Top Tweets for #MLSys2025
#UWAllen @UW & @nvidia researchers earned a #MLSys2025 Best Paper Award for boosting #LLM performance with FlashInfer—and showed “what’s possible when academia, industry & the open-source community innovate together,” says @ye_combinator. #AI #UWdiscovers https://t.co/cSxQXBJPpu
📢Exciting updates from #MLSys2025! All session recordings are now available and free to watch at https://t.co/gkofEOwcZp.
We’re also thrilled to announce that #MLSys2026 will be held in Seattle next May—submissions open next month with a deadline of Oct 30. We look forward to seeing your best work! #MLSys #AI #ML
1/ We had a great week at #MLSys2025!🧬🔬
🧠 GenBio AI Head of Infrastructure, @HongyiWang10, chaired the session on LLM and Diffusion Model Serving.
📝 We also presented “Radius: Range-based Gradient Sparsity for Large Foundation Model Pre-training."
And yes, we hit Topgolf too. 🏌️♂️
1/ We had a great week at #MLSys2025!🧬🔬
🧠 GenBio AI Head of Infrastructure, @HongyiWang10, chaired the session on LLM and Diffusion Model Serving.
📝 We also presented “Radius: Range-based Gradient Sparsity for Large Foundation Model Pre-training."
And yes, we hit Topgolf too. 🏌️♂️
Happy to have presented our paper "On Distributed Larger-Than-Memory Submodular Subset Selection" at #MLSys2025. This is a result of a collaboration between @SystemsGroupETH and @Google / @GoogleAI.
#MLSys2025 Ling Liu will give a keynote talk about "Responsible Finetuning of Large Language Models" at 10:30 today
🤔 𝗔 𝘀𝘁𝗮𝗻𝗱𝗼𝘂𝘁 𝘁𝗮𝗹𝗸 𝗯𝘆 𝗜𝗼𝗻 𝗦𝘁𝗼𝗶𝗰𝗮 𝗮𝘁 #MLSys2025 𝗵𝗶𝘁 𝗮 𝗻𝗲𝗿𝘃𝗲.
Why is there such explosive growth and consolidation around open-source LLM inference frameworks like @vllm_project and @lmsysorg SGLang?
Because the game is changing.
• 💻 Owning the compute is becoming essential as multi-modal agents and test-time reasoning shatter token-based pricing
• 🧠 Open-source models (LLaMA 4, DeepSeek R1) are hitting parity with closed alternatives - offering more control
• • 🧩 Model composability makes stacks more resilient, adaptable, and future-proof
And LLM inference frameworks offer the best solution for this inevitable future. They are the new distribution layer for open-source intelligence.
#MLSys2025 @AnimaAnandkumar will give a keynote Hardware-aware training and inference for large-scale AI at 10:30 today; make sure you don't miss it
Are you attending #mlsys2025? Come join us for drinks on Wednesday evening and meet the Inception team!
We are hosting a social at #mlsys2025 on Wednesday. Fill out this form if you're interested in attending: https://t.co/zSLrd9c3L9
Alongside #MLSys2025, @ZettaVentures, @a16z , and researchers from @Google and @cursor_ai are hosting an invite-only happy hour tonight (Tuesday, May 13).
Join us for some lively discussions around efficient model training, LLM innovations, distributed learning algorithms, hardware-efficient ML workflows, and more.
Request to attend at the link below or DM me.
https://t.co/drHaiGvqEm
2/ 🛠️ We’re hiring ML systems engineers to help design, train, and scale large multimodal foundation models for biology.
Let’s push the frontiers of AI + biology together. Come meet the GenBio AI team at #MLSys2025.
🔗 Learn more: https://t.co/n2lwNV8u8s
1/ 🚨 We’re at #MLSys2025 this week in Santa Clara!
🎤 Our Head of Infrastructure, @HongyiWang10, is chairing Session 10: LLM and Diffusion Model Serving.
🗓️ May 15, 1:15–2:40 PM PT
The session topics include:
• Prefix caching
• Fused Attention
• CPU offloading
• Scalable LLM serving
Full session details: https://t.co/XeXK2kMMy0
🎉 Congratulations to the FlashInfer team – their technical paper, "FlashInfer: Efficient and Customizable Attention Engine for LLM Inference Serving," just won best paper at #MLSys2025. 🏆
🙌 We are excited to share that we are now backing FlashInfer – a supporter and contributor to the project. We’ve chosen FlashInfer to release our top LLM inference kernels, including those from TensorRT-LLM, making them easy to integrate into @vllm_project, SGLang (@lmsysorg), and custom inference engines.
It started as a collaborative research project at @uwcse, @CarnegieMellon, and OctoAI (acquired by NVIDIA) with the goal of creating a flexible LLM inference kernel library that is engine agnostic, highly optimized, and easy to extend for new techniques such as algorithms for KV cache reuse. It is now a thriving open source project with production deployments and contributions from research and development teams across the AI systems community.
Check out FlashInfer today to get started to see our first Blackwell kernels for DeepSeek MLA available now: https://t.co/IIuI0Qdbrf
Congratulations again to Zihao Ye and all authors of the MLSys paper -- Lequn Chen, Wuwei Lin, Yineng Zhang, Stephanie Wang, Baris Kasikci, Arvind Krishnamurthy, Vinod Grover, Tianqi Chen, Ruihang Lai. And thank you to all community contributions, we look forward to continuing to grow this project.
