IBM Launches Power S1112 Server for On-Premise AI Inferencing

IBM (NYSE: IBM) has launched the new Power S1112 server for on-premise AI inferencing, just days after reporting preliminary Q2 2026 revenue of $17.2 billion and adjusted EPS of $2.93, both missing estimates and triggering a single-day stock plunge of over 25%. The server is aimed at enterprises that prefer running AI models locally rather than in the cloud.

The Product — Key Features

The Power S1112 is designed for running generative AI and inferencing workloads on-premise. It leverages IBM Power processors and offers high performance with low power consumption. The server supports open-source frameworks such as PyTorch and TensorFlow.

Pricing and Availability

IBM has not yet announced official pricing. The server is expected to be available for early ordering in Q3 2026, with commercial availability and pricing details to follow.

Competition

The Power S1112 competes with on-premise servers from Dell, HPE, and Nvidia, which offer similar AI inferencing solutions. IBM differentiates itself through integration with its software stack, including Red Hat OpenShift.

Potential Impact on IBM

The launch could help boost IBM's hardware and software revenue as enterprises increasingly seek on-premise solutions for data privacy. However, its near-term impact on the stock may be limited given the recent earnings disappointment.