Digital IC Design
Digital IC design,Verification,IPcore,Video,Audio,...
Jan 4, 2010
Dec 31, 2009
Dec 21, 2009
多媒体SOC严重同质化
cpu+video+graphic+image+audio+interface构成了一个高性能的soc,类似这样的芯片出了一颗又一颗,大部分核心ip都是license,这些产品的核心竞争力在哪呢?未来竞争力又体现在何处?做digital设计的人员出路又在何方?
Oct 13, 2009
restore source code encrypted by 'protectip'
protectip from synplicity(in the synplify install directory) can encrypts VHDL and Verilog source files, it's a simple perl script which call the openssl for encryption.
Sep 20, 2009
IC设计书籍推荐
做这个几年了,不知道算不算入门了。把一些曾经看过的书籍中觉得不错的推荐给大家。
May 10, 2009
Apr 30, 2009
Dirac-> vc-2 ?
About Dirac
The Dirac Project
Originally created by the BBC Research department, the Dirac project has expanded to include companies providing hardware equipment and software for handling Dirac video, as well as an active open-source development group.
Origin of the Name "Dirac"
The name "Dirac" is a reference to Paul A. M. Dirac, British physicist and winner of the 1933 Nobel Prize in Physics. The prize was shared with Erwin Schrödinger, for whom the Schroedinger implementation of Dirac was named.
read more:http://www.diracvideo.org/
Labels: Video
Mar 24, 2009
硅谷模式未必适合本土创业公司【转自老杳吧】
本人目前就职于一家本土集成电路创业公司,对这篇文章高度认同。
所谓硅谷集成电路创业模式一般包含三个阶段,一、获得VC对创意及团队认可,一般集成电路设计创业团队可以获得100-500万美元的启动资金创业;二、产品开发完成后创业公司会二次融资,额度一般在500-2000万美元用于产品推广;三、技术或产品获得市场认可后创业公司一般会选择直接IPO或被知名跨国公司收购,成功的硅谷创业公司一般都会经历上述三个阶段。
与硅谷模式不同,集成电路在中国发展几十年,上市的只有中星微、展讯和珠海炬力,能够被跨国公司收购的案例很少,十年前新涛科技被IDT以8500万美元收购以及上海掌微被Sirf 1.25亿美元收购之外鲜有成功,其实这两家也不能算是本土创业公司,总部都在硅谷,运营主体都设在美国,中国只设有部分研发,甚至所销售的产品也是以海外为主。
硅谷创业以核心技术为目标,能被跨国公司收购很正常,国内集成电路设计公司瞄准的是市场需求,至于技术是否领先则往往不是创业者最关注的,因此本土公司创业在技术层面很难引起跨国公司的关注,一旦开拓市场不利,被收购的价值为零,在中国集成电路设计界很少出现并购也是因为这个原因,因为中国集成电路设计公司有五百家之多,很多人预计未来会出现并购潮,老杳却认为很难,倒闭潮倒是非常有可能,真正拥有尖端核心技术的中国IC设计公司很少。
没有尖端核心技术并不能阻止本土IC设计公司的兴起,毕竟作为全球最大的电子产品制造基地,中国电子市场对不同种类、不同水准的集成电路需求非常旺盛,只有能够开发出适应市场需求的产品,即使没有最高端的技术本土IC设计公司一样可以做的非常不错,也是基于这一点,许多风险投资商认为中国是集成电路投资最后一片沃土。
与硅谷创业尽快开发出尖端技术不同,本土集成电路创业公司最大目标应当是尽快赚钱,无论产品高端低端,无论是否代表高科技,只有尽快实现营业收入,才能与客户建立紧密的关系,更好的把握市场脉搏,进而通过就客户推广新产品,目前阶段很多中国本土集成电路设计公司都是依赖中国成熟的电子制造产业生存,虽然产品品质与跨国公司相比略有差距,更好的性能价格比是竞争的根本,本土集成电路设计业的发展,反过来也促进了中国制造业的进步,制造出更加便宜、功能更加丰富的电子产品,应当说目前中国本土集成电路设计业已经与电子制造业形成了良好的互动和彼此依存的共生关系,山寨手机之所以流行,MTK固然居功至伟,中国本土集成电路设计业在成本上做的贡献同样不可小视,目前本土IC已经在FM、摄像头、蓝牙、电源管理、音频放大期,功率放大器、模拟电视甚至电池等部件上已经替代了或正在替代跨国公司成为主流供货商。其它诸如玩具用音频芯片等更在全球占据统治地位。
