The Ledger
Every statement that passed quotation and attribution checks. Mix any filter with any other: certainty 1/5, debate potential 5/5, or both at once.
why aren't all 21 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 0 checkable ones are still open, waiting for their date. predictions held up or didn't; assertions are supported or contradicted. on every card: ▮▮▮▮▮ certainty · ▮▮▮▮▮ debate potential. speakers are clickable
Sun: Synthetic data matches real-world data for multimodal model pre-training
“We were actually generating a lot of synthetic data and showing that, hey, you can actually, these synthetic data are actually as useful as real-world data when it comes to multimodal pre-training.”
Cerebras WSE-3 runs Llama inference 70x faster than NVIDIA GPUs
“Cerebris came out that the wafer scale engine three can serve llama 70 B at 2.1 thousand sorry, 202,100 tokens per second and serves llama four or five B at nearly 1000 tokens per second. So this, you know, to give you an understanding, like this is about 70 t…”
Houston: Groq and Cerebras outperform Nvidia on latency
“There's also, like, non-NVIDIA stacks, like the Grok, or Cerebris, or some of these custom silicon companies that are super interesting, and all, and outperformed the NVIDIA stack in terms of latency and things like that.”
Prakash predicts up to 5 million AI GPUs will sell in 2024
“There is four to five million GPUs that will be sold this year. NVIDIA and others.”
Patel: Nvidia manufactured 400k H100s last quarter and will sell 530k this quarter
“There's 400 to 500,000 being, 400,000 manufactured last quarter, and like five 30,000 this quarter being sold, right, of H-Hundreds”
Patel: Nvidia will sell over 3 million GPUs in 2024
“NVIDIA is going to sell well over three million, you know, total GPUs next year. You know, over a million H 100 this year alone, right?”
TurboQuant is inefficient on high-bandwidth data center GPUs like NVIDIA B200
“TurboQuant would not be like, it would not be used. Like Nvidia made it clear that this is not a good optimization. And we've seen it firsthand where the overhead of doing dequantization, quantization of You know, in the kernel itself, the turbo-quant kernel, …”
Ethan He: Megatron MoE was first to train trillion-parameter MoEs at 40% MFU
“The Megatron MOEs was the first It was the first framework open source to be able to train these MOEs at very large scales, like a hundred billion parameters to even trillion parameters efficiently at like 40% MFU.”
Parakhin: Bing Sydney first launched in India using Megatron, not OpenAI
“The funny thing, I mean, the most interesting anecdote is that Sydney was first shipped in India for and it was not noticed for a long time. And first implementation of Sydney didn't even have open AI model under it. It was during Megatron. Microsoft and the N…”
NVIDIA announces Rubin CPX as a dedicated prefill-specific hardware accelerator
“And like with our future generations, generations of hardware, we actually announced like with Rubin, this new accelerator that is pre-fill specific. It's called Rubin CPX.”
Jensen Huang prioritizes strategic investments in 'zero billion dollar markets'
“Jensen, He says, we're completely happy investing in zero billion dollar markets. We don't care if this creates revenue. It's important for us to know about this market. We think it will be important in the future. It can be zero billion dollars for a while.”
AMD's Composable Kernel library offers functionality similar to Nvidia's CUTLASS
“And then on the AMD side, they have this composable kernel library that does something very similar.”
Sohmers: NVIDIA's TF32 is actually a 19-bit precision format
“NVIDIA's TF-thirty-two number format is a nineteen-bit number format. They just call it thirty-two-bit.”
Ethan He: Hugging Face's sequential GEMM loop for Mixtral is inefficient
“Let's also look at the implementation of Mixtro eight by seven on Hagen-Phys transformer. You will soon notice the, in the expert operation there, You would iterate over all of the experts and compute each of the gem operations one by one. We found that this i…”
Fanelli: Singapore accounted for 15% of NVIDIA's Q3 2024 revenue
“Singapore was 15% of NVIDIA's revenue in Q three of 2024.”
Hotz: Tinygrad is about 5x slower than PyTorch on Nvidia GPUs
“The correctness for both forwards and backwards passes is there, but on Nvidia, it's about five X slower than PyTorch right now.”
NVIDIA Cosmos uses 50,000 to 60,000 tokens for five seconds of video
“Yeah, for example, like in Cosmos, I think just five seconds of video is like a 50, 50 K or a 60 K number of tokens. So like, if you do 50 seconds as a 500 K tokens, if you do longer than that, easily explode.”
NVIDIA chip design begins three to five years before market release
“The design process starts like- Exactly. ...three to five years before the chip gets to the market.”
NVIDIA Dynamo dynamically sizes and schedules Kubernetes prefill and decode workers
“Dynamo has a set of components that A, tell you how to scale. It tells you how many pre-fill workers and decoded workers it thinks you should have. And also provides a scheduling API for Kubernetes that allows you to actually represent and affect this scheduli…”
NVIDIA's 600M Parameter Parakeet Model Tops Speech Transcription Leaderboards
“And then Parakeet is Nvidia's new speech model, speech transcription model. That's number one on the leaderboards. And it's just like very enterprise tuned, like really, really rock solid, reliable at fairly small number of weights, like six hundred million pa…”