← Back to Chip Foundry Services

Glossary

28,973 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 1 of 580 (28,973 entries)

0-18-micron-process

0.18 micron process, 180 nm process, 0.18 micrometer process

A 0.18-micron process is the 180-nanometer technology generation. The conversion is exact: one micrometer is one thousand nanometers. The process label is not the thickness of…

0-25-micron-process

0.25 micron process, 250 nm process, quarter micron process

A 0.25-micron process is the quarter-micron, or 250-nanometer, generation. It is a useful reference point for deep-submicron CMOS because it connects finer transistor dimensions…

0-35-micron-process

0.35 micron process, 350 nm process, 0.35 micrometer process

A 0.35-micron process names the 350-nanometer technology generation. It sits between half-micron and quarter-micron generations in the conventional scaling sequence. The number is…

0-5-micron-process

0.5 micron process, 500 nm process, half micron process

A 0.5-micron process is the 500-nanometer generation of semiconductor fabrication. The name describes nominal technology dimensions, while the foundry design rules specify which…

0-8-micron-process

0.8 micron process, 800 nm process, 0.8 micrometer process

A 0.8-micron process belongs to the early submicron manufacturing generations. The nominal dimension is 800 nanometers. It identifies a process generation rather than the size…

1-bit sgd

distributed training

1-Bit SGD is a gradient quantization method that compresses each gradient component to a single bit (its sign) — transmitting only +1 or -1 for each gradient element, achieving…

1 f flicker noise prediction machine learning

1/f flicker noise prediction, 1/f flicker noise, flicker noise prediction, 1/f flicker, flicker noise, noise prediction

1/f Flicker Noise Prediction with Machine Learning Introduction & Motivation 1/f Flicker Noise Prediction with Machine Learning addresses a central problem in 1/f flicker noise…

1 f flicker noise prediction machine learning

1/f flicker noise prediction, 1/f flicker noise, flicker noise prediction, 1/f flicker, flicker noise, noise prediction

1/f Flicker Noise Prediction with Machine Learning Introduction & Motivation 1/f Flicker Noise Prediction with Machine Learning addresses a central problem in 1/f flicker noise…

1-kilobit DRAM

1Kb DRAM, DRAM density generation, dynamic RAM, memory scaling, one-transistor cell

The 1-kilobit DRAM is the first density generation of dynamic random-access memory, a chip storing 1,024 bits, and it marks the point where semiconductor memory became…

1-megabit DRAM

1Mb DRAM, megabit era, million-bit memory, three-dimensional cell capacitor, DRAM density generation

The 1-megabit DRAM is the density generation of dynamic memory storing roughly a million bits per chip, four times the 256-kilobit generation, reaching the market in…

1-micron process

one-micrometer node, feature-size milestone, micron boundary, scaling landmark, process generation

The 1-micron process is the technology generation at which the smallest features on a chip reached about one micrometer, a notable milestone in feature-size scaling marking…

常见汉字

常用汉字, 前100个汉字, common chinese characters, top 100 hanzi

常见汉字前 100 个 学汉字最好从最常用的字开始。下面是最常见的 100 个汉字: 序号 1–20(最常用,先重点学): 1 的 de — 所属标记 2 一 yī — 一 3 是 shì — 是 4 不 bù — 不 5 了 le/liǎo — 了,完成 6 在 zài — 在 7 人 rén — 人 8 有 yǒu — 有 9 我 wǒ — 我 10…

汉字101-200

常见汉字101, hanzi 101 200, common chinese 101

常见汉字 101–200 序号 101–150: 101 实 shí — 真实 102 日 rì — 太阳,天 103 军 jūn — 军队 104 者 zhě — ……的人 105 意 yì — 意思 106 无 wú — 没有 107 力 lì — 力量 108 它 tā — 它 109 与 yǔ — 和 110 长 cháng/zhǎng — 长…

3d-scene-reconstruction machine learning

3D Scene Reconstruction from Images Introduction & Motivation 3D Scene Reconstruction from Images addresses a central problem in vision processing and reconstruction-accuracy…

3d-convolution-video machine learning

3D Convolution for Video Introduction & Motivation 3D Convolution for Video addresses a central problem in vision processing and spatiotemporal-learning prediction: how to Build…

3d-pose-estimation machine learning

3D Pose Estimation from Video Introduction & Motivation 3D Pose Estimation from Video addresses a central problem in vision processing and 3d-accuracy-metric prediction: how…

