Glossary
IBM Power AI Infrastructure Glossary
Hardware-first IBM i AI terms: Power workload placement, inference partitions, ENZ0 expansion drawers, Matrix Math Acceleration, IBM i, and Db2 data proximity.
A
- AI Inference Partition ... A separate supported operating-system partition used to execute an AI model while an IBM i partition may remain the production application and data source.
- AI Workload Placement ... The decision about which system, operating environment, and partition executes an AI model or agent, separately from the IBM i application supplying its data.
- Artificial Intelligence Unit ... The IBM Research prototype chip that became the commercial Spyre Accelerator. Thirty-two cores, never sold.
- AS/400 ... IBM's original 1988 midrange business computer platform; the term is still used informally to refer to IBM i systems and shops even though IBM has rebranded the hardware and OS multiple times since.
C
- Capacity on Demand (CoD) ... Capacity on Demand is IBM's way to switch on processor cores and memory units that are already installed in a Power server but sitting inactive. For an IBM i shop it is how you give a new AI inference partition cores and memory without a hardware install and, for a permanent activation, without a server restart.
- Commercial Processing Workload (CPW) ... IBM's relative capacity rating for processors running IBM i, and the number that shows how much transaction headroom a Power server has left. IBM quotes it for the largest IBM i partition a machine allows, which on several Power 11 processor options is capped at 4 cores.
- Continuous batching ... Continuous batching lets one Spyre Accelerator serve many requests at once by handing an open seat to a waiting request the moment an earlier one finishes, instead of holding a fixed group together until the slowest one ends. For an IBM i shop it is the setting that decides how many people a single card can answer before anyone waits in line.
D
- Db2 for i ... The relational database management system built into every IBM i operating system, tightly integrated with IBM i rather than a separately installed database product.
- Differential DIMM (DDIMM) ... A DDIMM is the memory card an IBM Power 11 or Power 10 server takes in place of a plain DDR memory stick. It carries its own buffer chip and plugs into an Open Memory Interface slot, which is how a two-socket Power S1124 reaches the 2456 GBps memory bandwidth ceiling IBM lists for feeding data into an AI model.
- Dynamic Logical Partitioning ... Dynamic logical partitioning, or DLPAR, moves processors, memory and card slots in and out of a running IBM Power partition without restarting it. IBM documents it for the Spyre Accelerator, so an AI card can join or leave a Linux inference partition while the IBM i production partition keeps working.
I
- IBM Granite ... IBM's own family of open language models, published under the Apache 2.0 licence so they can be downloaded, tuned, and run on hardware you own.
- IBM i ... IBM's operating system for Power servers running traditional AS/400-lineage business applications, with Db2 for i built directly into the OS.
- IBM Open-Source AI Foundation for Power ... Supplies the ready-made, containerized AI services an IBM i shop deploys on a Spyre-equipped Power server, ordered from IBM as program number 5639-SAI. IBM lists it next to Red Hat AI Inference Server as a required part of a typical Spyre inference setup, so it is a line on the order rather than an optional download.
- IBM Power E1150 ... The Power E1150 is the midrange IBM Power 11 server, machine type 9043-MRU, and IBM lists it for 8, 12, 16 or 24 Spyre AI accelerator cards ... the same card ceiling as the smaller S1124. It is also the one Power 11 model an IBM i shop cannot move production onto, because IBM does not support IBM i on it.
- IBM Power E1180 ... The Power E1180 is the high-end IBM Power 11 server, machine type 9080-HEU, and the only Power model IBM lists for more than 24 Spyre AI accelerator cards ... up to 48 per system. An IBM i shop can keep production on IBM i 7.4 TR12 or later and run large AI inference in a Red Hat Enterprise Linux partition on the same machine.
- IBM Power S1122 ... The Power S1122 is the two-socket 2U IBM Power 11 server, machine type 9824-22A, and IBM lists it for 8 or 12 Spyre AI accelerator cards ... half the ceiling of the larger S1124. For an IBM i shop it is the smaller Power 11 route into on-site model inference, with one trap: only the ERGR processors let an IBM i partition grow past 4 cores.
- IBM Power S1124 ... The Power S1124 is the largest scale-out IBM Power 11 server, machine type 9824-42A, and IBM lists it for 8, 12, 16 or 24 Spyre AI accelerator cards ... but only when both processor modules are installed. An IBM i shop can keep production on IBM i 7.4 TR12 or later and run model inference in a Red Hat Enterprise Linux partition on the same 4U machine.
- IBM Spyre Enablement Stack for Power ... The drivers, runtime and models IBM requires before a Spyre Accelerator card runs anything on a Power 11 server. It arrives from Red Hat inside Red Hat AI Inference Server, so the card sits idle until the Linux partition that drives it has the stack installed and entitled.
- Inference serving ... The software layer that takes requests for a model, packs them onto the hardware efficiently, and returns answers. It decides how much work one accelerator card can actually do.
