Loading video...

Video Failed to Load

Go Home

I trained a 12M parameter LLM on my own ML framework using a Rust backend and CUDA kernels for flash attention, AdamW, and more. Wrote the full transformer architecture, and BPE tokenizer from scratch. The framework features: - Custom CUDA kernels (Flash Attention, fused LayerNorm, fused GELU) for 3x...

MankyDankyBanky's profile picture

Aadi Kulshrestha

3,630 subscribers

820,861 views • 4 months ago •via X (Twitter)

EducationScience & Technology
LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

I implemented Google Research's TurboQuant as a CUDA-native compression engine on Blackwell B200. 5x KV cache compression on Qwen 2.5-1.5B, near-loseless attention scores, generating live from compressed memory. 5 custom cuTile CUDA kernels ft: - fused attention (with QJL corrections) - online softmax -on-chip cache decompression - pipelined TMA loads Try it out: s/o Bryce, the CUDA Colonel and the cuTile team at NVIDIA for lending me Blackwell GPU access :) cc sunny madra Gavin
4:46

I implemented Google Research's TurboQuant as a CUDA-native compression engine on Blackwell B200. 5x KV cache compression on Qwen 2.5-1.5B, near-loseless attention scores, generating live from compressed memory. 5 custom cuTile CUDA kernels ft: - fused attention (with QJL corrections) - online softmax -on-chip cache decompression - pipelined TMA loads Try it out: s/o Bryce, the CUDA Colonel and the cuTile team at NVIDIA for lending me Blackwell GPU access :) cc sunny madra Gavin

ani

813,740 views • 5 months ago

Check out mistral.​rs, our #Rust-based open source inference engine allowing for fast #LLM serving for a variety of architectures including X-LoRA mixture-of-expert (MoE) models, Llama-3, Mistral/Mixtral, Gemma & many others. Built on the Hugging Face #Candle framework for #Rust w/ custom CUDA kernels in the backend (as well as support for Metal, Apple Accelerate, and Intel MKL for CPU use), you can easily create a REST API OpenAI compatible server or run via Python bindings. Key features include: ✅Prefix caching, continuous batching ✅Flash Attention V2 ✅Device offloading ✅GGUF or Hugging Face models ✅2, 3, 4, 5, 6 and 8 bit quantization ✅X-LoRA MoE non-granular scalings for fast inference ✅Grammar support ✅Continuous batching ✅LoRA support with weight merging ✅LlamaIndex 🦙 integration ...and much more. Incorporation into our GraphReasoning multi-agent modeling framework & LlamaIndex 🦙 allows you to combine in-context learning with adversarial agentic strategies, to dive deep into complex scientific analyses, such as to predict material behaviors, generate hypotheses, analyze papers and data, develop new research concepts, and much more. Check out mistral.​rs: Join our Discord here: Rust Trending Rust Language
0:35

Check out mistral.​rs, our #Rust-based open source inference engine allowing for fast #LLM serving for a variety of architectures including X-LoRA mixture-of-expert (MoE) models, Llama-3, Mistral/Mixtral, Gemma & many others. Built on the Hugging Face #Candle framework for #Rust w/ custom CUDA kernels in the backend (as well as support for Metal, Apple Accelerate, and Intel MKL for CPU use), you can easily create a REST API OpenAI compatible server or run via Python bindings. Key features include: ✅Prefix caching, continuous batching ✅Flash Attention V2 ✅Device offloading ✅GGUF or Hugging Face models ✅2, 3, 4, 5, 6 and 8 bit quantization ✅X-LoRA MoE non-granular scalings for fast inference ✅Grammar support ✅Continuous batching ✅LoRA support with weight merging ✅LlamaIndex 🦙 integration ...and much more. Incorporation into our GraphReasoning multi-agent modeling framework & LlamaIndex 🦙 allows you to combine in-context learning with adversarial agentic strategies, to dive deep into complex scientific analyses, such as to predict material behaviors, generate hypotheses, analyze papers and data, develop new research concepts, and much more. Check out mistral.​rs: Join our Discord here: Rust Trending Rust Language

