90/30 Club Reading Raptor: The First 3D-DRAM Accelerator for Generative Inference

​Special edition of 9030! Nithesh Kurella and Prashant Nair will be joining us to present their recent paper: "Early Silicon of Raptor: The First 3D-DRAM Accelerator for Generative Inference"

Paper link: https://ramyadhadidi.github.io/files/dMatrix-Raptor-ISCA.pdf

​The paper discusses d-Matrix’s Raptor, an accelerator that tackles a key bottleneck in generative AI: moving data between memory and compute. Drawing on early silicon results, we’ll explore how stacking DRAM directly with accelerator logic enables up to 100 TB/s of memory bandwidth per card—and the engineering challenges of making it power-efficient and reliable.

​Event Schedule:

​7pm-8pm: Quiet reading time, grab a snack and read! (optional)

​8pm-9pm: Group discussion about the paper 📝

​9pm-10pm: We have our space for a bit longer, stay to socialize or network!

​Our event is hosted within Mox SF, the gracious donors of the space where we will meet.