Cover Image for Cafe Compute Meetup: Dubai
Cover Image for Cafe Compute Meetup: Dubai
Avatar for Cerebras
Presented by
Cerebras
Cerebras makes the world's biggest chip, powering fast inference up to 15x faster than GPUs.

Cafe Compute Meetup: Dubai

Register to See Address
Dubai, United Arab Emirates
Registration
Approval Required
Your registration is subject to host approval.
Welcome! To join the event, please register below.
About Event

​Cerebras is excited to partner with Cognition to host a Cafe Compute Meetup

​Enjoy the fast vibes, great food, cool people, and more.

​Bring your laptop. This is a space to code, collab, and build in a cozy evening vibe.

​The event runs from 6 pm to 9 pm sharp. We'd love for you to be there from the start!

​6:00 pm - Doors open, grab something to drink and settle in

​6:30 pm - Quick intro to Cerebras + Cognition and what we're building today

​7:30 pm - Code, experiment, ask questions

​8:45pm - Open networking and wrap up

​9:00pm - That's a wrap!

​Cerebras is the world's fastest AI inference, up to 15x faster than leading GPUs. Cerebras Inference is powered by our Wafer-Scale Engine (WSE-3) - the world's largest AI chip. Experience the speed for yourself with OpenAI Codex-Spark, powered by Cerebras, or one of the open-source models like GLM 4.7. Get free compute at Cerebras.ai.

​Join our rapidly growing communities:
Discord ↗
Reddit ↗
X ↗

Cognition is the applied AI lab behind Devin, the AI software engineer, and Windsurf, the AI-native IDE. Devin takes on real engineering tasks end-to-end, from writing and testing code to opening pull requests, while Windsurf brings agentic coding directly into your editor. Experience it for yourself at the cafe: build with Devin and Windsurf, see how agents fit into a real dev workflow, and meet other people building with AI. Try it at devin.ae

Join our communities:
Join WhatsApp Group
DevinAI

Location
Please register to see the exact location of this event.
Dubai, United Arab Emirates
Avatar for Cerebras
Presented by
Cerebras
Cerebras makes the world's biggest chip, powering fast inference up to 15x faster than GPUs.