Markus J. Buehler

73,581 views • 2 years ago

>build a 1.5B VLA on top of MiniCPM-V 4.6 >compete with π0.5 and Qwen-VLA >use roughly half the visual-processing compute >remember previous steps without blowing up inference cost >pull the network plug >let the robot keep tracking — fully local, fully onboard >with CUDA Graph optimization and custom Triton fused kernels, MiniCPM-Robot inference throughput goes from 10 Hz → 33 Hz, and up to 36 Hz on the same NVIDIA H20 >release the models and framework as open source so why does robotics still rely on bigger models and the cloud? OpenBMB just dropped MiniCPM-Robot at WAIC two models + one inference framework 🧵
0:27

>build a 1.5B VLA on top of MiniCPM-V 4.6 >compete with π0.5 and Qwen-VLA >use roughly half the visual-processing compute >remember previous steps without blowing up inference cost >pull the network plug >let the robot keep tracking — fully local, fully onboard >with CUDA Graph optimization and custom Triton fused kernels, MiniCPM-Robot inference throughput goes from 10 Hz → 33 Hz, and up to 36 Hz on the same NVIDIA H20 >release the models and framework as open source so why does robotics still rely on bigger models and the cloud? OpenBMB just dropped MiniCPM-Robot at WAIC two models + one inference framework 🧵

arc.

48,279 views • 1 month ago

The latest MLX has a CUDA back-end! To get started: pip install "mlx[cuda]" With the same codebase you can develop locally, run your model on Apple silicon, or in the cloud on Nvidia GPUs. MLX is designed around Apple silicon - which has a unified memory architecture. It uses the same design with CUDA. So there's no need to move arrays around from CPU memory to GPU memory. Note, this is early days - some operations are missing and performance is still being optimized. But it's already quite fast for Transformer training, text generation, and more! Here's a demo using mlx-lm to generate text with Llama 3 8B (bf16) on an A100:
0:20

The latest MLX has a CUDA back-end! To get started: pip install "mlx[cuda]" With the same codebase you can develop locally, run your model on Apple silicon, or in the cloud on Nvidia GPUs. MLX is designed around Apple silicon - which has a unified memory architecture. It uses the same design with CUDA. So there's no need to move arrays around from CPU memory to GPU memory. Note, this is early days - some operations are missing and performance is still being optimized. But it's already quite fast for Transformer training, text generation, and more! Here's a demo using mlx-lm to generate text with Llama 3 8B (bf16) on an A100:

Awni Hannun

42,810 views • 1 year ago

Just launched #CES2026, the new open-source NVIDIA Nemotron Speech ASR model is here to solve latency drift and redundant compute. Its cache-aware streaming architecture eliminates the need for buffered inference, giving you stable, sub-100ms latency (24ms median T-T-F) and up to 3x more throughput on your GPU. 🤗 Read the technical blog with real-world results from Daily and Modal on Hugging Face:
1:24

Just launched #CES2026, the new open-source NVIDIA Nemotron Speech ASR model is here to solve latency drift and redundant compute. Its cache-aware streaming architecture eliminates the need for buffered inference, giving you stable, sub-100ms latency (24ms median T-T-F) and up to 3x more throughput on your GPU. 🤗 Read the technical blog with real-world results from Daily and Modal on Hugging Face:

NVIDIA AI Developer

138,518 views • 8 months ago

Congrats to the Kimi.ai team! This is awesome. Great to see this level of research coming from open-source frontier model labs. I liked the paper so much I built a Rust implementation of it ;) Full AttnRes + Block AttnRes with two-phase inference, built using Burn (tensor library and Deep Learning Framework, in Rust, by Tracel AI). Runs on CPU, CUDA, Metal, wgpu. Includes an interactive TUI that trains a model live and visualizes depth attention evolving from uniform to selective in real time. Repo link and more on what is implemented in the comments.
0:43