应当说本土IC更依赖市场而不是技术,这是本土IC创业公司与硅谷创业的最大区别,基于此本土创业者要更加重视市场和盈利,也正因为如此,本土集成电路设计公司被收购的价格会很低,这一点与电脑、电视领域的竞争格局并没有太大的区别。
虽然创业阶段本土与硅谷有较大的区别,如果希望成为一家伟大的公司,二者却殊途同归,说到底就像联想的"贸工技"和华为的"技工贸"的区别,拥有了市场的联想可以进一步拓展核心技术,这是本土IC发展的捷径,拥有了核心技术的华为进一步拓展市场,这是硅谷创业的归宿,当然如果创业公司无法最终生存,那又是另外一件事。(作者:老杳)
Dec 25, 2008
Dec 24, 2008
diff in linux
diff在linux下用来比较两个文件或者文件夹的内容,具体使用可以参考http://linux.about.com/library/cmd/blcmdl1_diff.htm,其中-b可以比较不同类型的文件,如dos,unix,但经实践证明,如果两个文件类型不一样,且文件比较大,如大于1MB时,比较时间会很长,why?
如果我们把Windows下产生的文件在linux下执行一次dos2unix,转换成unix类型,然后再用diff比较,时间将大大减少。仅试过文本文件。
Labels: Script
Nov 8, 2008
Virtex5 DSP48E Synplify8.6.2 bug?
最近工作中遇到的一个问题,不知道是否是工具的bug,在此备案。
问题描述:设计中所用到的一个乘法器,在fpga验证时不能工作,用逻辑分析仪查看得知乘法器输入正确,但输出有问题。通过多次修改并用综合后的网表仿真得知是synplify在综合时调用DSP48E时处理出错,具体分析如下:
环境:Xilinx的Virtex5 LX系列FPGA,Synplify8.6.2综合,ISE9.2实现,Modelsim后仿
设计背景:乘法器的输入,一端为寄存器直接输出,另一端是比较复杂的组合逻辑,输出结果再做加法等运算,然后送入寄存器。
DSP48E简介:
Virtex-5 DSP48E Slice 包含 Virtex-4 DSP48 的所有功能以及多种新功能。这些新功能包括
一个更宽的 25 x 18 乘法器和一个扩展后用作逻辑单元的加/ 减功能。
看框图:
![]() |
其中输入可以有两级寄存器,所以在做优化的时候会把外面的寄存器放到里面,而这两级寄存器分别有两个EN端,当只用其中一个时,会使用REG2,可以从仿真的model中间看到详细的描述,但Synpilfy在处理这种情况时,使用的是REG1 ,此时CE2常为1,所以导致输入数据不能在正确的时刻被latch,进而导致输出错误。
在搜索时发现Xilinx自己的工具,如coregen之类的在处理DSP48E时也出现过不少bug,synplify也是从8.6版本才开始支持virtex5的,所以我觉得很有可能是工具在这有bug,目前synplify的版本已达9.6以上,估计新版本不会再有此问题。
总结:
1、尽量不要用工具最新的功能,关注工具新版的release notes
2、FPGA做逻辑验证的时候最好不要使用FPGA的特殊单元,也不要让工具做太多的优化,除非timing确实有问题
3、出了问题应该先分析问题,不要一上来就尝试各种修改的办法。(这点我开始认为自己的coding style有问题,让工具犯傻了,所以修改了好几次,均已失败告终)
Labels: AsicDesign
Sep 30, 2008
IPcore Release Package
- Synthesizable Verilog RTL
- Bit-accurate C model
- Verilog testbench
- Detailed product documentation
- Design specifications
- Integration guidelines
- Complete verification suite
- VHS tape with example noisy signals
- Extensive tests including corner-case scenarios
- Golden test result vectors
- FPGA test board (with testbench and supporting files)
- Support and training
- to be added...
