Work

1.536TB/s/mm2 Bandwidth Scalable Attention Accelerator with 22.5GOPS Throughput High Speed SoftMax for Quantized Transformers in Intel 3

Year: 2025

Type: conference-paper

DOI: 10.23919/vlsitechnologyandcir65189.2025.11075126

Authors: Prerna Budhkar, Srivatsa Rangachar Srinivasa, Mirco Sciulli, Gauthaman Murali, Ragh Kuttappa, Paolo Aseron, Trang Nguyen, Vinayak Honkote, Tanay Karnik

Institution: Intel (United States)

Language: English

Topic: 3D IC and TSV technologies

Subfield: Electrical and Electronic Engineering

Field: Engineering

Domain: Physical Sciences

SDG: Industry, innovation and infrastructure

Open Access status: closed

OpenAlex ID: W4412965838

FWCI: 2.358

Cites: 1

Cited by: 1

Related to: 10