CuTe C++ in its programming model, APIs, abstraction level, and performance.
@compiler-explorer.com is so valuable for the #CUDA community. Thanks @matt.godbolt.org for the amazing work. Well deserved! #NVIDIA
Thrilled to share that @compiler-explorer.com has received a $10K grant from NVIDIA's FOSS Fund! This support will help keep our GPU instances running and allow developers worldwide to experiment with CUDA code. Thank you NVIDIA AI Dev for supporting open source tools! #NVIDIAGrant
April 24, 2025 at 5:32 AM
@compiler-explorer.com is so valuable for the #CUDA community. Thanks @matt.godbolt.org for the amazing work. Well deserved! #NVIDIA
Reposted by Federico Busato
Rust-GPU / Rust-CUDA: Ecosystem of libraries and tools for writing and executing fast GPU code fully in Rust. ★4168 https://github.com/Rust-GPU/Rust-CUDA
Rust-GPU / Rust-CUDA
Ecosystem of libraries and tools for writing and executing fast GPU code fully in Rust.
github.com
April 12, 2025 at 8:32 PM
Rust-GPU / Rust-CUDA: Ecosystem of libraries and tools for writing and executing fast GPU code fully in Rust. ★4168 https://github.com/Rust-GPU/Rust-CUDA
Reposted by Federico Busato
Jensen Huang provided nearly all of the picks and shovels for the AI gold rush. James Surowiecki on how the Nvidia CEO became the new king of tech:
The New King of Tech
How Jensen Huang built Nvidia into a nearly $3 trillion business
bit.ly
April 9, 2025 at 12:15 AM
Jensen Huang provided nearly all of the picks and shovels for the AI gold rush. James Surowiecki on how the Nvidia CEO became the new king of tech:
Reposted by Federico Busato
Network Security 101 😂
March 27, 2025 at 5:36 PM
Network Security 101 😂
Reposted by Federico Busato
We're excited to join Bluesky in time for #PiDay! 🥳 Today is a day to celebrate the infinite possibilities of engineering and mathematics. #PiDay2025 #Pi #Mathematics #Engineering #IEEE #IEEEXplore
March 14, 2025 at 1:29 PM
We're excited to join Bluesky in time for #PiDay! 🥳 Today is a day to celebrate the infinite possibilities of engineering and mathematics. #PiDay2025 #Pi #Mathematics #Engineering #IEEE #IEEEXplore
Reposted by Federico Busato
git commit -m "first Bluesky post"
March 17, 2025 at 4:30 PM
git commit -m "first Bluesky post"
Sometimes good news
Wealthy colleges are flexing their financial aid packages and will attract more of the best students.
"Harvard announced on Monday that it plans to offer free tuition for students whose families earn $200,000 and below"
"Harvard announced on Monday that it plans to offer free tuition for students whose families earn $200,000 and below"
Harvard Will Make Tuition Free for More Students
Harvard is the latest elite school to announce that families with incomes of $200,000 and under will not pay tuition as a way to bolster diversity.
www.nytimes.com
March 18, 2025 at 6:08 AM
Sometimes good news
March 14, 2025 at 5:08 AM
Awesome work @aartbik.bsky.social! We will see a lot of sparse computation in the future. This is a great first step #nvidia #cpp #sparse
The sparse work continues with the introduction of the Universal Sparse Tensor (UST) type to the MatX GPU-accelerated numerical computing C++ library.
github.com/NVIDIA/MatX/...
github.com/NVIDIA/MatX/...
Sparse Tensor Support in MatX · NVIDIA MatX · Discussion #912
MatX is in the process of adding experimental support for sparse tensors. The implementation is based on the Universal Sparse Tensor (UST) type that uses a tensor format DSL (Domain Specific Langua...
github.com
March 14, 2025 at 5:01 AM
Awesome work @aartbik.bsky.social! We will see a lot of sparse computation in the future. This is a great first step #nvidia #cpp #sparse