v1.3.7 · 2026-09-05
The terms the site uses, each defined once.
Where sources define a term differently, the entry records the convention Scrutica uses. Related terms at the foot of an entry link to the ones this glossary also defines.
Extreme ultraviolet lithography uses 13.5 nm light to print circuit patterns on wafers. A tin-plasma light source and multilayer mirrors operate in a vacuum.
Deep ultraviolet lithography uses 193 nm argon-fluoride or 248 nm krypton-fluoride light. Multiple patterning adds exposures and processing steps to print finer features.
A manufacturer’s name for a semiconductor process generation, such as N4 or N7. The number in a modern node name does not specify a physical gate length or permit direct comparison between manufacturers.
A thin disc of semiconductor material on which chip dies are fabricated. Wafer starts per month measures how many wafers enter production.
The fraction of manufactured dies that meet the required specifications. Yield depends on the design and manufacturing process.
Chip-on-Wafer-on-Substrate: TSMC’s packaging family for connecting logic dies and high-bandwidth memory. CoWoS-S uses a silicon interposer; CoWoS-R uses redistribution layers, and CoWoS-L combines them with local silicon interconnects.
High Bandwidth Memory: stacks of DRAM dies connected vertically by through-silicon vias. Several stacks can supply memory to one accelerator package.
A semiconductor manufacturer that fabricates chips designed by customers. A customer must use a design compatible with the foundry’s process.
A chip company that designs chips and contracts out their fabrication. NVIDIA and AMD use this model.
Integrated Device Manufacturer: a company that both designs and fabricates chips. Intel and Texas Instruments are examples.
Outsourced Semiconductor Assembly and Test: a company that packages and tests fabricated chips. ASE and Amkor provide these services.
Fin Field-Effect Transistor: a transistor whose gate surrounds the sides and top of a raised silicon channel, called a fin.
Gate-All-Around transistor: a transistor whose gate surrounds the channel on all sides. Intel’s RibbonFET uses ribbon-shaped silicon channels.
A patterned template, also called a reticle, from which a lithography system projects circuit features onto a wafer.
Techniques that connect multiple dies within one package, including side-by-side integration and vertical stacking. They allow logic and memory manufactured separately to work together.
A floating-point arithmetic operation. FLOP/s measures the rate of these operations; multiplying a sustained rate by elapsed seconds gives a cumulative operation count.
One trillion (10¹²) floating-point operations per second. Hardware ratings specify the arithmetic format and whether structured sparsity contributes to the reported rate.
A 16-bit floating-point format with one sign bit, five exponent bits and ten fraction bits. Mixed-precision training can combine FP16 arithmetic with higher-precision operations.
A 16-bit floating-point format with one sign bit, eight exponent bits and seven fraction bits. Its exponent range matches FP32, with less precision for individual values.
Thermal Design Power: a manufacturer’s thermal design rating, in watts, used to size cooling. Its specification and operating conditions matter when using it as a power estimate.
NVIDIA’s interconnect for high-bandwidth communication between supported processors. Link bandwidth and the number of processors connected depend on the generation and system topology.
A network technology used for high-bandwidth, low-latency communication between servers, including servers in AI training clusters.
NVIDIA hardware units that accelerate matrix multiply-and-accumulate operations. Supported number formats and throughput vary by GPU generation.
The rate at which data can move between a processor and its memory, usually expressed in GB/s or TB/s.
NVIDIA’s module form factor for GPUs installed on a server baseboard. The power limit and available interconnect depend on the GPU and system configuration.
PCI Express: an interface for connecting devices within a computer. GPU cards can use PCIe alongside other links.
The operations used to train a model, commonly counted in FLOP. A reported total needs a defined scope, including which training stages it counts.
Using a trained model to produce an output. Compute demand depends on the model, input and output lengths, and how requests are processed together.
An empirical relationship between model performance and inputs such as parameter count, training data and compute. Kaplan et al. (2020) and Hoffmann et al. (2022) study such relationships for language-model loss.
A compute-constrained balance of model size and training data studied by Hoffmann et al. (2022). Their results favored scaling parameters and training tokens together, with about 20 tokens per parameter in the Chinchilla training run.
The number of learned weights in a model. The parameters used for an individual input can be a subset of the total in a mixture-of-experts model.
Model FLOP Utilization: model operations per second divided by the hardware’s theoretical peak rate. The model-operation count excludes activation recomputation; hardware FLOP utilization can include that extra work.
Further training of an existing model, often to adapt it to a task or dataset. It may update all model weights or a smaller set of parameters.
METR’s estimate of the task duration at which an AI agent reaches a specified success rate, commonly 50%. Duration refers to the time human experts need for the tasks.
Power Usage Effectiveness: total facility energy divided by IT equipment energy over the same period. IT equipment includes servers, storage and networking.
Megawatt: one million watts, a unit of power. Energy consumption also requires a duration; one megawatt sustained for one hour is one megawatt-hour.
