The infrastructure powering IBM's Gen AI model development
Authors:
Talia Gershon,
Seetharami Seelam,
Brian Belgodere,
Milton Bonilla,
Lan Hoang,
Danny Barnett,
I-Hsin Chung,
Apoorve Mohan,
Ming-Hung Chen,
Lixiang Luo,
Robert Walkup,
Constantinos Evangelinos,
Shweta Salaria,
Marc Dombrowa,
Yoonho Park,
Apo Kayi,
Liran Schour,
Alim Alim,
Ali Sydney,
Pavlos Maniotis,
Laurent Schares,
Bernard Metzler,
Bengi Karacali-Akyamac,
Sophia Wen,
Tatsuhiro Chiba
, et al. (122 additional authors not shown)
Abstract:
AI Infrastructure plays a key role in the speed and cost-competitiveness of developing and deploying advanced AI models. The current demand for powerful AI infrastructure for model training is driven by the emergence of generative AI and foundational models, where on occasion thousands of GPUs must cooperate on a single training job for the model to be trained in a reasonable time. Delivering effi…
▽ More
AI Infrastructure plays a key role in the speed and cost-competitiveness of developing and deploying advanced AI models. The current demand for powerful AI infrastructure for model training is driven by the emergence of generative AI and foundational models, where on occasion thousands of GPUs must cooperate on a single training job for the model to be trained in a reasonable time. Delivering efficient and high-performing AI training requires an end-to-end solution that combines hardware, software and holistic telemetry to cater for multiple types of AI workloads. In this report, we describe IBM's hybrid cloud infrastructure that powers our generative AI model development. This infrastructure includes (1) Vela: an AI-optimized supercomputing capability directly integrated into the IBM Cloud, delivering scalable, dynamic, multi-tenant and geographically distributed infrastructure for large-scale model training and other AI workflow steps and (2) Blue Vela: a large-scale, purpose-built, on-premises hosting environment that is optimized to support our largest and most ambitious AI model training tasks. Vela provides IBM with the dual benefit of high performance for internal use along with the flexibility to adapt to an evolving commercial landscape. Blue Vela provides us with the benefits of rapid development of our largest and most ambitious models, as well as future-proofing against the evolving model landscape in the industry. Taken together, they provide IBM with the ability to rapidly innovate in the development of both AI models and commercial offerings.
△ Less
Submitted 13 January, 2025; v1 submitted 7 July, 2024;
originally announced July 2024.
Communication-Aware Energy Efficient Trajectory Planning with Limited Channel Knowledge
Authors:
D. Bonilla Licea,
M. Bonilla,
M. Ghogho,
S. Lasaulce,
V. S. Varma
Abstract:
Wireless communications is nowadays an important aspect of robotics. There are many applications in which a robot must move to a certain goal point while transmitting information through a wireless channel which depends on the particular trajectory chosen by the robot to reach the goal point. In this context, we develop a method to generate optimum trajectories which allow the robot to reach the g…
▽ More
Wireless communications is nowadays an important aspect of robotics. There are many applications in which a robot must move to a certain goal point while transmitting information through a wireless channel which depends on the particular trajectory chosen by the robot to reach the goal point. In this context, we develop a method to generate optimum trajectories which allow the robot to reach the goal point using little mechanical energy while transmitting as much data as possible. This is done by optimizing the trajectory (path and velocity profile) so that the robot consumes less energy while also offering good wireless channel conditions. We consider a realistic wireless channel model as well as a realistic dynamic model for the mobile robot (considered here to be a drone). Simulations results illustrate the merits of the proposed method.
△ Less
Submitted 18 November, 2020;
originally announced November 2020.