Congrats to the Kimi.ai team! This is awesome. Great to see this level of research coming from open-source frontier model labs. I liked the paper so much I built a Rust implementation of it ;) Full AttnRes + Block AttnRes with two-phase inference, built using Burn (tensor library and Deep Learning Framework, in Rust, by Tracel AI). Runs on CPU, CUDA, Metal, wgpu. Includes an interactive TUI that trains a model live and visualizes depth attention evolving from uniform to selective in real time. Repo link and more on what is implemented in the comments.

abdel

95,651 views • 6 months ago

Inkling, Thinking Machines' first open model, dropped today: 975B total / 41B active MoE, up to 1M context, reasoning natively over text, images, and audio. Serving and RL support are already live: you can run and shape it on an open stack, starting now. Day 0 support on SGLang SGLang and Miles RadixArk👇 - Inkling's new architecture (ShortConv, attention with relative positional embedding, shared expert sink MoE) is natively implemented and deeply optimized, with prefill full CUDA graph and MXFP8 KV cache - Full parameter and LoRA RL in a customized Megatron backend, train inference consistency via customized kernels, routing replay, and cross-runtime parameter synchronization - DFlash speculative decoding from Modal for low-latency serving Launching now, blog and cookbook in the comments ⬇️
1:01

Inkling, Thinking Machines' first open model, dropped today: 975B total / 41B active MoE, up to 1M context, reasoning natively over text, images, and audio. Serving and RL support are already live: you can run and shape it on an open stack, starting now. Day 0 support on SGLang SGLang and Miles RadixArk👇 - Inkling's new architecture (ShortConv, attention with relative positional embedding, shared expert sink MoE) is natively implemented and deeply optimized, with prefill full CUDA graph and MXFP8 KV cache - Full parameter and LoRA RL in a customized Megatron backend, train inference consistency via customized kernels, routing replay, and cross-runtime parameter synchronization - DFlash speculative decoding from Modal for low-latency serving Launching now, blog and cookbook in the comments ⬇️

LMSYS Org

146,298 views • 1 month ago

Ben Thompson explains how LLMs greatly diminished Nvidia's CUDA moat even as they sent the stock to the moon "So the weird thing about large language models is they were obviously incredible for Nvidia. That's why their stock went to the moon." "They have been on and off the most valuable company in the world." "It was also very bad for Nvidia. And the reason it was bad for Nvidia is that the play with CUDA is to build a developer ecosystem on top of CUDA." "But CUDA only works on Nvidia GPUs. So you get CUDA for free. It's easier to use, and it's a tremendous investment. Nvidia almost went under trying to build CUDA at a time when no one understood what they were doing or why they were wasting money on it." "And that's why Jensen Huang will get bristly, particularly when people question their rent-seeking or profit, whatever. It's like, no, they earned their spot fair and square." "Absolutely. It shouldn't be forgotten. They have earned every dollar they've gotten through 25 years of taking massive risks." "It bottomed out in October 2022. I wrote an article like three weeks before ChatGPT came out, tracing their bottoming-out history and their search for what was next." "'Nvidia in the Valley.' So, go back to this GTC. So I wrote an article at the time called 'Nvidia Waves and Moats'." "And what was interesting about that GTC was, number one, it was very boring. All the cool stuff kind of got scrubbed out." "Now, Jensen Huang has brought that stuff back, so the last few GTCs he's more talking about other things. Now it comes across as, oh, you're still looking for something beyond the LLM." "Because the problem with the LLM is it shifts the developer platform far above where Nvidia sits. All the activity is happening on top of LLMs. And so no one who's writing an AI application today is using CUDA." "Now, some people are, if you're training your own model and you're doing some low-level things or non-LLM things." "But the vast majority of the energy and all the money and the ecosystem is far removed from CUDA." "They have no idea and don't need to know or care what chips their application is running on. They're just on the OpenAI API, or the Anthropic API, or using Bedrock on Amazon, and it's sitting on Trainium, and they're using a Chinese open-source model. It's totally abstracted away, and this is why LLMs were bad for Nvidia." "Now, again, all the money they made along the way is worth it, but their moat has been tremendously diminished." "CUDA is still a moat if you need to do stuff that requires CUDA. But the vast majority of stuff, in energy, doesn't require CUDA, like in a post-LLM world."
3:04