- InstructLab ... An open source project from IBM Research and Red Hat that lets a subject matter expert add knowledge to a model without being a data scientist.
N
- Nanosheet transistor ... The gate-all-around transistor design IBM demonstrated at 2 nm in 2021, now the industry's leading-edge architecture.
- NanoStack ... IBM's 0.7 nm transistor architecture, announced June 25, 2026. The first published transistor below one nanometre.
- NorthPole ... An IBM Research chip with no off-chip memory at all, claiming 25 times the energy efficiency of a conventional architecture. Research only, not for sale.
- NVMe U.2 ... NVMe U.2 is the 2.5 inch flash drive bay that holds every bit of internal storage on a Power 11 server, wired as native PCIe straight to the processor. For an IBM i shop staging data for AI inference it settles two things: how much flash fits inside the box, and whether your one expansion drawer goes to drive bays or to Spyre Accelerator cards.
P
- PagedAttention ... PagedAttention is the memory manager inside the vLLM serving engine. It hands out an AI model's working memory in small fixed blocks instead of one worst-case reservation, which is what lets one Spyre Accelerator card in a Power 11 server answer many requests at once instead of a few.
- PCIe Gen5 ... PCIe Gen5 is the slot generation that sets how fast an adapter card can talk to an IBM Power processor, at 32 GT/s per lane. For an IBM i shop sizing AI work it matters for one narrow reason: every Gen5 slot IBM documents on Power 11 is eight lanes wide, so it carries the same 64 GB/s as a Gen4 sixteen lane slot.
- PowerVM ... IBM's hypervisor for Power servers, built into system firmware. It is what lets a Linux partition running an AI model sit on the same machine as the production IBM i partition, each with its own processors, memory and cards.
- ppc64le ... ppc64le is the architecture code for 64-bit little endian Linux on IBM Power, and it is the only kind of build that will run on a Power server. Every part of an AI stack, from the serving engine to the Spyre Accelerator drivers, has to exist as a ppc64le build before an IBM i shop can serve a model in a Linux partition on the same machine.
- Prefix caching ... Prefix caching lets a Spyre Accelerator skip re-reading the opening part of a prompt it has already read. For an IBM i shop whose requests all begin with the same long instruction block, it is the setting that stops the card from paying for that block on every single request.
- Processor compatibility mode ... The setting on each partition that decides which generation of Power processor features the operating system inside it is allowed to use. For an IBM i shop it is the reason a Linux AI partition on a new Power 11 server can still be running as a Power10 or POWER9 machine, which changes what the inference software underneath can reach.
R
- Red Hat AI Inference Server ... Red Hat's model serving product, built on the open source vLLM engine. IBM lists it as one of two supported ways to drive a Spyre Accelerator on IBM Power.
- Red Hat Enterprise Linux AI ... A bootable Red Hat Enterprise Linux image that arrives with IBM Granite models and the InstructLab tuning tools already fitted together.
- Red Hat OpenShift AI ... Red Hat's larger AI platform, and the second option IBM accepts for driving a Spyre Accelerator on IBM Power.
S
- Shared Processor Pool ... A group of Power cores that several partitions draw from, with one ceiling over the whole group. It is how an IBM i shop gives a Linux AI partition real processor capacity on the production server while holding both that partition and the IBM i license count under a fixed limit.
- SMT8 (Simultaneous Multithreading) ... SMT8 is IBM's name for running eight hardware threads at once on one Power processor core. For an IBM i shop it sets how much work an inference partition can hold, and IBM i controls it with the QPRCMLTTSK system value and the QWCCHGPR API.
- Spyre Accelerator ... IBM's PCIe AI inference card. On IBM Power 11 it carries 32 cores, 128 GB of memory, and more than 300 TOPS inside a 75 watt envelope.
T
- Telum II ... IBM's z17 mainframe processor, with an AI accelerator on the chip itself. It does not go in a Power server, but it is where IBM's current AI hardware argument was proven.
- Tensor parallelism ... Tensor parallelism splits one AI model across several Spyre Accelerator cards so the cards answer a single request together. On IBM Power 11 it is the setting that decides whether a model fits at all and how long a prompt it can take, and IBM publishes verified setups at only one, two or four cards per model.
- Time to first token ... Time to first token, or TTFT, measures the wait between sending a question to an AI model and seeing the first word of the answer come back. For an IBM i shop it is the number that says whether local AI on a Power 11 server with a Spyre Accelerator feels instant or feels broken, because it counts queue time and the whole prompt read, not just the writing.
- TOPS ... Trillions of operations per second. A rough headline rating for AI hardware, useful for comparison and misleading on its own.
V
- Virtual I/O Server ... VIOS is a dedicated IBM Power partition that shares one set of physical disk, Fibre Channel and Ethernet adapters out to every other partition on the server. For an IBM i shop that means a Linux partition running an AI model can get its storage and its network connection from VIOS instead of buying its own physical cards.
- vLLM ... The open source model serving engine inside Red Hat AI Inference Server, known for PagedAttention and continuous batching.