Cerebras Systems

Meet Cerebras Inference – the fastest inference for generative AI! 🏎️ Speed: 1,800 tokens/sec for Llama 3.1-8B and 450 tokens/sec for Llama 3.1-70B, 20x faster than NVIDIA GPU-based hyperscale clouds. 💸 Price: Cerebras Inference offers the industry’s best price-performance at 10c per million tokens for Llama 3.1-8B and 60c per million tokens for Llama-3.1 70B. 🎯 Accuracy: Cerebras Inference uses native 16-bit weights for all models, ensuring the highest accuracy responses. 🔓 Access: Cerebras Inference is open to everyone today via chat and API access. All powered by our third-generation Wafer Scale Engine (WSE-3). Try it now 👉 https://lnkd.in/gEJJ2pfY Press Release: https://lnkd.in/gtF5fxHt Blog: https://lnkd.in/gZ46q4cD

Verified by Artificial Analysis, Cerebras Inference achieves 1,850 tokens/sec on Llama 3.1 8B and 450 tokens/sec on Llama 3.1 70B! By dramatically reducing processing time, we're enabling more complex AI workflows and enhancing real-time LLM intelligence. This includes a new class of intelligent agents that can “think faster” than ever before. Cerebras Inference will power a new era of Instant AI. 👉Try it today: https://lnkd.in/gEJJ2pfY 👉Read our blog: https://lnkd.in/gZ46q4cD 👉Check out Artificial Analysis for more data: https://lnkd.in/gY6NAFqW

🎙️Learn about Cerebras Inference on the Gradient Dissent Podcast! 🎙️ In the latest episode of Gradient Dissent, Andrew Feldman joins Lukas Biewald to discuss the incredible advancements we're making in AI Inference. Andrew dives deep into how Cerebras Inference sets new benchmarks in speed, accuracy, and cost efficiency. Don’t miss this insightful conversation! 🎧 🎙️ YouTube: https://lnkd.in/gxd9kjrw 🎙️ Apple: https://lnkd.in/gnNQtEjY Check out Weights & Biases' blog about Cerebras Inference: https://lnkd.in/gmfmyi88

Celebrating our product launch in Times Square NYC! 🚕 Join the high-speed inference party at inference.cerebras.ai 🎉

⚡Cerebras Announces DocChat: GPT-4 Level Conversational QA Trained in a Few Hours⚡ DocChat is our latest document-based conversational AI model series, that includes two models: 🦙 Cerebras Llama3-DocChat - An LLM that was trained to GPT-4 level document-based Q&A performance in a few hours 🐲 Cerebras Dragon-DocChat - A multi-turn retriever model that was fine-tuned in a few minutes DocChat models demonstrate top-of-the-line performance across major QA benchmarks and are fully open-source (model weights, training recipes, and datasets) Read more in our blog: https://lnkd.in/gQ27z_nq Check out the model here: https://lnkd.in/gQu-EuC4

🔬Cerebras Accelerates GenAI progress across Life Sciences and Healthcare   Generative AI is transforming the life sciences and healthcare industries. Cerebras is helping leaders like GSK and Mayo develop AI breakthroughs by training their own models for molecular design, drug discovery, and personalized medicine. Read our blog: https://lnkd.in/gDvTs636

🌍 Cerebras Wafer-Scale Engine Achieves 210x Speedup Over NVIDIA H100! Researchers from Rice University and TotalEnergies have demonstrated a groundbreaking 210x speedup using the Cerebras Wafer-Scale Engine (WSE) over the NVIDIA H100 GPU using a Finite-Volume solver. This type of solver is essential to many scientific computations such as carbon capture simulations. This advancement could be pivotal for designing geological carbon capture and storage projects, crucial for climate change mitigation. 🔍 Discover how the Cerebras WSE is redefining high-performance computing and paving the way for more effective climate solutions. https://lnkd.in/gMn5Qxir Research credit: Ryuichi Sai, Francois Hamon, PhD, John Mellor-Crummey, Mauricio Araya Polo

🌟 Join Us at the AI Hardware & Edge AI Summit 2024! 🌟 We're excited to be sponsoring this year's AI Hardware and Edge AI summit, taking place September 9-12 in San Jose, CA. Swing by our booth to experience groundbreaking AI innovations and meet the brilliant minds behind them! 🎤 Keynote Spotlight: Our CTO, Sean Lie, will be delivering a keynote on "Wafer-Scale AI: Unlocking GPU-Impossible Speed." This is your chance to see the future of AI technology in action! 🤝 Let’s Connect! Want to chat with us at the summit? Let us know! https://lnkd.in/gmWFBAKh

Cerebras Co-Founder Deconstructs NVIDIA Blackwell Delays From intricate interposer designs to alignment issues and thermal expansion complications, Cerebras Co-Founder and Chief System Architect Jean-Philippe Fricker provides a detailed look into the hurdles faced by GPU architectures as they try to go bigger. Watch the full interview here: https://lnkd.in/gaqxht5j

🚀 Cerebras Outperforms GPUs by 95x and CPUs by 570x in Stencil Computations Researchers at Rice University and TotalEnergies have developed the StencilPy framework, which optimizes stencil computations across various hardware architectures. On the Cerebras Wafer-Scale Engine (WSE), StencilPy not only delivers a 95x speedup over GPUs and a 570x speedup over CPUs, but it also reduces the code required by a factor of 7. The stencil simulation that produced these results was a high-order, 25-point star-shaped stencil used for seismic simulations, specifically for the acoustic isotropic approximation of the wave equation—crucial for applications like seismic imaging, weather forecasting, and computational fluid dynamics. Dive into our blog for more info: https://lnkd.in/gDSRZ9mV Ryuichi S. John Mellor-Crummey Jinfan Xu Mauricio Araya Polo