3d-gaussian-splatting machine learning

3D Gaussian Splatting for Real-Time Rendering Introduction & Motivation 3D Gaussian Splatting for Real-Time Rendering addresses a central problem in vision processing…

3d-object-generation machine learning

3D Object Generation Introduction & Motivation 3D Object Generation addresses a central problem in generative generation and 3d-realism-quality synthesis: how to Design generative…

3d-vision-language machine learning

3D Vision-Language Models Introduction & Motivation 3D Vision-Language Models addresses a central problem in multimodal learning and 3d-semantic-alignment alignment: how to Learn…

10nm-process

10 nm process, what is 10 nm process

A 10 nm process name identifies a supplier’s technology generation rather than a universal ten-nanometer feature. Different manufacturers reached their named platforms on…

130nm-process

130 nm process, what is 130 nm process

A 130 nm process is a manufacturing generation also written as 0.13 micron. The label describes a technology platform with coordinated transistor, wiring, and design-rule choices.…

14nm-process

14 nm process, 14nm process, samsung-14nm-finfet, samsung 14nm finfet

The 14 nm label was used by several manufacturers for distinct FinFET technology platforms. Intel and Samsung both documented fin-based processes, but their names do not establish…

16-kilobit DRAM

16Kb DRAM, soft errors, alpha particle errors, DRAM reliability, memory scaling

The 16-kilobit DRAM is the density generation storing 16,384 bits per chip, four times the 4-kilobit generation, and it is the node where dynamic memory moved decisively into…

16-megabit-dram

16-megabit dram, what is 16-megabit dram

A 16-megabit DRAM stores 16,777,216 data bits in the conventional binary organization. That is 2,097,152 bytes, or 2 MiB, before considering how a system combines chips.…

16nm-process

16 nm process, 16nm process, tsmc-16nm-finfet, tsmc 16nm finfet

TSMC’s 16 nm process family introduced FinFET-based logic into its advanced foundry offerings. The raised fin increases the gate-controlled channel perimeter and improves…

193nm-lithography

193 nm lithography, 193 nanometer lithography

193 nm lithography specifies the exposure wavelength used by ArF systems. It does not specify a 193-nanometer transistor or require a process with that node name. The same…

1nm node pathfinding

1nm process technology, 1nm node requirements, sub 1nm scaling, 1nm manufacturing feasibility

The 1 nm technology node does not yet have a fixed definition the way earlier nodes did — it is being pathfound, not specified, because the transistor itself must change…

1T1C

1T1C cell, 1T1C DRAM cell, one transistor one capacitor, DRAM cell architecture, DRAM storage capacitor, access transistor, charge sharing sense amplifier, DRAM sensing, stacked capacitor DRAM, trench capacitor DRAM, DRAM retention time, DRAM refresh

1T1C: The One-Transistor-One-Capacitor DRAM Cell — Charge Storage Physics, Array Sensing, and the Capacitor Scaling Moat The 1T1C cell — one access transistor plus one storage…

2-5-volt-logic

2.5 volt logic, 2.5v logic

2.5-volt logic describes a nominal supply or signaling domain, not a universal interface specification. A device must be characterized for the relevant voltage range. Merely…

2.5d extraction

2.5d, signal & power integrity

2.5D Extraction is electromagnetic extraction balancing 2D efficiency with partial 3D effects for interconnect analysis - It offers faster runtime than full 3D while capturing key…

2 5d interposer tsv yield prediction with machine learning

2 5d interposer tsv yield prediction machine learning, 2 5d, interposer tsv yield, 5d tsv, 25d interposer tsv yield prediction, 25d interposer tsv, tsv yield prediction, tsv yield

2.5D Interposer TSV Yield Prediction with Machine Learning Introduction & Motivation 2.5D Interposer TSV Yield Prediction with Machine Learning addresses a central problem in 2.5D…

汉字201-300

常见汉字201, hanzi 201 300, common chinese 201

常见汉字 201–300 序号 201–250: 201 平 píng — 平安 202 形 xíng — 形状 203 反 fǎn — 相反 204 题 tí — 题目 205 认 rèn — 认识 206 条 tiáo — 条 207 结 jié — 结果 208 友 yǒu — 朋友 209 义 yì — 意义 210 活 huó — 生活 211…

20nm-process

20 nm process, what is 20 nm process

A 20 nm process is a logic-manufacturing generation whose implementation must be identified by supplier. In TSMC’s historical sequence, 20SoC is associated with planar logic…

