IBM and NVIDIA Partnership Cuts Data Mart Refresh Time from 15 to 3 Minutes

IBMIBM

IBM and NVIDIA will integrate GPU-native acceleration into IBM watsonx.data’s Presto engine, cutting Nestlé’s global data mart refresh from 15 to three minutes and delivering 83% cost savings plus a 30× price-performance boost. The collaboration brings NVIDIA Blackwell Ultra GPUs to IBM Cloud in Q2 2026 for GPU-optimized AI deployments.

1. Collaboration Overview

IBM and NVIDIA have expanded their alliance to help enterprises move AI from pilot to production by embedding GPU acceleration across data analytics, intelligent document processing, on-premises regulated infrastructure, cloud deployments and consulting services.

2. Nestlé Proof-of-Concept Delivers Speed and Savings

In a production test using IBM watsonx.data’s Presto engine accelerated by NVIDIA cuDF, Nestlé’s Order-to-Cash data mart refresh time dropped from 15 minutes to three minutes, achieving 83% cost savings and a 30× price-performance improvement on terabyte-scale data across 44 tables in 186 countries.

3. Infrastructure and Cloud Rollout

IBM will deploy NVIDIA Blackwell Ultra GPUs on IBM Cloud in early Q2 2026, certify IBM Storage Scale System 6000 for GPU-intensive workloads, and integrate GPU-native computing into Red Hat AI Factory to support on-premises and regulated AI deployments.

Sources

FMIF