Labels: Docs
Sep 29, 2008
清除所有.svn目录
一、在linux下
删除这些目录是很简单的,命令如下
find . -type d -name ".svn"|xargs rm -rf
或者
find . -type d -iname ".svn" -exec rm -rf {} \;
全部搞定。
二、在windows下
1、在项目平级的目录,执行dos命令:
xcopy project_dir project_dir_1 /s /i
2、或者在项目根目录执行以下dos命令
for /r . %%a in (.) do @if exist "%%a\.svn" rd /s /q "%%a\.svn"
3、直接用windows的搜索功能,打开搜索隐藏文件的选项,然后搜索.svn,再一起删除即可
Labels: Script
Sep 27, 2008
Power Savings Techniques
From Spreadsheet to Sign-off by George Cuan, Cisco Systems, Inc.
Labels: AsicDesign
Aug 31, 2008
async fifo design
1. fifo body: dual ports sram or registers array
2. read and write address for sram
3. empty and full flag for the fifo body, it's the most important, and there are two methods presented in the reference paper.
4. almost empty or almost full flag, if needed
5. fifo content depth or fifo space depth, if needed
6. some special case for the special application, eg. the predicable r/w clock frequence ratio.
Reference paper:
http://www.sunburst-design.com/papers/
Labels: AsicDesign
Jun 5, 2008
H.264 Baseline Decoder from opencores
Nova is a low-power realtime H.264/AVC baseline decoder of QCIF resolution, targeting mobile applications. It is a dedicated, full hardwired and self-contained ASIC design without utilizing any GPP/DSP cores. It has been successfully verified on Xilinx Virtex-4 FPGA and 0.18um ASIC chip. The measured power consumption is 293uW@1V for 30fps QCIF decoding.
Introduction page:http://www.opencores.org/projects.cgi/web/nova/overview
Labels: AsicDesign, Video
May 28, 2008
Berkeley University EECS Course WEB Sites
EECSBA1 Strategic Computing and Communications Technology
EECS20N Structure and Interpretation of Systems and Signals
EE24 Freshman Seminar
EE40 Introduction to Microelectronic Circuits
EE42 Introduction to Digital Electronics
EE43 Introductory Electronics Laboratory
EE98 EE 98 Seminar Home Pages
EE100 Electronic Techniques for Engineering
EE104 Linear and Nonlinear Circuits
EE105 Microelectronic Devices and Circuits
EE117 Electromagnetic Fields and Waves
EE117B Electromagnetic Fields and Waves II
EE118 Introduction to Optical Communication Systems
EE119 Introduction to Optical Engineering
EECS120 Signals and Systems
EE121 Introduction to Digital Communication Systems
EE122 Introduction to Communication Networks
EE123 Digital Signal Processing
EE125 Introduction to Robotics
EE126 Probability and Random Processes
EE128 Feedback Control
EE129 Neural and Nonlinear Information Processing
EE130 Integrated-Circuit Devices
EE131 Semiconductor Electronics
EE136 Introduction to Quantum and Optical Electronics
EE140 Linear Integrated Circuits
EE141 Introduction to Digital Integrated Circuits
EE142 Integrated Circuits for Communications
EE143 Microfabrication Technology
EE145L Introductory Electronic Transducers Laboratory
EE145M Intro Microcomputer Interfacing Lab
EE145A (renamed to EE145L)
EE145B Image Processing and Reconstruction Tomography
EE146 unknown
EECS150 Components and Design Techniques for Digital System...
