1,344 technical terms and definitions
amd instinct, amd gpu compute, radeon compute, rocblas, miopen, composable kernel, hipify, cdna architecture, mi300x rocm
ROCm is AMD's open-source GPU compute stack: the CUDA-equivalent layer of compiler, runtime and math libraries that turns an Instinct accelerator into a machine-learning device…
hipamd port cuda, rocm software stack, roofline model amd, amd mi300x gpu
HIP/ROCm AMD GPU Programming: CUDA Portability and MI300X — enabling GPU-agnostic code and AMD CDNA acceleration HIP (Heterogeneous Interface for Portability) enables…
GPU programming, AMD, portable
ROCm HIP GPU Programming is an open-source GPU programming framework enabling portable code development targeting both AMD and NVIDIA GPUs through unified application interface…
rohs 2 directive restricted substance compliance prediction, rohs directive restricted, directive restricted substance, restricted substance compliance, rohs directive
RoHS 2 Directive Restricted Substance Compliance Prediction with Machine Learning Introduction & Motivation RoHS 2 Directive Restricted Substance Compliance Prediction…
standards
RoHS compliance is the conformance to regulations restricting hazardous substances in electrical and electronic equipment - it is a fundamental requirement for market access…
rohs compliance heavy metal content prediction, rohs compliance heavy, compliance heavy metal, heavy metal content, metal content prediction, rohs compliance, compliance heavy
RoHS Compliance Heavy Metal Content Prediction with Machine Learning Introduction & Motivation RoHS Compliance Heavy Metal Content Prediction with Machine Learning addresses…
rohs compliance screening drift prediction, rohs compliance screening, compliance screening drift, screening drift prediction, rohs compliance, compliance screening
RoHS Compliance Screening Drift Prediction with Machine Learning Introduction & Motivation RoHS Compliance Screening Drift Prediction with Machine Learning addresses a central…
roi, business & strategy
ROI is return on investment, a ratio that compares net gain against total invested capital for a project or program - It is a core method in advanced semiconductor program…
roland, graph neural networks
Roland is a dynamic graph-learning approach for streaming recommendation and interaction prediction - Incremental representation updates handle new edges and nodes without full…
ai safety
Role-play jailbreaks is the jailbreak technique that frames harmful requests as fictional or character-based scenarios to bypass safety refusals - it exploits narrative framing…
prompting techniques
Role Prompting is a prompting technique that assigns the model a specific persona or expertise frame to shape response behavior - It is a core method in modern LLM workflow…
roll-to-roll flexible electronics defect prediction, roll-to-roll flexible electronics, flexible electronics defect, electronics defect prediction, roll-to-roll flexible
Roll-to-Roll Flexible Electronics Defect Prediction with Machine Learning Introduction & Motivation Roll-to-Roll Flexible Electronics Defect Prediction with Machine Learning…
revert, previous version
Rollback Rollback mechanisms provide an essential safety net for AI system deployments, enabling immediate reversion to a previously known stable version if production issues are…
quality & reliability
Rolled Throughput Yield is the compounded probability of a unit passing all sequential process steps without defects - It reveals cumulative quality loss across multi-step…
rty, production
Rolled throughput yield optimization is the system-level improvement of cumulative first-pass success across all process steps in a manufacturing chain - because RTY is…
rolled throughput yield, rty, production
Rolled Throughput Yield (RTY) is the cumulative probability of passing all process steps without defects — calculated by multiplying individual step yields, revealing true process…
time series models
Rolling Forecast is walk-forward forecasting where training and evaluation windows advance through time. - It simulates real deployment by repeatedly retraining or updating models…
rolling update pod disruption, rolling update pod, update pod disruption, rolling update, update pod, pod disruption
Rolling Update Pod Disruption Introduction & Motivation Rolling Update Pod Disruption addresses a central problem in cloud process control…
Expanding and Rolling Window Backtesting Protocols Introduction & Motivation Expanding and Rolling Window Backtesting Protocols addresses a central problem in timeseries…
rollup zk snark proof, rollup zk snark, zk snark proof, rollup zk, zk snark, snark proof
Rollup ZK SNARK Proof Introduction & Motivation Rollup ZK SNARK Proof addresses a central problem in blockchain process control and proof-verification-latency prediction: how…
flash, memory, array, design, read, access
ROM and Flash Memory Array Design and Read Access is the physical and circuit implementation of read-only and flash memories enabling high-density storage with efficient access…
nlp
Romanization is a specific type of transliteration where text is converted into the Roman (Latin) script — often used in NLP to standardize inputs from diverse scripts (Cyrillic…
rome, model editing
ROME is the Rank-One Model Editing method that updates selected transformer weights to modify a targeted factual association - it is a prominent single-edit approach…
optimization
The roofline model is a one-picture performance framework: it plots attainable compute throughput against arithmetic intensity, so you can see at a glance whether a kernel is…
compute bound, memory bound, performance model
Roofline Model — a visual framework for understanding whether a computation is limited by compute throughput or memory bandwidth, guiding optimization efforts. The Model Where: OI…
roofline performance, compute bound memory bound, roofline gpu, performance modeling