Ben Thompson explains how LLMs greatly diminished Nvidia's CUDA moat even as they sent the stock to the moon "So the weird thing about large language models is they were obviously incredible for Nvidia. That's why their stock went to the moon." "They have been on and off the most valuable company in the world." "It was also very bad for Nvidia. And the reason it was bad for Nvidia is that the play with CUDA is to build a developer ecosystem on top of CUDA." "But CUDA only works on Nvidia GPUs. So you get CUDA for free. It's easier to use, and it's a tremendous investment. Nvidia almost went under trying to build CUDA at a time when no one understood what they were doing or why they were wasting money on it." "And that's why Jensen Huang will get bristly, particularly when people question their rent-seeking or profit, whatever. It's like, no, they earned their spot fair and square." "Absolutely. It shouldn't be forgotten. They have earned every dollar they've gotten through 25 years of taking massive risks." "It bottomed out in October 2022. I wrote an article like three weeks before ChatGPT came out, tracing their bottoming-out history and their search for what was next." "'Nvidia in the Valley.' So, go back to this GTC. So I wrote an article at the time called 'Nvidia Waves and Moats'." "And what was interesting about that GTC was, number one, it was very boring. All the cool stuff kind of got scrubbed out." "Now, Jensen Huang has brought that stuff back, so the last few GTCs he's more talking about other things. Now it comes across as, oh, you're still looking for something beyond the LLM." "Because the problem with the LLM is it shifts the developer platform far above where Nvidia sits. All the activity is happening on top of LLMs. And so no one who's writing an AI application today is using CUDA." "Now, some people are, if you're training your own model and you're doing some low-level things or non-LLM things." "But the vast majority of the energy and all the money and the ecosystem is far removed from CUDA." "They have no idea and don't need to know or care what chips their application is running on. They're just on the OpenAI API, or the Anthropic API, or using Bedrock on Amazon, and it's sitting on Trainium, and they're using a Chinese open-source model. It's totally abstracted away, and this is why LLMs were bad for Nvidia." "Now, again, all the money they made along the way is worth it, but their moat has been tremendously diminished." "CUDA is still a moat if you need to do stuff that requires CUDA. But the vast majority of stuff, in energy, doesn't require CUDA, like in a post-LLM world."

Fireside Alpha

12,980 views • 21 days ago

Ubuntu 26.04 LTS, codenamed #ResoluteRaccoon, is now available to download. 🦝 Resolute Raccoon builds on the resilience-focused improvements introduced in interim releases, with TPM-backed full-disk encryption, improved support for application permission prompting, Livepatch updates for Arm-based servers, and Rust-based utilities for enhanced memory safety. This release also brings native support for industry-leading AI/ML toolkits like NVIDIA CUDA and AMD ROCm, making Ubuntu 26.04 LTS the ideal platform for AI development and production workloads. Install now: Learn more about the release:
1:20

Ubuntu 26.04 LTS, codenamed #ResoluteRaccoon, is now available to download. 🦝 Resolute Raccoon builds on the resilience-focused improvements introduced in interim releases, with TPM-backed full-disk encryption, improved support for application permission prompting, Livepatch updates for Arm-based servers, and Rust-based utilities for enhanced memory safety. This release also brings native support for industry-leading AI/ML toolkits like NVIDIA CUDA and AMD ROCm, making Ubuntu 26.04 LTS the ideal platform for AI development and production workloads. Install now: Learn more about the release:

Ubuntu

653,887 views • 4 months ago

Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. Subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.
1:24

Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. Subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.

