_version_ 1866908665950240768
author DeBole, Michael V.
Appuswamy, Rathinakumar
McGlohon, Neil
Taba, Brian
Esser, Steven K.
Akopyan, Filipp
Arthur, John V.
Amir, Arnon
Andreopoulos, Alexander
Carlson, Peter J.
Cassidy, Andrew S.
Datta, Pallab
Flickner, Myron D.
Gandhasri, Rajamohan
Garreau, Guillaume J.
Ito, Megumi
Klamo, Jennifer L.
Kusnitz, Jeffrey A.
McClatchey, Nathaniel J.
McKinstry, Jeffrey L.
Nayak, Tapan K.
Otero, Carlos Ortega
Penner, Hartmut
Risk, William P.
Sawada, Jun
Sivagnaname, Jay
Smith, Daniel F.
Sousa, Rafael
Terrizzano, Ignacio
Ueda, Takanori
Gray-Donald, Trent
Cox, David
Modha, Dharmendra S.
author_facet DeBole, Michael V.
Appuswamy, Rathinakumar
McGlohon, Neil
Taba, Brian
Esser, Steven K.
Akopyan, Filipp
Arthur, John V.
Amir, Arnon
Andreopoulos, Alexander
Carlson, Peter J.
Cassidy, Andrew S.
Datta, Pallab
Flickner, Myron D.
Gandhasri, Rajamohan
Garreau, Guillaume J.
Ito, Megumi
Klamo, Jennifer L.
Kusnitz, Jeffrey A.
McClatchey, Nathaniel J.
McKinstry, Jeffrey L.
Nayak, Tapan K.
Otero, Carlos Ortega
Penner, Hartmut
Risk, William P.
Sawada, Jun
Sivagnaname, Jay
Smith, Daniel F.
Sousa, Rafael
Terrizzano, Ignacio
Ueda, Takanori
Gray-Donald, Trent
Cox, David
Modha, Dharmendra S.
contents A vertically integrated, end-to-end, research prototype system combines 288 NorthPole neural inference accelerator cards, offline training algorithms, a high-performance runtime stack, and a containerized inference pipeline to deliver a scalable and efficient cloud inference service. The system delivers 115 peta-ops at 4-bit integer precision and 3.7 PB/s of memory bandwidth across 18 2U servers, while consuming only 30 kW of power and weighing 730 kg in a 0.67 m^2 42U rack footprint. The system can run 3 simultaneous instances of the 8-billion-parameter open-source IBM Granite-3.3-8b-instruct model at 2,048 context length with 28 simultaneous users and a per-user inter-token latency of 2.8 ms. The system is scalable, modular, and reconfigurable, supporting various model sizes and context lengths, and is ideal for deploying agentic workflows for enterprise AI applications in existing data center (cloud, on-prem) environments. For example, the system can support 18 instances of a 3-billion-parameter model or a single instance of a 70-billion-parameter model.
format Preprint
id arxiv_https___arxiv_org_abs_2511_15950
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Scalable NorthPole System with End-to-End Vertical Integration for Low-Latency and Energy-Efficient LLM Inference
DeBole, Michael V.
Appuswamy, Rathinakumar
McGlohon, Neil
Taba, Brian
Esser, Steven K.
Akopyan, Filipp
Arthur, John V.
Amir, Arnon
Andreopoulos, Alexander
Carlson, Peter J.
Cassidy, Andrew S.
Datta, Pallab
Flickner, Myron D.
Gandhasri, Rajamohan
Garreau, Guillaume J.
Ito, Megumi
Klamo, Jennifer L.
Kusnitz, Jeffrey A.
McClatchey, Nathaniel J.
McKinstry, Jeffrey L.
Nayak, Tapan K.
Otero, Carlos Ortega
Penner, Hartmut
Risk, William P.
Sawada, Jun
Sivagnaname, Jay
Smith, Daniel F.
Sousa, Rafael
Terrizzano, Ignacio
Ueda, Takanori
Gray-Donald, Trent
Cox, David
Modha, Dharmendra S.
Distributed, Parallel, and Cluster Computing
Artificial Intelligence
Hardware Architecture
A vertically integrated, end-to-end, research prototype system combines 288 NorthPole neural inference accelerator cards, offline training algorithms, a high-performance runtime stack, and a containerized inference pipeline to deliver a scalable and efficient cloud inference service. The system delivers 115 peta-ops at 4-bit integer precision and 3.7 PB/s of memory bandwidth across 18 2U servers, while consuming only 30 kW of power and weighing 730 kg in a 0.67 m^2 42U rack footprint. The system can run 3 simultaneous instances of the 8-billion-parameter open-source IBM Granite-3.3-8b-instruct model at 2,048 context length with 28 simultaneous users and a per-user inter-token latency of 2.8 ms. The system is scalable, modular, and reconfigurable, supporting various model sizes and context lengths, and is ideal for deploying agentic workflows for enterprise AI applications in existing data center (cloud, on-prem) environments. For example, the system can support 18 instances of a 3-billion-parameter model or a single instance of a 70-billion-parameter model.
title A Scalable NorthPole System with End-to-End Vertical Integration for Low-Latency and Energy-Efficient LLM Inference
topic Distributed, Parallel, and Cluster Computing
Artificial Intelligence
Hardware Architecture
url https://arxiv.org/abs/2511.15950