FlashInfer paper: https://t.co/2TusYX9Oqz
Blackwell MLA kernel: https://t.co/STRgvOZiMv
FlashInfer won #MLSys2025 best paper🏆, with backing from @NVIDIAAIDev to bring top LLM inference kernels to the community
🎉 Congratulations to the FlashInfer team – their technical paper, "FlashInfer: Efficient and Customizable Attention Engine for LLM Inference Serving," just won best paper at #MLSys2025. 🏆
🙌 We are excited to share that we are now backing FlashInfer – a supporter and contributor to the project. We’ve chosen FlashInfer to release our top LLM inference kernels, including those from TensorRT-LLM, making them easy to integrate into @vllm_project, SGLang (@lmsysorg), and custom inference engines.
It started as a collaborative research project at @uwcse, @CarnegieMellon, and OctoAI (acquired by NVIDIA) with the goal of creating a flexible LLM inference kernel library that is engine agnostic, highly optimized, and easy to extend for new techniques such as algorithms for KV cache reuse. It is now a thriving open source project with production deployments and contributions from research and development teams across the AI systems community.
Check out FlashInfer today to get started to see our first Blackwell kernels for DeepSeek MLA available now: https://t.co/IIuI0Qdbrf
Congratulations again to Zihao Ye and all authors of the MLSys paper -- Lequn Chen, Wuwei Lin, Yineng Zhang, Stephanie Wang, Baris Kasikci, Arvind Krishnamurthy, Vinod Grover, Tianqi Chen. And thank you to all community contributions, we look forward to continuing to grow this project.
FlashInfer paper: https://t.co/2TusYX9Oqz
Blackwell MLA kernel: https://t.co/STRgvOZiMv
#MLSys2025 full house keynote from @istoica05 on An AI stack: from scaling AI workloads to evaluating LLMs
🎉 Congratulations to the FlashInfer team – their technical paper, "FlashInfer: Efficient and Customizable Attention Engine for LLM Inference Serving," just won best paper at #MLSys2025. 🏆
🙌 We are excited to share that we are now backing FlashInfer – a supporter and contributor to the project. We’ve chosen FlashInfer to release our top LLM inference kernels, including those from TensorRT-LLM, making them easy to integrate into @vllm_project, SGLang (@lmsysorg), and custom inference engines.
It started as a collaborative research project at @uwcse, @CarnegieMellon, and OctoAI (acquired by NVIDIA) with the goal of creating a flexible LLM inference kernel library that is engine agnostic, highly optimized, and easy to extend for new techniques such as algorithms for KV cache reuse. It is now a thriving open source project with production deployments and contributions from research and development teams across the AI systems community.
Check out FlashInfer today to get started to see our first Blackwell kernels for DeepSeek MLA available now: https://t.co/IIuI0Qdbrf
Congratulations again to Zihao Ye and all authors of the MLSys paper -- Lequn Chen, Wuwei Lin, Yineng Zhang, Stephanie Wang, Baris Kasikci, Arvind Krishnamurthy, Vinod Grover, Tianqi Chen. And thank you to all community contributions, we look forward to continuing to grow this project.
FlashInfer paper: https://t.co/2TusYX9Oqz
Blackwell MLA kernel: https://t.co/STRgvOZiMv
We are hosting a social at #mlsys2025 on Wednesday. Fill out this form if you're interested in attending: https://t.co/zSLrd9c3L9
#MLSys2025 make sure to attend 10:30am keynote @istoica05 An AI stack: from scaling AI workloads to evaluating LLMs. Checkout full schedule at https://t.co/y4F7jO3wfG
The PyTorch Foundation is a Gold Sponsor of #MLSys2025 this week in Santa Clara.
Visit the booth and explore talks from Soumith Chintala, Ion Stoica, and Exec Dir Matt White on open source AI and scalable ML systems.
🔗 https://t.co/n6EyYOBozg
#PyTorch #OpenSourceAI #AIInfrastructure
Most Popular Users
Elon Musk 
@elonmusk
240.7M followers
Barack Obama
@barackobama
119.2M followers
Donald J. Trump
@realdonaldtrump
111.7M followers
Cristiano Ronaldo
@cristiano
110.9M followers
Narendra Modi
@narendramodi
107M followers
Rihanna
@rihanna
97.7M followers
NASA
@nasa
92.2M followers
Justin Bieber
@justinbieber
91M followers
KATY PERRY
@katyperry
87.8M followers
Taylor Swift
@taylorswift13
81.7M followers
Lady Gaga
@ladygaga
73.2M followers
Virat Kohli
@imvkohli
70.2M followers
Kim Kardashian
@kimkardashian
69.9M followers
YouTube
@youtube
68.7M followers
Bill Gates
@billgates
64M followers
Neymar Jr
@neymarjr
63M followers
The Ellen Show
@theellenshow
62.4M followers
CNN
@cnn
61.9M followers
Selena Gomez
@selenagomez
60.9M followers
X
@x
60.8M followers