VLLM
product on 6 shows · 14 statements across 9 episodes · said 115 times in 32 episodes since 2023
Latent Space 75 
the a16z Podcast 25 
the MAD Podcast 11 
20VC 2 
the Y Combinator Startup Podcast 1 
No Priors 1
Mentions by year, every show
tap a year for its mentions
Latent Space 75the a16z Podcast 25the MAD Podcast 1120VC 2No Priors 1the Y Combinator Startup Podcast 1
2026 45 mentions in 9 episodes 5 per episode
- How Open Source Became AI's Backbone | Inferact with a16z
- Dylan Patel: NVIDIA's New Moat & Why China is "Semiconductor Pilled”
- Next 100x in AI: Inference, Networking, & Self-Optimizing Models — Philip Kiely & Ali Taha, Baseten
- Agent Inference at the "Speed of Light" — How NVIDIA moves like a $4.3 Trillion Startup
- Nebius Co-Founder on AI Infrastructure Bubbles | How Price Elastic is Demand for Compute · 20VC with Harry Stebbings
- ⚡️ Google's Open AI Strategy — Omar Sanseviero, Google DeepMind
- Multi-GPU Kernels, Intelligence per Watt, Heterogeneous Inference, and More | YC Paper Club · Y Combinator
- The AI Memory Problem: Why Long Context Isn’t Enough — Dan Biderman, Engram Co-founder & CEO
- 1 more episode that year, every mention in 2026 →
2025 47 mentions in 11 episodes 4 per episode
- The Shape of Compute (Chris Lattner of Modular)
- DeepSeek V3, SGLang, and the state of Open Model Inference in 2025 (Quantization, MoEs, Pricing)
- Outlasting Noam Shazeer, Crowdsourcing Chai AI w/ 1.4m DAU — with William Beauchamp, Chai Research
- One Year of MCP — with David Soria Parria and AAIF leads from OpenAI, Goose, Linux Foundation
- 🕰️ The Oral History of Windsurf (ft. Varun Mohan, Scott Wu, Jeff Wang, Kevin Hou, Anshul R)
- A Technical History of Generative Media
- Dylan Patel on GPT-5’s Router Moment, GPUs vs TPUs, Monetization
- No Priors Ep. 127 | With SemiAnalysis Founder and CEO Dylan Patel
- 3 more episodes that year, every mention in 2025 →
2024 22 mentions in 11 episodes 2 per episode
- [Paper Club] Writing in the Margins: Chunked Prefill KV Caching for Long Context Retrieval
- Making AI Work: Fine-Tuning, Inference, Memory | Sharon Zhou, CEO, Lamini
- Windsurf: The Enterprise AI IDE
- Why Compound AI + Open Source will beat Closed AI — with Lin Qiao, CEO of Fireworks AI
- Answer.ai & AI Magic with Jeremy Howard
- A Brief History of the Open Source AI Hacker - with Ben Firshman of Replicate
- Best of 2024: Synthetic Data / Smol Models, Loubna Ben Allal, HuggingFace [LS Live! @ NeurIPS 2024]
- In the Arena: How LMSys changed LLM Benchmarking Forever
- 3 more episodes that year, every mention in 2024 →
2023 1 mention in 1 episode
every mention on every show, scene by scene, with the transcript →
14 statements about VLLM, every show
Burger: vLLM Runs on 500,000 GPUs at Any Moment
“Today we're here with Simon Moe, co-founder of Infraact, and a lead maintainer of VLLM, the open source inference engine, now running on half a million GPUs at any moment.”
Simon Mo: vLLM supports over 1,000 model architectures
“For VRM, we support more than a thousand model architecture up to today, and a lot of those are proprietary, but also a lot of those are open-weight, right?”
Mo: Major chipmakers use vLLM as an internal benchmark
“And additionally, VLM also work closely with all the hardware vendors. So that means across like NVIDIA, AMD, Google, and Amazon, Intel, and a lot more, their newest chip will make sure VLM can run on them. And then a lot of cases they use VLM as a benchmark t…”
Lambert: Kernel differences between vLLM and Hugging Face cause RL numerical instability
“VLLM and HuggingFace use different kernels to do the actual internal computation of the model. So these kernels are the things that make things like vLLM really fast. But these things, this then results in subtle numerical differences between the completions t…”
Ermon: Inception Labs built proprietary engine for production inference traffic
“So just like you would normally serve an LLM using a VLLM or SGLang or a Tensor or TLLM, we have built our own inference engine. And so we are supporting production traffic already with our own inference engine. We support continuous batch and quantization.”
Mohan: Codeium would not have survived if it relied on vLLM
“If we use VLLM, we would not be talking with you right now.”
Morris: vLLM and SGLang are here to stay and will grow more complex
“I also think, ah, VLLM and SGLang seem, like, really good and important and here to stay. Like, they'll probably just get larger and more complex to accommodate future systems”
Lattner: Modular MAX is 'more open source' than vLLM
“This thing's more open source than VLM because VLM depends on all these crazy binary CUDA kernels and stuff like this that are just opaque blobs from NVIDIA, right?”
Lattner calls vLLM a 'hot mess' due to too many stakeholders
“VLM seems much more like a massive community with a lot of stakeholders, a lot of stuff going on, and it's kind of a hot mess.”
Zhang: SGLang outperforms vLLM and has better usability than TensorRT-LLM
“I think for the common use case, maybe not, not the DeepSeq VIII, for the common use case, I think SGLAN's performance is better than FLM, and its usability is better than TensorFlow TLM.”
Zhang: vLLM does not support DeepSeek MLA while SGLang does
“Something like DeepSeq V-II, they proposed attention parent named MLA, multi-latent attention, and I think SGLAN is the only framework to support that. Maybe LightLM and TRTM also support, but VLM doesn't support.”
Zhang: SGLang achieves higher cache hit rates using block size of one
“Redix cache, I think it's the technology of the prefix caching, and it is a special case for something like block size is one, you know, for VLM or for other frameworks, they use something block size, 32, and SGLAN use the block size one. I think if you use th…”
Varun Mohan: Codeium would have failed if it used vLLM
“If we use VLLM, we would not be talking with you right now.”