Modular
@Modular
Followers
19K
Following
474
Statuses
654
The future of AI development starts here. Sign up to our 📪 Newsletter → https://t.co/gpuHGRyHTs. We are hiring → https://t.co/cPTAes0HMt 🚀
Joined January 2022
MAX 24.6 is here and available for download today! Say hello to MAX GPU: the first vertically integrated gen AI serving stack, with SoTA performance on NVIDIA A100 and support for deployment across all major clouds 🤯 Don't miss the announcement:
1
17
109
And while you're at it, check out @clattner_llvm's keynote, where he explores how MAX is redefining AI inference and serving—unlocking faster progress, better performance, and less pain for AI engineers. 🧑🚀 🚀
0
3
8
To understand how AI can break free from hardware lock-in, we must understand how we got here. Stay tuned for part 3 of this series, where @clattner_llvm will discuss how CUDA got to be so successful in the first place. Hint: its success isn’t really about performance.
0
1
7
What exactly is “CUDA”? 🤔 In part 2 of our Democratizing Compute series, @clattner_llvm pulls back the curtain, shedding light on why @deepseek_ai and others are bypassing it entirely.
4
21
160
Tune into the replay of our latest community meeting on YouTube! ▶️ We discussed Owen's structured async Mojo proposal and highlighted two standout community projects: Brian's EmberJSON for Mojo JSON parsing and Martin's Modo for generating Mojo docs.
1
2
11
Don’t miss Monday's community meeting! We’ll dive into Owen’s async Mojo proposal, Brian’s parsing library, Martin’s Mojo docs generator, and a live demo from Bill, combining MAX Serve and Open WebUI to build a RAG and web search app 🤩 RSVP in the forum:
0
1
17
In the weeks ahead, @clattner_llvm and the @Modular team will continue this multipart series, discussing: 🎯 Why has CUDA been so successful? ❓ Why do other hardware makers struggle to provide comparable AI software? 🚀 How can we as an industry move forward? Anything else you'd like to see? Let us know in a reply!
🐋 #DeepSeek is rewriting the story that only those with vast compute resources can drive cutting-edge AI. In our first article of a series, @clattner_llvm discusses how novel ideas, backed by highly focused teams, can unlock efficiency breakthroughs
2
0
6
Unlock the power of MAX Serve and AI agents! 🤖 Our latest blog post walks through integrating OpenAI's function calling, FastAPI, and Llama3.1 to build an interactive app that pulls in real-time data based on user queries:
0
3
16
RT @GoodPutMax: The latest product launched by @Modular is called #MAX :) thanks for a very insightful conversation, @clattner_llvm ! https…
0
1
0
🚀 Deploy Llama 3 with RAG and web search in minutes! MAX 🧑🚀 + @OpenWebUI gives you a complete platform for shipping SOTA expert knowledge systems. Zero config needed, runs on laptop or GPU server. Get building now! #AI #LLM
0
3
17
MAX 24.6, MAX GPU, and community projects—what a combination! Check out the latest community highlights, including a MAX Graph implementation of Moondream, a small but mighty vision language model. 🌛 Don’t miss it:
0
3
12
Mojo 🔥 24.6 brings important changes to argument conventions and lifetime management, making Mojo's memory and ownership model more intuitive while maintaining strong safety guarantees 💪 Deep dive into Mojo 🔥 24.6 with our latest blog post:
1
6
52
RT @kylejohnmorris: Join us at the 2nd Democratize Intelligence Summit! 🗓️ January 24th, 2025 10 am - 5:40 pm PST 📍San Francisco Regist…
0
9
0
Missed yesterday's community meeting? Catch the replay on YouTube to dive into how the Modular team benchmarked the first MAX GPU release and learn about @bradlarson's new MAX-powered image processing framework, MAX-CV: ▶️
0
3
11
Kick off 2025 with this year’s first Modular community meeting on Monday, January 13th! 🥳 We’ll cover MAX GPU benchmarking and learn about Brad Larson’s new MAX-powered image processing framework, max-cv 🖼️ RSVP in Discord so you don't miss out:
0
3
22
1
0
1