Alexander Whedon

13,483,053 views • 4 months ago

Compute has always been core to the Verne Robotics story. Getting there meant negotiating long, expensive contracts with big neoclouds, and projecting compute needs years into the future. Verne wasn't ready for that. Then came the Y Combinator cluster, built on Together AI. Verne trained a model 3x larger, on almost 100x more data. More data in, bigger foundation models out.
1:10

Compute has always been core to the Verne Robotics story. Getting there meant negotiating long, expensive contracts with big neoclouds, and projecting compute needs years into the future. Verne wasn't ready for that. Then came the Y Combinator cluster, built on Together AI. Verne trained a model 3x larger, on almost 100x more data. More data in, bigger foundation models out.

Together AI

21,649 views • 18 days ago

Finally, I'm done with the basics of backend engineering on the all-in-one resources to learn backend engineering I'm working on This basic section includes - Internet - HTTP - Servers - Web Dev fundamentals - Fundamentals of Operating Systems - A server-side language -> JavaScript | Go | Rust | Node - A framework -> Express | Nestjs | Django | Laravel | Spring Boot This is the boring part. Don't worry. We are getting into the fun part soon. Still working on it, updating it, Adding more, and making it better for you. Follow me. Retweet. Comment "Backend." I will DM you the PDF version when it's out. Enjoy the web version, and send in your feedback on how we can improve it for everyone.
0:37

Finally, I'm done with the basics of backend engineering on the all-in-one resources to learn backend engineering I'm working on This basic section includes - Internet - HTTP - Servers - Web Dev fundamentals - Fundamentals of Operating Systems - A server-side language -> JavaScript | Go | Rust | Node - A framework -> Express | Nestjs | Django | Laravel | Spring Boot This is the boring part. Don't worry. We are getting into the fun part soon. Still working on it, updating it, Adding more, and making it better for you. Follow me. Retweet. Comment "Backend." I will DM you the PDF version when it's out. Enjoy the web version, and send in your feedback on how we can improve it for everyone.

Solomon Eseme

92,216 views • 3 years ago

Mixtral 8x7B Instruct with AWQ & Flash Attention 2 🔥 All in ~24GB GPU VRAM! With the latest release of AutoAWQ - you can now run Mixtral 8x7B MoE with Flash Attention 2 for blazingly fast inference. All in < 10 lines of code. The only real change except loading AWQ weights is to pass attn_implementation="flash_attention_2" over to the .from_pretrained call whilst loading the model. Here's a full run through: 1. Install AutoAWQ and transformers pip install autoawq git+ com/huggingface/transformers.git 2. Initialise the tokeniser and the model from transformers import AutoModelForCausalLM, AutoTokenizer, TextStreamer model_id = "casperhansen/mixtral-instruct-awq" tokenizer = AutoTokenizer.from_pretrained(model_id) model = AutoModelForCausalLM.from_pretrained( model_id, low_cpu_mem_usage=True, device_map="cuda:0", attn_implementation="flash_attention_2") 3. Initialise the TextStreamer streamer = TextStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True) 4. Tokenise the inputs tokens = tokenizer( text, return_tensors='pt' ).input_ids.to("cuda:0") 5. Generate! generation_output = model.generate( tokens, streamer=streamer, max_new_tokens=512 ) That's it! 🤗
1:38