EECS152 Computer Architecture and Engineering
EE192 Mechatronics
EE194 EE 194 Seminar Home Pages
EEH196A Senior Honors Thesis Research
EE198 EE 198 Seminar Home Pages
EE199 Independent Study
EE201 Strategic Computing and Communications Technology
EE210 Applied Electromagnetic Theory
EE210B Applied Electromagnetic Theory
EE213 Soft X-Rays and Extreme Ultraviolet Radiation
EE217 Microwave Circuits
EE219 unknown
EE219A Computer-Aided Verification of Electronic Circuits
EE219B Logic Synthesis for Hardware Systems
EE219C Computer-Aided Verification
EE220 Neural & Nonlinear Information Processing
EE221A Linear System Theory
EE222 Nonlinear Systems--Analysis, Stability and Control
EE223 Stochastic Systems: Estimation and Control
EE224A Digital Communications
EE224B Fundamentals of Wireless Communications
EE225D Audio Signal Processing
EE225A Digital Signal Processing
EE225B Digital Image Processing
EE225C VLSI Signal Processing
EE226 unknown
EE226A Random Processes in Systems
EE227A Introduction to Convex Optimization
EE228A Communication Networks
EE229 Information Theory and Coding
EE229B Error Control Coding
EE230 Solid State Electronics
EE231 Solid State Devices
EE232 Lightwave Devices
EE233 Lightwave Systems
EE235 Nanoscale Fabrication
EE236A Quantum and Optical Electronics
EE238 Superconductive Devices and Circuits
EE240 Analog Integrated Circuit Design and Analysis
EE241 Advanced Digital Integrated Circuits
EE242 Advanced Integrated Circuits for Communications
EE243 Advanced IC Processing and Layout
EE244 Computer-Aided Design of Integrated Circuits
EECS245 Introduction to MEMS Design
EE246 Microelectromechanical Systems (MEMS)
EE247 Analysis and Design of VLSI Analog-Digital Interfac...
EE249 Embedded System Design
EE290D Advanced Topics in Semiconductor Technology
EE290E Advanced Topics in Electromagnetics and Plasmas
EE290F Advanced Topics
EE290G (renamed to EE245)
EE290H Semiconductor Manufacturing
EE290I Advanced Topics in Wireless Communication
EE290J Advanced Topics in Electrical Engineering
EE290N Advanced Topics in System Theory
EE290O Advanced Topics in Control
EE290Q Advanced Topics in Communication Networks
EE290S Advanced Topics in Communications and Information T...
EE290T Advanced Topics in Signal Processing
EE290X Advanced Topics in Management and Social Issues in ...
EE290Y Organic Materials in Electronics
EE290A Advanced Topics in Computer-Aided Design
EE290B Advanced Topics in Solid State Devices
EE290C Advanced Topics in Circuit Design
EE291 Control and Optimization of Distributed Parameters ...
EE291E Hybrid Systems and Intelligent Control
EE297 Field Studies in Electrical Engineering
EE298 EE 298 Seminar Home Pages
EE299 Individual Research
EE301 Teaching Techniques for Electrical Engineering
Labels: Docs
May 22, 2008
Support earthquake relief in China
Official figures (as of May 22, 10:00 CST) state that 51,151 are confirmed dead, including 50,651 in Sichuan province, and 288,431 injured.[4] Tens of thousands are missing, approximately 14,000 of them buried, and eight provinces were affected.[6] The earthquake left about 4.8 million people homeless.[7] It was the deadliest and strongest earthquake to hit China since the 1976 Tangshan earthquake, which killed over 240,000 people.
Labels: blog
Mar 23, 2008
Mar 10, 2008
registered output
Labels: AsicDesign
