?? Announcing the world’s fastest AI platform SambaNova Cloud runs Llama 3.1 405B @ 132t/s at full precision ? Llama 3.1 405B @ 132 tokens/sec ? 3.1 70B @ 570 tokens/sec ? Faster Inference than GPUs Start developing #FastAI ? https://cloud.sambanova.ai AI at Meta
This is really fast and cool. Tried it and compared with Groq - great results (considering model is served with full precision)!
Hi, our team at https://RAGDrive.com have shipped out a version that now supports SambaNova Systems as our very first inference provider. Would love to talk to your team more about this: [email protected]
Congratulations on the launch of SambaNova Cloud, the world's fastest AI platform! It's impressive to see Llama 3.1 405B running at 132 tokens/sec at full precision, and even more so to know that 3.1 70B runs at 570 tokens/sec. It's exciting to see that SambaNova Cloud offers faster inference than GPUs, which will undoubtedly benefit many businesses and organizations. I'm looking forward to exploring the platform and seeing how it can help me develop fast AI solutions. Thank you for sharing this exciting news! #FastAI
Thank you for such a great news! Is the speed of generating content important? In practice right content is! can't wait to test the product!!!
Fantastic. Can't wait to try this on AI Agents.
Insightful
Impressive performance stats SambaNova Systems Exciting to see SambaNova pushing the boundaries of AI with such speed and precision.
Setting a new bar for 405B and 70B — developers are going to hit the ground running FAST with SambaNova Cloud today! ??