Mixtral 8x7B Instruct with AWQ & Flash Attention 2 🔥 All in ~24GB GPU VRAM! With the latest release of AutoAWQ - you can now run Mixtral 8x7B MoE with Flash Attention 2 for blazingly fast inference. All in < 10 lines of code. The only real change except loading AWQ weights is to pass attn_implementation="flash_attention_2" over to the .from_pretrained call whilst loading the model. Here's a full run through: 1. Install AutoAWQ and transformers pip install autoawq git+ com/huggingface/transformers.git 2. Initialise the tokeniser and the model from transformers import AutoModelForCausalLM, AutoTokenizer, TextStreamer model_id = "casperhansen/mixtral-instruct-awq" tokenizer = AutoTokenizer.from_pretrained(model_id) model = AutoModelForCausalLM.from_pretrained( model_id, low_cpu_mem_usage=True, device_map="cuda:0", attn_implementation="flash_attention_2") 3. Initialise the TextStreamer streamer = TextStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True) 4. Tokenise the inputs tokens = tokenizer( text, return_tensors='pt' ).input_ids.to("cuda:0") 5. Generate! generation_output = model.generate( tokens, streamer=streamer, max_new_tokens=512 ) That's it! 🤗

Vaibhav (VB) Srivastav

128,912 views • 2 years ago

I added KV caching and INT8 KV quantization to our transformer inference, improving throughput by 35x. All of this was done from scratch in Rust + CUDA, on top of a homemade ML framework. On a 4-token prompt with 252 generated tokens: - Original: 0.76 tok/s - KV cache fp32: 27.21 tok/s - KV cache int8 (quantized): 27.29 tok/s Try it out yourself here: In practice: - KV caching gave us about a 35x end-to-end speedup - INT8 KV cache kept roughly the same speed as fp32 but cut KV cache memory by 3.78x FP32 cache used 4.5 MB in this run while the INT8 cache used only 1.19 MB This simple change to inference created a huge impact on performance. To learn more about the KV cache and other optimizations like this, check out the blog at
2:36

I added KV caching and INT8 KV quantization to our transformer inference, improving throughput by 35x. All of this was done from scratch in Rust + CUDA, on top of a homemade ML framework. On a 4-token prompt with 252 generated tokens: - Original: 0.76 tok/s - KV cache fp32: 27.21 tok/s - KV cache int8 (quantized): 27.29 tok/s Try it out yourself here: In practice: - KV caching gave us about a 35x end-to-end speedup - INT8 KV cache kept roughly the same speed as fp32 but cut KV cache memory by 3.78x FP32 cache used 4.5 MB in this run while the INT8 cache used only 1.19 MB This simple change to inference created a huge impact on performance. To learn more about the KV cache and other optimizations like this, check out the blog at

Reese Chong

53,026 views • 4 months ago

New course: Build and Train an LLM with JAX, built in partnership with Google and taught by Chris Achard. JAX is the open-source library behind Google's Gemini, Veo, and other advanced models. This short course teaches you to build and train a 20-million parameter language model from scratch using JAX and its ecosystem of tools. You'll implement a complete MiniGPT-style architecture from scratch, train it, and chat with your finished model through a graphical interface. Skills you'll gain: - Learn JAX's core primitives: automatic differentiation, JIT compilation, and vectorized execution - Build a MiniGPT-style LLM using Flax/NNX, implementing embedding and transformer blocks - Load a pretrained MiniGPT model and run inference through a chat interface Come learn this important software layer for building LLMs!
1:54

New course: Build and Train an LLM with JAX, built in partnership with Google and taught by Chris Achard. JAX is the open-source library behind Google's Gemini, Veo, and other advanced models. This short course teaches you to build and train a 20-million parameter language model from scratch using JAX and its ecosystem of tools. You'll implement a complete MiniGPT-style architecture from scratch, train it, and chat with your finished model through a graphical interface. Skills you'll gain: - Learn JAX's core primitives: automatic differentiation, JIT compilation, and vectorized execution - Build a MiniGPT-style LLM using Flax/NNX, implementing embedding and transformer blocks - Load a pretrained MiniGPT model and run inference through a chat interface Come learn this important software layer for building LLMs!