Cooling that carries heat away in a liquid, through cold plates attached to components or by immersing equipment in a suitable fluid.
A group of connected GPU servers configured to work together. Distributed training depends on the accelerators available to a job and the communication between them.
The fraction of available capacity used over a stated period. For GPUs, the measure may refer to occupied device time or to a particular activity counter; the denominator needs to be specified.
The links and switches that connect processors and servers. A cluster may combine local accelerator links with an InfiniBand or Ethernet network between servers.
A cloud provider focused on supplying GPU computing capacity. CoreWeave, Lambda and Nebius offer such services.
An operator of large-scale computing infrastructure, often spread across many data centers. The term includes major cloud providers and companies that operate infrastructure for their own services.
A dependency in which only one supplier can meet the requirement for a particular input or qualified process. Other suppliers may serve the wider market.
A directed connection representing a recorded business relationship. A supplier–customer pair can have separate connections for different products or services.
Herfindahl-Hirschman Index: the sum of squared market shares, expressed as percentages. A single supplier has an HHI of 10,000; splitting supply among more similarly sized firms reduces the index.
A supply-chain dependency whose disruption is difficult to avoid because suitable alternatives are limited.
Disruption that spreads through business dependencies. Scrutica’s simulator follows recorded relationships and weights transmitted impact by supply share, criticality and a scenario-level decay factor.
A subscription source of business relationships. Its records can include supplier–customer links and attributes such as relationship value or price correlation.
A Pearson correlation between two companies’ stock prices over a stated window. Scrutica’s licensed relationship records can include a three-month value.
The ability to replace an input or supplier with an alternative that meets the required specifications. It depends on the product and the time available to qualify or expand replacement supply.
Export Administration Regulations: US rules administered by the Bureau of Industry and Security that govern covered exports, reexports and transfers, along with specified activities by US persons.
A list maintained by BIS whose entries specify licence requirements and review policies for transactions involving the named entities.
Bureau of Industry and Security: the US Department of Commerce agency that administers the EAR, issues licences and maintains the Entity List.
An authorization under the EAR for a transaction that meets specified conditions, without an individual licence application.
Under the EAR, a release of controlled technology or source code to a foreign person in the United States can constitute an export to that person’s relevant country of citizenship or permanent residency.
A multilateral arrangement for coordinating controls on conventional arms and dual-use goods and technologies. Participating states agree control lists and related guidance.
Export restrictions tied to the intended use of an item or specified activities. They can apply in addition to controls based on the item’s technical classification.
EAR rules that assess controlled US-origin content in a foreign-made item. The applicable percentage depends on the item and destination, with some categories subject to no minimum.
Tera operations per second: 10¹² operations per second. A quoted rate needs the operation type, precision and sparsity convention.
General-purpose AI models, as defined by the EU AI Act, have significant generality and can perform a wide range of distinct tasks, including when incorporated into other systems.
Under the EU AI Act, risk associated with a general-purpose model’s high-impact capabilities that can spread at scale across the value chain, with significant effects on matters such as safety or fundamental rights.
Regulation (EU) 2024/1689, which sets rules for AI systems and general-purpose AI models, with obligations that depend on the applicable category.
Policies that use access to computing resources, or information about their production and use, to influence AI development and deployment.
Goods, software or technology that can serve civilian and military purposes.
The spread of a capability to additional actors. In AI, this can involve access to models, computing resources or the expertise needed to use them.
National initiatives to develop AI capabilities, including domestic computing infrastructure and access arrangements with commercial providers.
Scrutica’s ratio of reported disbursed spending to announced government spending. When the government figure is absent, the calculation falls back to the headline commitment, which may include private capital.
An interval produced by a statistical procedure with a specified repeated-sampling coverage rate. A 95% procedure would contain the true parameter in 95% of repeated applications under its assumptions.
Information about where a value came from and how it was transformed. It can include a source document, the date it was accessed and the calculation applied.
A boolean field used to mark values represented as estimated or inferred. Its meaning depends on the record and the source’s treatment of that value.
Ranges obtained by varying assumptions in an estimate. Scrutica’s facility estimator varies inputs within each available hardware, power or cost calculation.
Breadth-first search explores a graph outward in stages from a starting point. Scrutica uses it to follow modeled disruptions through recorded business relationships.
In Scrutica’s cascade model, a scenario-level multiplier applied when transmitting disruption across a relationship. Supply share and criticality contribute additional multipliers.
Allocating capital expenditure among components such as buildings, power infrastructure and computing equipment.
One petaFLOP/s (10¹⁵ floating-point operations per second) sustained for 86,400 seconds: 8.64 × 10¹⁹ FLOP.
Cloud rental price per petaflop-day = hourly instance price × 24 ÷ (GPU count × dense BF16 TFLOP/s per GPU ÷ 1,000).