You Say You Wanted Evolution 

IBM Fusion began as a collection of data services for containerized applications.  However, it has since transformed into a comprehensive application data platform designed to support virtualization, hybrid cloud and AI/ML workloads. 

This evolution continues with the new release of Fusion 2.11, which offers multiple enhancements around operations, optimization and infrastructure. 

Let’s take a look a few of them! 

Let’s Get Flexible

Fusion 2.11 brings more flexibility to the configuration infrastructure. A base Fusion system now only requires three (3) servers in a rack – with either 32 or 64 GB of memory and up to ten (10) NVMe drives. These can serve as combined control/compute nodes and enable the addition of worker nodes for expandability. 

This also allows for more flexible multi-rack expansion options where an additional rack can be procured and installed with only a single server or GPU node. 

In addition, these new options facilitate a new Two Availability Zone (“2AZ”) topology. This allows two Fusion racks to be set up in a “stretched” configuration perfect for a high-availability deployment.

This adds further flexibility to virtual machines by providing them with other Fusion services such as H/A, backup and restore and Fusion Data Cataloging.

Enter SAN, Man

Fusion Access for SAN, which was in Tech Preview in the previous release is now production-ready in Fusion 2.11!

This solution provides a scalable clustered file system for applications running on Fusion Software in a bare-metal server environment with SAN attached storage. 

Developed primarily to scale with OpenShift Container Platform clusters and virtual machine workloads, Fusion Access for SAN can support up to approximately 3000 VMs on six (6) bare-metal hosts.

In addition, by leveraging the existing SAN infrastructure, Fusion Access for SAN can overcome the data transfer bottlenecks that can occur over a traditional LAN, especially for large volumes of data.

AI Will Always Love You

The new 2.11 release provides several enhancements designed to support the growing depth and breadth of the AI space. 

First, we’ve added support for the new NVIDIA H1200 GPUs. In addition to the NVIDIA L40S and the new Dell GPU server, clients now have more options when it comes to designing the configuration for any high resource-consumption application.

In addition, this new release also expands the breadth and scope of the new Content Aware Storage service that was introduced in the last release.  

The new Docling multi-modal infrastructure allows users to process documents without relying on the NVIDIA services. This allows for a streamlined integration of the RAG (retrieval augmented generation) pipeline and better control of the data being ingested. 
Finally, we’ve added some new security features in the CAS Service. CAS Access Control allows administrators to define access levels for specific users and user groups instead of having to redefine these values with every query.  And File level security adds a layer of protection to the ingest data, ensuring that users can only query on data they are authorized to read. 
 

Would You Like to Know More?

To learn more, join the Advanced Technology Group on October 16th for the “Accelerate with ATG Webinar – What’s New in Fusion 2.11” 

Click here to register for the ATG Webinar!

You can also learn more at IBM TechXchange, which will be held in Orlando, FL this year from October 6th – 9th .

This conference will feature in-depth in-person sessions around such topics as OpenShift Virtualization, Fusion Disaster Recovery and Fusion and AI. 

Click here to register for IBM TechXchange!