← Back to Chip Foundry Services

Glossary

2,905 technical terms and definitions

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z All
Showing page 57 of 59 (2,905 entries)

Cycle Time

manufacturing operations

Cycle Time is the elapsed time required to complete one unit at a specific process step - It determines step capacity and queue behavior in production flow. What Is Cycle Time?…

Cycle Time Management

operations

Cycle time management is the control of total elapsed time from wafer release to completion by reducing wait, transport, and rework delays across the route - it is a primary…

Cycle Time Queue Prediction Machine Learning

fab cycle-time queue prediction, fab cycle-time queue, cycle-time queue prediction, fab cycle-time, cycle-time queue, queue prediction

Fab Cycle-Time Queue Prediction with Machine Learning Introduction & Motivation Fab Cycle-Time Queue Prediction with Machine Learning addresses a central problem in semiconductor…

Cycle Time Reduction

production

Cycle time reduction is the systematic reduction of total elapsed time from process start to finished output - it targets queue delay, handoff friction, and imbalance so products…

Cycle Time Variability Forecast Prediction Machine Learning

cycle time variability forecast prediction, cycle time variability, time variability forecast, variability forecast prediction, cycle time, time variability, variability forecast

Cycle Time Variability Forecast Prediction with Machine Learning Introduction & Motivation Cycle Time Variability Forecast Prediction with Machine Learning addresses a central…

Cyclegan

generative models

CycleGAN is the pioneering generative adversarial network architecture that enables unpaired image-to-image translation using cycle consistency loss — learning to translate images…

Cyclegan-Unpaired Machine Learning

CycleGAN for Unpaired Translation Introduction & Motivation CycleGAN for Unpaired Translation addresses a central problem in generative generation and cycle-consistency-quality…

Cyclegan Voice

audio & speech

CycleGAN Voice is unpaired voice-conversion using cycle-consistent adversarial learning between speaker domains. - It converts source speech style to target style without…

Cyclic-Learning-Rates Machine Learning

Cyclic Learning Rate Schedules Introduction & Motivation Cyclic Learning Rate Schedules addresses a central problem in optimization algorithm tuning and improvement: how…

Cyclic Mechanical Stress Fatigue Acceleration Machine Learning

cyclic mechanical stress fatigue acceleration prediction, cyclic mechanical stress, mechanical stress fatigue, stress fatigue acceleration, fatigue acceleration prediction

Cyclic Mechanical Stress Fatigue Acceleration Prediction with Machine Learning Introduction & Motivation Cyclic Mechanical Stress Fatigue Acceleration Prediction with Machine…

Cyclic Stress Test

reliability

Cyclic stress test is testing that alternates stress levels in repeated cycles to activate fatigue-related failure mechanisms - Periodic thermal or electrical cycling introduces…

Cyclomatic Complexity

code ai

Cyclomatic Complexity is a software metric developed by Thomas McCabe in 1976 that counts the number of linearly independent execution paths through a function or method…

Czochralski

crystal growth, silicon ingot, pure silicon

Czochralski process is the primary method for growing single-crystal silicon ingots from molten ultra-pure silicon — producing the 99.999999999% (11-nines) pure silicon wafers…

Czochralski

crystal growth, silicon ingot, pure silicon

Czochralski process is the primary method for growing single-crystal silicon ingots from molten ultra-pure silicon — producing the 99.999999999% (11-nines) pure silicon wafers…

Czochralski Crystal Growth Defect Prediction Machine Learning

czochralski crystal growth defect prediction, czochralski crystal growth, crystal growth defect, growth defect prediction, czochralski crystal, crystal growth, growth defect

Czochralski Crystal Growth Defect Prediction with Machine Learning Introduction & Motivation Czochralski Crystal Growth Defect Prediction with Machine Learning addresses a central…

Czochralski Crystal Pull Defect Prediction with Machine Learning

czochralski crystal pull defect prediction machine learning, czochralski crystal, pull defect prediction, czochralski crystal pull defect prediction, czochralski crystal pull

Czochralski Crystal Pull Defect Prediction with Machine Learning Introduction & Motivation Czochralski Crystal Pull Defect Prediction with Machine Learning addresses a central…

Czochralski Crystal Pulling Diameter Control Prediction Machine Learning

czochralski crystal pulling diameter control prediction, czochralski crystal pulling, crystal pulling diameter, pulling diameter control, diameter control prediction

Czochralski Crystal Pulling Diameter Control Prediction with Machine Learning Introduction & Motivation Czochralski Crystal Pulling Diameter Control Prediction with Machine…

Czochralski Method

Czochralski process, CZ silicon, Jan Czochralski, Teal and Little, single-crystal growth, silicon boule, oxygen in silicon

The Czochralski method is a way of growing a large single crystal by dipping a small seed crystal into a melt and slowly drawing it upward while rotating it, so that the liquid…

Capacitor

on chip capacitor, mim capacitor, mom capacitor, mos capacitor, decoupling capacitor

capacitor is a two-terminal element that stores charge and electric-field energy according to Q = C × V. Capacitors stabilize power rails, define analog time constants, sample…

Chip Design

ic design, fabless design, rtl to gdsii, eda tools, design rule check, drc, lvs, static timing analysis, tapeout, process design kit, pdk, place and route, ppa tradeoff

Design is the half of the industry that never touches a wafer: the discipline of turning an intended function into a file a foundry can manufacture, under constraints — power…

Capacitor Contact Etch

dram capacitor contact etch, plate contact etch, PC etch dram, dram pc via, capacitor plate bias via

The capacitor contact etch is the final via cut in the DRAM capacitor module, the step that drills down from the periphery metal level to reach the shared top electrode plate so…

