Viewing profile — helloericsf

Wow, didn't know there were talks about relaxing GDPR. Can you share a few links? Many thanks.

Thanks for sharing. I bet their DPO and EU customers are super interested in the findings. The CEO should have handled it better, IMO.

Not a DBA, how do you do DB permission rollout gating?

Thanks for reaching out. Just reposted.

If you're in SF, you don't want to miss this. The Qwen team is making their first public appearance in the United States, with the VP of Qwen Lab speaking at the meetup below durin…

The base plan limit is not hard to hit. Then you're on the usage based rocket.

Honestly depends on when they got in. Seed investors? They're probably fine with their preferences. Series B and beyond? That's where it gets messy. What round you thinking?

How does it stack up against the new Grok 4 model?

3 repos DualPipe https://github.com/deepseek-ai/DualPipe EPLB https://github.com/deepseek-ai/eplb profile-data https://github.com/deepseek-ai/profile-data

Github: https://github.com/deepseek-ai/DeepGEMM - Up to 1350+ FP8 TFLOPS on Hopper GPUs - No heavy dependency, as clean as a tutorial - Fully Just-In-Time compiled - Core logic at …

HF: https://huggingface.co/Wan-AI/Wan2.1-T2V-14B Github: https://github.com/Wan-Video/Wan2.1 Wan2.1, a comprehensive and open suite of video foundation models that pushes the bound…

this might help: https://x.com/main_horse/status/1894215779521794058/photo/1

- Efficient and optimized all-to-all communication - Both intranode and internode support with NVLink and RDMA - High-throughput kernels for training and inference prefilling - Low…

They don't have h100. wink,wink.

Don't think the decision is based on infra, or any technical reasons. It's more on the service support side. How a 200-person company supports 44M iPhone users in China?

What do you mean by "lower"? To my understanding, they will open 5 infra related repos this week. Let's revisit your comparison question on Friday.

X: https://x.com/deepseek_ai/status/1893836827574030466 BF16 support Paged KV cache (block size 64) 3000 GB/s memory-bound & 580 TFLOPS compute-bound on H800