Bryce Adelstein Lelbach
@blelbach
Followers
17K
Following
1K
Media
1K
Statuses
12K
Principal Engineer at @NVIDIA working on programming languages. @adspthepodcast co-host. C++ Library Evolution chair emeritus. Frequent flyer. Horology fan.
Manhattan, NY
Joined March 2011
The latest revision of @INCITS/@isostandards COBOL comes out this year The goals of COBOL sound normal today: - Portable - Freely available - Designed by the community In 1959 it was radical & unprecedented It was also conceived of & led by women This is the story of COBOL
9
104
346
Full Self Driving Supervised drives like the best chauffeur imaginable Try it yourself at a nearby Tesla location
0
119
1K
.@TripIt continue to be the buggiest app I use. The new web UI is awful. Now it takes many clicks to see or edit info. Date/time entry is pedantic and painful. The app crashes or logs out, or fails edits constantly. Yet it has to be used since the web UI can't do some things.
1
1
2
.@GrabTH has the best rideshare app UI & it's not close When riding from an airport, they have step by step instructions with pictures guiding you to the pickup area They have separate entrances/pickup locations programmed for almost everywhere I wish they were in the US
0
0
2
💡 Empire AI is proving that purpose-built AI infrastructure can deliver billions in projected economic and societal benefits, turning AI investment into real ROI for the public sector. By democratizing access to #acceleratedcomputing resources at scale, Empire AI is
3
13
82
Thrilled to announce we're open-sourcing the CUDA Tile dialect and bytecode! https://t.co/wgUD9k4OxE What's included: • CUDA Tile MLIR dialect • Bytecode serialization/deserialization support • MLIR Python bindings for programmatic IR construction •
github.com
CUDA Tile IR is an MLIR-based intermediate representation and compiler infrastructure for CUDA kernel optimization, focusing on tile-based computation patterns and optimizations targeting NVIDIA te...
11
117
762
A year that redefined industries, and accelerated the future of computing. @TIME sat down with NVIDIA CEO Jensen Huang, exploring how NVIDIA’s breakthroughs in AI and accelerated computing are shaping the next industrial revolution. This incredible progress wouldn’t be
26
72
357
Sleigh the season with the most personal gift around. Get them a Cameo video!
0
101
1K
Congratulations to our founder and CEO, Jensen Huang, on being named @FT Person of the Year. This recognition reflects decades of bold technology bets, a distinctive leadership style, and a culture he has built at NVIDIA that believes in tackling the world’s most complex
The FT's Person of the Year has been the driving force behind the massive AI boom which has the capacity to reshape the global economy. https://t.co/Dbn0nC2wM5
122
259
2K
📢 Episode 264 is out! 📢 In this episode, @code_report chats with @KevlinHenney about aphantasia, anendophasia, synesthesia, techno-optimism, the future and more! 🥳 https://t.co/KNzA3ELJq1
0
4
7
Our co-founder, @ChrisMlachowsky, has been selected to join the National Academy of Inventors’ 2025 Class of Fellows. Since 1993, Chris has developed core technologies, authored dozens of patents, and helped transform NVIDIA into a global leader. Congratulations to Chris on
The National Academy of Inventors is proud to announce the 2025 Class of Fellow inductees! This distinguished group includes 169 U.S. Fellows and 16 International Fellows, representing 127 research universities, government agencies, and non-profit institutions around the world.
21
21
221
🚀 Want to become a CUDA ninja? Start with the new CUDA Programming Guide - Section 4 is your gold mine! It’s packed with features most developers don’t even know exist, and it can unlock serious performance gains, smarter debugging, and cleaner GPU code.
17
138
900
We're so close to 1000, but we just reached 2000 on LinkedIn (that ratio is concerning, yes). Please @giffmana need your help here :D
0
5
41
btw today at 3pm PST (in ~4 hours) we're having Vicki Wang from NVIDIA giving a @GPU_MODE talk on CuTe DSL, its features, and how to use the most of it if you're currently competing in the NVFP4 Blackwell competition this will be very helpful, but it's open to anyone!
3
20
203
🚨 NVIDIA CUDA 13.1 brings the largest update to the #CUDA platform in 20 years. Discover the new features and updates for improving performance and driving #acceleratedcomputing, including: ✅ Announcing CUDA Tile: Future-proof kernels by abstracting Tensor Cores. ✅
9
50
276
America needs you! Join U.S. Immigration and Customs Enforcement today.
3K
7K
33K
How will AI supercomputing transform the semiconductor industry? 🤔 Join NVIDIA’s Tim Costa at the Global Semiconductor Executive Summit as he breaks down how #acceleratedcomputing and AI are powering a new era of smart manufacturing—from AI driven chip design and
6
11
58
As #NeurIPS2025 comes to a close, we wanted to share a few of our favorite moments 📸 Thanks to all the researchers from around the world who joined us. 💚 From pioneering new scientific frontiers to building on decades of iterative research, we loved celebrating with our AI
4
12
86
TileGym: https://t.co/k9jPIYEyrW > This repository aims to provide helpful kernel tutorials and examples for tile-based GPU programming. Notably includes: > End-to-end integration examples with popular LLMs (Llama 3.1, DeepSeek V2)
github.com
Helpful kernel tutorials and examples for tile-based GPU programming - NVIDIA/TileGym
1
82
507
The largest advancement of the CUDA platform since its creation in 2006 is here 👀 Introducing CUDA Tile, a tile-based programming model that provides the ability to write algorithms at a higher level and abstract away the details of specialized hardware, such as tensor cores.
28
123
785
Did you know HomeView® goes in-depth on closing costs? Check out our free homebuyer education course today.
4
33
278
CUDA 13.1 is a big upgrade CUDA Tile basically rewires how devs talk to NVIDIA GPUs It lowers the skill barrier, boosts efficiency, and pulls even more AI workloads into the CUDA gravity well This is Jensen extending the moat again TPUs and pretenders just fell further behind
7
49
383
CUDA Tile is the future of GPU programming, and it's in Python. 🐍 cuTile Python in CUDA 13.1 lets you focus on your algorithm, not Tensor Core specifics. Simplify your HPC/AI code now. ✨ 🔗 https://t.co/5MTRNYGCRp
18
84
472