22nm-process

22 nm process, 22nm process, intel-22nm-trigate, intel 22nm trigate

The 22 nm generation includes Intel’s transition to three-dimensional Tri-Gate transistors. Intel announced the manufacturing approach in May 2011 and associated it with…

248nm-lithography

248 nm lithography, 248 nanometer lithography

248 nm lithography names an exposure wavelength rather than a technology node. In semiconductor manufacturing the term ordinarily refers to KrF excimer-laser deep-ultraviolet…

256-kilobit DRAM

256Kb DRAM, DRAM density generation, three-dimensional capacitor, US-Japan memory competition, cell capacitance

The 256-kilobit DRAM is the density generation of dynamic memory storing 262,144 bits per chip, four times the 64-kilobit generation, reaching the market in the first half of…

256-megabit-dram

256-megabit dram, what is 256-megabit dram

A 256-megabit DRAM stores 268,435,456 bits, or 32 MiB. It is four times the bit capacity of a 64-megabit device. The number describes stored information rather than the amount…

25d

25d packaging, 2.5d, 2.5d packaging, advanced packaging 25d, silicon interposer packaging, cowos

Interposers are the substrates that carry 2.5D integration: a die-to-die wiring layer dense enough to route thousands of connections between chiplets that sit side by side.…

28nm-fd-soi

28 nm fd-soi, what is 28 nm fd-soi

A 28 nm FD-SOI platform uses a thin silicon device layer above a buried insulating oxide. The channel body is designed to be fully depleted in its intended operating regime.…

28nm-process

28 nm process, what is 28 nm process

A 28 nm process is a manufacturing-platform generation rather than a universal transistor dimension. Its devices, wiring pitches, gate stack, libraries, and qualification rules…

2D

materials, semiconductor, MoS2, WSe2, graphene

2D Materials in Semiconductors: MoS2, WSe2, and Graphene is atomically-thin layered materials exhibiting unique electronic properties enabling transistors, optoelectronic devices…

2d dopant profile

scm profiling, ssrm metrology

Two-Dimensional Dopant Profiling is a metrology technique that maps dopant concentration across both depth and lateral dimensions in semiconductor structures. What Is 2D Dopant…

2d material integration

graphene integration, tmdc devices

2D material integration brings channel materials only one to a few atoms thick into a technology stack built for three-dimensional silicon, and that mismatch in dimensionality is…

2d material semiconductor mos2

transition metal dichalcogenide, tmd monolayer transistor, mos2 channel transistor, 2d semiconductor device

2D Semiconductor Materials (MoS₂/TMDs) is the family of transition metal dichalcogenide crystals with monolayer thickness exhibiting direct bandgaps and strong light-matter…

2d material transistor

mos2 transistor, molybdenum disulfide, mos2, tmd fet, graphene transistor, transition metal dichalcogenide, tmd transistor

2D Material Transistors are the post-silicon device concept using atomically-thin layered semiconductors (MoS₂, WSe₂, black phosphorus) as channel materials — providing ultimate…

2d material transistor mobility drift prediction machine learning

2d material transistor mobility drift prediction, 2d material transistor, material transistor mobility, transistor mobility drift, mobility drift prediction, 2d material

2D Material Transistor Mobility Drift Prediction with Machine Learning Introduction & Motivation 2D Material Transistor Mobility Drift Prediction with Machine Learning addresses…

2d material transistors

mos2 transistor fabrication, molybdenum disulfide, mos2, tmd fet, tmdc channel devices, 2d material transfer, 2d heterostructure integration

Two-Dimensional (2D) Material Transistors: Physics, Materials & Scaling Two-Dimensional (2D) Material Transistors represent the ultimate channel-scaling frontier for semiconductor…

2d sinusoidal position encoding

2d, computer vision

2D sinusoidal position encoding is a fixed mathematical position representation that extends the original 1D sinusoidal encoding from "Attention Is All You Need" to two spatial…

2nm node challenges

2nm process technology, 2nm node requirements, 2nm manufacturing challenges, 2nm scaling issues

The 2 nm technology node is not a single process step but a cascade of simultaneous re-architectures — the transistor, the lithography, the contact, the power delivery, and…

2nm-process

2nm process, what is 2nm process

A generation associated with nanosheet logic. A 2nm process is an advanced logic platform defined by a semiconductor manufacturer. Leading examples use gate-all-around nanosheet…