STAC Summit New York – Bitfusion’s Elastic GPU Infrastructure in Finance

July 12, 2018

Bitfusion, the Elastic AI Software Company, presented its Elastic GPU software platform and use cases for the finance industry at the recent STAC Summit New York. STAC Summits bring together industry leaders in architecture, app dev, infrastructure, engineering, and operational intelligence to discuss important technical challenges in the finance industry.

AI is becoming a key element of applications across the finance sector, from trading and portfolio management to fraud detection to underwriting and beyond. As these applications proliferate and a plethora of new applications emerge, AI infrastructure leveraging GPUs and FPGAs is evolving to provide leaps ahead in performance, productivity and efficiency based on virtualization and elasticity.

With Bitfusion, GPU farms can reach top performance for risk management machine learning (ML) workloads. The Bitfusion platform can assign on-demand ML training and inference workloads to available GPUs and partial GPUs, increasing GPU utilization to 90% or more. In addition, with Bitfusion there is no need for data migration from the compute farm to the GPU farm, resulting in enhanced security and data governance.

STAC members can access Bitfusion’s STAC Summit presentation at https://stacresearch.com/spring2018NYC.

About Bitfusion
AI and machine learning are changing every aspect of the data center. GPUs (and soon newer purposed-built AI hardware) are deployed at scale in Enterprises to accommodate newer workloads. However, unlike storage, compute and networking, GPUs are deployed in a scattered, siloed and uncoordinated way across the enterprise. This drives negative metrics of productivity, profitability, CapEx, OpEx and agility. With Bitfusion, GPUs can be deployed as a shared and common pool, responding in real time to any demand of a machine learning workload. Bitfusion virtualizes a GPU cluster of any size by allowing any workload to remotely attach over Ethernet to a partial GPU, one or more GPUs anywhere in the cluster, for only the duration of the runtime code. See: www.bitfusion.io.

# # #

Bitfusion Media Contact: 


« Back to News & Events