Cell Sti Etch

shallow trench isolation DRAM, active area isolation, etch loading cell array, pad nitride strip, STI oxide recess

The cell STI etch cuts the shallow trench isolation that electrically separates one DRAM cell's active area from its neighbor, and inside the array this trench has to be narrower…

Confront the 1k Density Wall

4k dram, dram4k 1973

Confront the 1K DRAM Density Wall The 1968 Dennard one-transistor, one-capacitor cell solved the structural problem of dense dynamic memory, and the parts that followed it through…

Compare Pin Counts Before and After Multiplexing

4k dram, dram4k 1973

Compare Pin Counts Before and After Multiplexing Without multiplexed addressing, a 4K-bit DRAM needing roughly twelve address bits would have required twelve dedicated address…

Compare Tms4030 Design Choices to Mk4096

dram cost per bit, dram4k 1973

Compare TMS4030 Design Choices to the MK4096 The TMS4030 reached 4K density using TI's own cell and sense-amplifier design choices, which differed from Mostek's in process detail…

Carry Lessons Forward to 16k and Beyond

4k dram, dram4k 1973

Carry Lessons Forward to 16K and Beyond The engineering lesson of the 4K generation, that address-pin count had to be managed with multiplexing rather than simply adding more…

Close Naming the 64k Generation as the Next Leap

mostek mk4096, dram4k 1973

Close by Naming the 64K Generation as the Next Leap The 4K generation, anchored by Mostek's MK4096 and contested commercially by Texas Instruments' TMS4030, closed the gap between…

Confront the Problem Dov Frohman Was Actually Solving at Intel

eprom eeprom history step 1 confront the problem dov frohman was actually solving at int, eprom eeprom history overview

Confront the Problem Dov Frohman Was Actually Solving at Intel 1. A Reliability Problem, Not a Memory Project In 1971, Dov Frohman was an engineer at Intel working on a question…

Close the Eprom Eeprom History and Hand Off by Name to NAND Flash

eprom eeprom history step 13 close the eprom eeprom history and hand off by name to nand, eprom eeprom history series close

Close the EPROM/EEPROM History and Hand Off, by Name, to NAND Flash 13. Closing This Series' Arc This series told a specific, previously-uncovered chapter of non-volatile memory…

Chemical Etching

selective etching, selective chemical etch, GAA channel release, sacrificial layer removal

Chemical etching. uses reaction chemistry chosen to remove a target material faster than adjacent masks, stop layers, channels, spacers, liners, or substrates. Selectivity is…

Confront Why the First Working Mos Devices Were P Channel Not N Channel

experimental mos bipolar logic step 5 positive oxide charge forced early pmos polarity choice, early pmos channel polarity choice

Confront Early PMOS Polarity: Why the First Working Devices Were P-Channel, Not N-Channel 1. Why Every Early Experimental MOS Device Used a P-Channel, Against the Intuitive Choice…

Confront Resistor Transistor Logic Fan Out and Noise Margin Limits

experimental mos bipolar logic step 9 measured rtl noise margin falling with fan out count, rtl fan out noise margin limit

Confront RTLs Fan-Out and Noise-Margin Limits 1. Why RTL's Noise Margin Was the Family's Actual Practical Ceiling Measured across real RTL gates driving real fan-out loads…

Compare Experimental Mos Static Power Against Bipolar Standby Power

experimental mos bipolar logic step 16 mos off state leakage versus bipolar standing current, mos static power versus bipolar standby power

Compare Experimental MOS Static Power Against Bipolar Standby Power 1. Why Static Power, Not Switching Power, Was MOS's Clearest Measured Advantage Measuring a PMOS gate and…

Confront the Experimental Mos Speed Deficit Against Bipolar Switching Speed

experimental mos bipolar logic step 17 low mobility thick oxide and channel resistance slow early pmos, early pmos speed deficit versus bipolar

Confront the Experimental MOS Speed Deficit Against Bipolar Switching Speed 1. Why Early PMOS Could Not Yet Match Bipolar Switching Speed at Any Reasonable Voltage Measured…

Characterize Yield Limits That Kept Experimental Mos Out of Production Volume

experimental mos bipolar logic step 18 threshold scatter contamination and oxide pinhole defects limit yield, experimental mos yield production limit

Characterize Yield: Why Experimental MOS Struggled to Reach Production Volume 1. Why Laboratory Success Did Not Translate Directly Into a Shippable Product Line Across RCA's…

Curtaining Artifact

FIB cross section, focused ion beam milling, failure analysis imaging artifact, xenon plasma FIB, ion beam sputter rate

Curtaining is a milling artifact that shows up in a focused ion beam cross-section as a series of vertical streaks running straight through the image, caused by the ion beam…

Closing the Arc and Handing Off by Name to CMOS Logics Own Commercial Era

first commercial mos linear ics step 20 overview, closing the arc and handing off by name to cmos logics own c

Closing the Arc and Handing Off, by Name, to CMOS Logic's Own Commercial Era 1. What This Series Covered This series told two genuinely separate but contemporaneous commercial…

Cloud Computing

cloud ai, aws, azure, gcp, gpu cloud, managed ml

cloud computing is the on-demand delivery of compute, storage, networking, databases, platforms, and applications through pooled datacenter infrastructure. Cloud platforms are…

Compiler Optimization

gpu kernel fusion optimization, operator fusion deep learning, kernel launch overhead, fused kernel computation, fusion compiler optimization

compiler optimization is the transformation of a correct program into semantically equivalent code that runs faster, uses less energy, occupies less memory, or maps efficiently…