Andrew Ng

193,825 views • 6 months ago

If you have compute hungry AI ML tasks, you can spin up an Intel Gaudi2 instance and watch it just rip. I wrote a quick Python script to use the facebook/detr-resnet-50 model (🙏🏻 Hugging Face) to detect and label objects in a video. It's interesting that no one is talking about Intel Gaudi vs NVIDIA A100 or H100. Gaudi2 isn't cheap at $10.42/hour but it feels like a viable alternative to fighting someone for an A100 or H100. Here's my code if you want to try it out:
0:26

If you have compute hungry AI ML tasks, you can spin up an Intel Gaudi2 instance and watch it just rip. I wrote a quick Python script to use the facebook/detr-resnet-50 model (🙏🏻 Hugging Face) to detect and label objects in a video. It's interesting that no one is talking about Intel Gaudi vs NVIDIA A100 or H100. Gaudi2 isn't cheap at $10.42/hour but it feels like a viable alternative to fighting someone for an A100 or H100. Here's my code if you want to try it out:

Ryan Carson

41,328 views • 2 years ago

Cerebras just had the biggest IPO of the year. Founder Andrew Feldman says the 3 most important things he had to convince investors of while doing the roadshow were that demand for inference is going to 1,000,000x, the GPU isn't the only way to do compute, and that the CUDA moat is overstated. What he said: "Jensen said some time ago on Brad Gerstner's podcast that the demand for inference will grow by a 1,000,000x, and nobody believed him. And at the same time, you saw Sam Altman displaying real vision and going out and trying to lock up huge amounts of compute, memory, data centers, and power, because he saw it too." "[We tried] to share what that means — what exponential demand means. And that we're still so early, and yet the demand for AI compute is overwhelming." "The other thing is that there are lots of ways to do this. The GPU isn't the only way. You've got TPUs, Trainium, and us. There are lots of different ways to build a solution here." "And finally — the notion that CUDA is this grand lock-in is overplayed. Gemini 3, which is an excellent model, was trained on TPUs with no CUDA. The Anthropic models were trained on Trainium with no CUDA. Some of the best models, some of the most interesting things are being done without CUDA. And that lock-in might be overplayed." $CBRS
1:46

Cerebras just had the biggest IPO of the year. Founder Andrew Feldman says the 3 most important things he had to convince investors of while doing the roadshow were that demand for inference is going to 1,000,000x, the GPU isn't the only way to do compute, and that the CUDA moat is overstated. What he said: "Jensen said some time ago on Brad Gerstner's podcast that the demand for inference will grow by a 1,000,000x, and nobody believed him. And at the same time, you saw Sam Altman displaying real vision and going out and trying to lock up huge amounts of compute, memory, data centers, and power, because he saw it too." "[We tried] to share what that means — what exponential demand means. And that we're still so early, and yet the demand for AI compute is overwhelming." "The other thing is that there are lots of ways to do this. The GPU isn't the only way. You've got TPUs, Trainium, and us. There are lots of different ways to build a solution here." "And finally — the notion that CUDA is this grand lock-in is overplayed. Gemini 3, which is an excellent model, was trained on TPUs with no CUDA. The Anthropic models were trained on Trainium with no CUDA. Some of the best models, some of the most interesting things are being done without CUDA. And that lock-in might be overplayed." $CBRS

TBPN

36,474 views • 4 months ago

Qwen 3.8 Flash Next (125B-A6B) now runs natively in mlx-serve on Apple Silicon. No Python, Zig + Metal. (Preview Release) M4 Max, 4-bit pack (~75 GB resident): - ~70 tok/s serial, ~98 tok/s with MTP when tested using `npx llmprobe` - prefill ~700 tok/s, sparse attention past 2k tokens with fused kernels - images and video input supported Hyper-connections, n-gram embeddings gathered straight from disk, QSA block selection, 5 new Metal kernels. All open source, parity-tested against the HF reference. Kernel work: - hyper-connection read: 3 Metal kernels, split-K, write folded into the next read - GDN decode: 2 fused dispatches per layer - hd-256 attention: mask arm, sparse prefill + verify stay fused - all greedy-equivalent to the unfused path If you port any of this to your engine, a mention would be nice. MLX-Serve Preview 26.8.11 build: 4bit Speed Pack: 4-8bit Quality Pack: As for the model, i'm not sure yet, it feels verbose, and Opus like. Thinks a lot. Goes on long self testing loops/talking with itself, but need more time with it. It seems like a good model for agentic things.
0:25

