当前位置:网站首页>Introduction to the types and repair methods of chip Eco
Introduction to the types and repair methods of chip Eco
2022-07-08 01:11:00 【NanDigits】
Classify according to the category of repair content , It can be divided into functions ECO and Timing ECO. function ECO It is to modify the logic function of the chip ,Timing ECO It's repair Setup/hold Time. function ECO It can be changed manually , You can also use tools , such as cadence conformal eco,nandigits Of gof eco.Timing ECO You can use tools directly on the back end (Innovus、ICC/ICC2) Li reform , It can also be used. PrimeTime etc. STA Tools to automatically Fix, You can also manually Fix A small part ( Sometimes Timing dissatisfaction , It could be Transition perhaps cap Too big ).
Of course, if Timing Difference is too big , From the chip architecture 、 The clock tree structure is further optimized . It may also be that the speed limit of the process has been reached , It's time to change the process .
According to whether the chip has been streamed , Can be divided into PreMask ECO and PostMask ECO.Mask Is the meaning of the mask , Before and after manufacturing the hood .PreMask ECO You can change all the layers made , and PostMask ECO Only the metal layer can be changed , Even a few specified metal layers .
In chip design , There is always some uncertainty , We need to consider reserving a variety of options . such as , Reserve in advance OTP/MTP Come on trim Circuit parameters and logic , Private registers to switch alternate functions and modes , utilize TOP Metal Option、EEPROM、Flash Waiting is also a common and easy-to-use method . as long as BUG It can be repaired or bypassed by software , Or reduce the working clock frequency , Or adjust Process corner, Anyway, the chip that does not affect customers' use is a good chip . It really has little influence BUG, Let the sales reduce the price .
边栏推荐
- Chapter VIII integrated learning
- 11. Recurrent neural network RNN
- Implementation of adjacency table of SQLite database storage directory structure 2-construction of directory tree
- Handwriting a simulated reentrantlock
- AI遮天传 ML-回归分析入门
- Chapter IV decision tree
- Codeforces Round #804 (Div. 2)
- 13. Enregistrement et chargement des modèles
- Complete model training routine
- Markdown learning (entry level)
猜你喜欢
Several frequently used OCR document scanning tools | no watermark | avoid IQ tax
10.CNN应用于手写数字识别
Complete model training routine
How does starfish OS enable the value of SFO in the fourth phase of SFO destruction?
Su embedded training - Day6
3. MNIST dataset classification
Su embedded training - Day9
For the first time in China, three Tsinghua Yaoban undergraduates won the stoc best student thesis award
Cancel the down arrow of the default style of select and set the default word of select
Chapter 7 Bayesian classifier
随机推荐
Redis, do you understand the list
SDNU_ ACM_ ICPC_ 2022_ Summer_ Practice(1~2)
Letcode43: string multiplication
新库上线 | CnOpenData中华老字号企业名录
Ag9310 for type-C docking station scheme circuit design method | ag9310 for type-C audio and video converter scheme circuit design reference
USB type-C mobile phone projection scheme | USB type-C docking station scheme | TV / projector type-C converter scheme | ag9300ag9310ag9320
1. Linear regression
Ag7120 and ag7220 explain the driving scheme of HDMI signal extension amplifier | ag7120 and ag7220 design HDMI signal extension amplifier circuit reference
New library online | cnopendata China Star Hotel data
STL -- common function replication of string class
3.MNIST数据集分类
Class head up rate detection based on face recognition
第四期SFO销毁,Starfish OS如何对SFO价值赋能?
Study notes of single chip microcomputer and embedded system
Introduction to paddle - using lenet to realize image classification method II in MNIST
Serial port receives a packet of data
网络模型的保存与读取
y59.第三章 Kubernetes从入门到精通 -- 持续集成与部署(三二)
Image data preprocessing
Vscode is added to the right-click function menu