Roofline Model Analysis is the visual performance modeling framework that plots achievable performance (FLOP/s) against arithmetic intensity (FLOP/byte) to determine whether…
arithmetic intensity, compute bound memory bound, roofline analysis, performance ceiling
The Roofline Model is the visual performance analysis framework that plots achievable computation throughput (FLOPS) against arithmetic intensity (FLOPS/byte) — creating…
compute bound memory bound, arithmetic intensity analysis, roofline gpu cpu, operational intensity optimization
Roofline Model Performance Analysis is the visual performance modeling framework that characterizes the performance ceiling of a compute kernel as limited by either computational…
memory bound vs compute bound, operational intensity, hpc optimization roofline, flops vs memory bandwidth
The Roofline Performance Model is the universally adopted graphical heuristic utilized by supercomputing architects and software optimization engineers to visually diagnose…
audio & speech
Room Simulation is acoustic augmentation that simulates reverberation and room impulse responses - It exposes models to realistic far-field and reverberant conditions during…
quality & reliability
Root Cause is the underlying process, design, or system condition that directly enables a problem to occur - It identifies what must be changed to prevent recurrence. What Is Root…
Causal Root Cause Analysis and Fault Attribution Introduction & Motivation Causal Root Cause Analysis and Fault Attribution addresses a central problem in causal inference…
root-cause analysis cycle time prediction, root-cause analysis cycle, analysis cycle time, cycle time prediction, root-cause analysis, analysis cycle, cycle time, time prediction
Root-Cause Analysis Cycle Time Prediction with Machine Learning Introduction & Motivation Root-Cause Analysis Cycle Time Prediction with Machine Learning addresses a central…
root cause analysis fishbone defect prediction, root cause analysis, cause analysis fishbone, analysis fishbone defect, fishbone defect prediction, root cause, cause analysis
Root Cause Analysis Fishbone Defect Prediction with Machine Learning Introduction & Motivation Root Cause Analysis Fishbone Defect Prediction with Machine Learning addresses…
rca, production
Root cause analysis for equipment is the structured method for identifying the underlying technical and systemic causes of recurring or high-impact equipment problems - it focuses…
operations
Root cause analysis (RCA) is a systematic investigation technique used to identify the fundamental underlying cause(s) of a system failure, rather than just addressing…
root cause analysis, rca, production
Root Cause Analysis (RCA) is a systematic investigation methodology to identify the fundamental underlying cause of defects, failures, or process excursions, not just symptoms.…
quality & reliability
Root Cause Investigation is a structured analysis to identify the fundamental process or system cause behind an observed problem - It is a core method in modern semiconductor…
rope data structure, rope data, data structure
Rope Data Structure Introduction & Motivation Rope Data Structure addresses a central problem in data-structures process control and concatenation-operation-logarithmic…
rope, rotary position embedding
Rotary Position Embedding (RoPE) is the method most modern large language models use to tell the Transformer where each token sits in the sequence. Unlike the original absolute…
optimization
Rotary Position Embedding (RoPE) is the method most modern large language models use to tell the Transformer where each token sits in the sequence. Unlike the original absolute…
rotary encoder resolution drift prediction machine learning, rotary encoder, resolution drift prediction, rotary encoder resolution drift prediction, encoder resolution drift
Rotary Encoder Resolution Drift Prediction with Machine Learning Introduction & Motivation Rotary Encoder Resolution Drift Prediction with Machine Learning addresses a central…
rope positional encoding, rotary attention, position rotation matrix, rope llm
Rotary Position Embedding (RoPE) is the method most modern large language models use to tell the Transformer where each token sits in the sequence. Unlike the original absolute…
RoPE, angle embeddings, transformer positional encoding, relative position
Rotary Position Embedding (RoPE) is a positional encoding method that encodes token position as rotation angles in complex plane, applying multiplicative rotation to query/key…
rope positional encoding, relative position encoding, rope extrapolation, ntk aware scaling rope
Rotary Position Embedding (RoPE) is the positional encoding method used in most modern LLMs (Llama, PaLM, Qwen, Mistral) that encodes position information by rotating the query…
positional encoding transformers, rope attention mechanism, relative position encoding, position embedding interpolation
Rotary Position Embedding (RoPE) is the method most modern large language models use to tell the Transformer where each token sits in the sequence. Unlike the original absolute…
rope, positional encoding, alibi, positional attention
Rotary Position Embedding (RoPE) is the method most modern large language models use to tell the Transformer where each token sits in the sequence. Unlike the original absolute…
graph neural networks
RotatE is a knowledge graph embedding model that represents each relation as a rotation in complex vector space — mapping entity pairs through element-wise phase rotations…
graph neural networks
RotatE is a complex-space embedding model that represents relations as rotations of entity embeddings - It encodes relation patterns through phase rotations that preserve…
self-supervised learning
Rotation Prediction is an early self-supervised pretext task where the model is trained to predict which rotation (0°, 90°, 180°, 270°) was applied to an input image — requiring…