Qwen 3.8 Flash Next (125B-A6B) now runs natively in mlx-serve on Apple Silicon. No Python, Zig + Metal. (Preview Release) M4 Max, 4-bit pack (~75 GB resident): - ~70 tok/s serial, ~98 tok/s with MTP when tested using `npx llmprobe` - prefill ~700 tok/s, sparse attention past 2k tokens with fused kernels - images and video input supported Hyper-connections, n-gram embeddings gathered straight from disk, QSA block selection, 5 new Metal kernels. All open source, parity-tested against the HF reference. Kernel work: - hyper-connection read: 3 Metal kernels, split-K, write folded into the next read - GDN decode: 2 fused dispatches per layer - hd-256 attention: mask arm, sparse prefill + verify stay fused - all greedy-equivalent to the unfused path If you port any of this to your engine, a mention would be nice. MLX-Serve Preview 26.8.11 build: 4bit Speed Pack: 4-8bit Quality Pack: As for the model, i'm not sure yet, it feels verbose, and Opus like. Thinks a lot. Goes on long self testing loops/talking with itself, but need more time with it. It seems like a good model for agentic things.

David Dalcu

121,176 views • 16 days ago

Excited to announce the new preview for Microsoft Foundry Agents 🎉! You can now build, run, and deploy your agent using any model, any framework, any harness in the cloud 🧑‍💻 - check out the demo below This is not just any cloud compute environment; it's an agent-optimized platform with: 🖥️ Persistent microVMs - securely scale up and down without losing context 🛠️ Built-in tools (1000+) 👀 Observability and evaluations 👷 Guardrails 🔐 Private networking... and more
8:10

Excited to announce the new preview for Microsoft Foundry Agents 🎉! You can now build, run, and deploy your agent using any model, any framework, any harness in the cloud 🧑‍💻 - check out the demo below This is not just any cloud compute environment; it's an agent-optimized platform with: 🖥️ Persistent microVMs - securely scale up and down without losing context 🛠️ Built-in tools (1000+) 👀 Observability and evaluations 👷 Guardrails 🔐 Private networking... and more

Jeff Hollan

157,946 views • 4 months ago

Ever wanted to quickly turn a PDF into clean text to paste into your favorite AI agent, without having use CLIs or open the browser? We built exactly that. Using Tauria Penny ps, with a Rust backend powered by LiteParse and a React frontend, we created a cross-platform desktop app that lets you: • Drag and drop PDFs • Convert them into clean Markdown • Preview page screenshots with extracted bounding boxes overlaid for easy inspection It's a simple way to understand your document's structure and verify the parsing results before handing them off to an LLM. Check out the demo below! 👇 GitHub: Get started with LiteParse:
0:31

Ever wanted to quickly turn a PDF into clean text to paste into your favorite AI agent, without having use CLIs or open the browser? We built exactly that. Using Tauria Penny ps, with a Rust backend powered by LiteParse and a React frontend, we created a cross-platform desktop app that lets you: • Drag and drop PDFs • Convert them into clean Markdown • Preview page screenshots with extracted bounding boxes overlaid for easy inspection It's a simple way to understand your document's structure and verify the parsing results before handing them off to an LLM. Check out the demo below! 👇 GitHub: Get started with LiteParse:

LlamaIndex 🦙

21,173 views • 2 months ago