Celebrating the Power 11 Family’s New Addition

Rob McNelly provides details on the recent S1112 announcement and reflects on his upcoming writing/blogging anniversary

There was a new server announcement that you might have missed:

“The IBM Power S1112 is a compact, one-socket Power11 server engineered for IBM i, AIX and Linux. It brings Power11 innovation to distributed and edge locations, delivering enterprise-grade performance and availability in an entry-class form factor.

“Optimized for IBM i entry-level deployments, the S1112 supports the P05 Software Tier, enabling lower licensing costs and a more economical path to Power11 adoption. Its design aligns with the needs of smaller IBM i environments while preserving the scalability and resiliency of the Power platform.

Beyond IBM i, the S1112 provides the flexibility to consolidate and grow mixed environments—running AIX, Linux, or hybrid workloads—on a single, high-performance system.”

You’ll want to read the whole thing, but I’ll share (in part) the S1112 feature summary:

• One entry single-chip processor module per system server:
  • 10-core Typical 3.05 to 4.0 Ghz (max) Power11 Processor (#EJSV) – Rack version only
  • 4-core Typical 3.60 to 4.0 Ghz (max) Power11 Processor (#EJMT) – Available in both rack and tower form factors, with support for deconfiguration to a single active processor core (feature #2319). This flexibility enables even the smallest customers to start with a cost-optimized footprint while maintaining a clear path for growth.

• Up to 512 GB of system memory distributed across 4 DDIMM slots per system server

• PCIe slots
  • Two x16 Gen4 half-height, half-length direct slots
  • Two x8 Gen4 half-height, half-length direct slots

• Storage: Four NVMe slots with 12.8 TB maximum capacity.

• Optional PCIe I/O expansion drawer with PCIe slots:
  • Up to one drawer (#ENZ0) – Rack version only
  • Each I/O drawer holds one six-slot PCIe fanout module (#ENZF)
  • Each fanout module attaches to the system node through a PCIe optical or copper cable adapter (#EJ24).

I do not know what exposure you have had with Power11 servers, but just the other day I was reminded of the noticeable difference between Power9 and Power11 when I performed a Live Partition Mobility operation on an LPAR. I could immediately see the difference in responsiveness, and this was even before shutting down the LPAR, changing the processor mode from Power9 to Power11 and rebooting.

The S1112 was just one of numerous July announcements, which are listed here. I’ll highlight the IBM i 7.6 Technology Refresh 2 and IBM i 7.5 Technology Refresh 8 along with AIX enhancements and Virtualization and platform enhancements. Also be sure to read Brandon Pederson’s blog post about IBM Power Autonomous Operations, his take on the S1112 announcement, IBM Bob, the Agentic Engine for IBM i, etc.

Recent IBM Support alerts

In less exciting but still important news, IBM Support recently issued these security alerts:

* Vulnerability AIX impacts due to cURL libcurl: A logic error allows libcurl to reuse an existing negotiate-authenticated connection even when a new request specifies different credentials for the same server.

* Multiple vulnerabilities in the curl/libcurl are related to unauthorized access.

* BIND vulnerability affects AIX/VIOS: A remote attacker can send or trigger processing of malformed BRID/HHIT DNS resource records, causing the named daemon to terminate unexpectedly.

* Multiple vulnerabilities in Python affects AIX/VIOS: Integer overflow and use after may cause memory corruption and denial of service.

* Vulnerability in Python affects AIX/VIOS: Use-after-free can cause denial of service.

* A vulnerability in zlib could allow a remote attacker to cause a denial of service. AIX uses zlib as part of its data compression functions.

* Vulnerabilities in AIX could cause denial of service, heap buffer overflow, crash and heap corruption, forge authentication, loss of confidentiality, credential exposure, bypass integrity protection and heap buffer overread. OpenSSL is used by AIX as part of AIX’s secure network communications.

* Multiple vulnerabilities in the curl/libcurl are related to use-after-free, reusing an invalid connection, bypass Public Suffix List checks, improper host key validation and invalid authorization.

Finally, here’s a pretty big one if you are running the affected fibre adapters:

“Some FC adapters may not recover from temporary SAN errors.

“Certain fibre channel adapters may not recover from SAN errors that would normally cause only temporary link down events. Affected adapters include any that use the cvfcdd driver. Identify if these adapters are in use by the VIOS or AIX partition by running the following as root:

lsdev -c adapter | grep fcs | grep 7710

or

odmget -q ddins=pci/cvfcdd CuDv

“This issue can occur when SFP monitoring or ‘fcstat -e’ command is executed before a link transition (link up or link down). As a result, environments that regularly run ‘fcstat -e’ to collect adapter statistics are at higher risk of encountering the problem.

“Once triggered, the adapter remains in a link-down state and recovery may require an LPAR reboot.

As a temporary workaround until the corrective fix is applied, avoid running ‘fcstat -e’ and SFP-monitoring utilities on systems equipped with affected adapters.”

Keep VUG-ing Along

A periodic reminder to check out the IBM Power Virtual User Group monthly meetings. Even if you can’t catch the live events, you can always dive into the replays. There have been some very good VUG events in the past few months; I recommend these in particular:

* April 2026: Meet IBM Bob with Tim Rowe

* March 2026: Top Tips for Successful Administration of your VIO servers with Jaqui Lynch

* January 2026: NIM 201 with Jaqui Lynch

* September 2025: Advanced Power11 Features with Todd Boyd

A Happy, Summery Anniversary

On a personal note, this is the time of year that I celebrate my first AIXchange post, for what was at the time known as IBM Systems Magazine. My blog, which was updated weekly, went live on July 16, 2007.

And it is a celebration for me. While at that time I’d been a contributor to the magazine’s website for three years, getting the chance to write regularly has opened many doors. Next year will be 20 years of regular writing for me, and seven years since the transition to TechChannel and transitioning to a monthly column.

Regardless of the start date I choose, I will always be grateful that so many of you have followed me on this journey. I don’t plan on slowing down any time soon.

---