2007년 12월 31일 월요일

2007년 마지막 날입니다..

안녕하세요... PLDWorld 집쥔입니다...

어영부영 뭐 특별히 한것도 없는데, 벌써 2007년도 마지막 날이 되었습니다... 년초마다 거창한 계획을 세우기는 하지만, 뭐 하나 제대로 진득하게 하지 못하는 저로서는 2007년 한해도 마냥 아쉬운 한해이기만 합니다...

하지만 그렇다고 과거의 후회에 묻혀 살수는 없는법...!!! 다시한번 힘내서 앞으로 나아갈 수 있도록 노력하려 합니다...

여러분 모두 무자년 한해, 무쟈게 복많이 받으셔서 부자되는 한해 되시길...

감사합니다...

2007년 12월 27일 목요일

Marvell Power Supply Chip Shrinks Power Bricks, Energy Use

Marvell Power Supply Chip Shrinks Power Bricks, Energy Use

marvellpower.jpg

Marvell's launching a new chip for more efficient power supplies that cuts down energy use by automatically adjusting the amount of juice drawn by a computer depending on what it actually needs, slashing waste—it can chop peak energy use by up to 50 percent, according to Marvell.

It also supposedly rolls the functionality of 20 components into the single processor, allowing for the "Honey I Shrunk the Power Brick" magic (by up to a third) at lower costs for OEM manufacturers. When we talked with Hubie Notohamiprodjo, general manager of industrial control and power management, he said we could probably expect products with the new chips to start showing up in a few months.

While we'll see them in desktop and notebook power supplies first—Marvell's hitting them because of the size of computers' carbon footprint and its potential reduction—they could also make their way into flatscreen TVs, portables and other consumer electronics. It was tipped that they're talking to at least one manufacturer now. Even if green tech's not really your bag—or hell, even if it didn't actually save that much power —carrying around a smaller notebook power brick is reason enough for us to give it a thumbs up.

Digital PFC Controllers Smart Technology for Power Supplies

Marvell is the first to market with a DSP-based power factor correction (PFC) controller for AC-DC power supplies that are designed to significantly cut PC energy usage, while also helping to reduce the system's carbon footprint and save on energy bills. By integrating an average of 20 discrete components to build notebook adapters and desktop PC power supplies, Marvell allows OEMs to increase reliability while reducing size and cost.

According to the U.S. Environmental Protection Agency (EPA), there are more than 10 billion AC-DC power supplies used in computing, telecommunications, and consumer electronics worldwide, an estimated 2.5 billion in the United State alone. More efficient power supply designs could cut U.S. energy usage in half, saving nearly $3 billion and about 24 million tons of carbon dioxide emissions per year.

Marvell's smart technology for power supplies, the DSP determines the amount of power required for users' applications with mixed-mode technology that optimizes operation throughout the cycle by changing from Pulse-Width Modulation (PWM) to Pulse-Frequency Mode (PFM). This keeps the peak current at the lowest level improving energy efficiency, as well as reducing overall system size. The adaptive drive capability of the DSP intelligently adjusts the driving level to the switch; reducing switching loses, while adaptive adjustments also reduce electromagnetic interference (EMI) for a wide range of switches. Adaptive current limits are also automatically adjusted for varying current threshold limits worldwide, which adds system protection and reliability.

Marvell® 88EM8011 for Desktop Power Supplies

Desktop power supplies based on Marvell's 88EM8011 technology will deliver more energy efficiency as voltage and current are in-phase, wasting less energy than present technology. The Marvell 88EM8011 reduces Total Harmonic Distortion (THD) to around one percent or less (current technology THD averages 160 percent.)

It is also designed to be EPA Energy Star compliant, which requires that 80 percent of the power consumed is actually used by the PC, as well as 80Plus program compliant. 80 PLUS is an electric utility-funded incentive program to integrate more energy-efficient power supplies into desktop computers and servers.

Marvell 88EM8041 for Notebook Adapters

At half-a-pound in size, Marvell-based notebook adapters are nearly half the size of current adaptor technology. Using in-phase voltage and current, these products are more energy efficient with THD at one percent or less. With the Marvell 88EM8041, OEMs can reduce the size and weight by a notebook adaptor up to one-third. With integrated DSP-technology that incorporates an average 20 discrete components, design complexity and manufacturing costs are reduced, while reliability is increased.

10:59 PM on Mon Nov 5 2007
by Matt Buchanan


 

"resilient packet ring" - Google 검색

http://www.google.co.kr/search?num=100&complete=1&hl=ko&newwindow=1&rlz=1T4GGIH_koKR233KR236&q=%22resilient+packet+ring%22&btnG=%EA%B2%80%EC%83%89&lr=

http://www.google.co.kr/search?num=100&complete=1&hl=ko&newwindow=1&rlz=1T4GGIH_koKR233KR236&q=%22resilient+packet+ring%22&btnG=%EA%B2%80%EC%83%89&lr=lang_ko&aq=-1&oq=

謹賀新年

전자 메일 보내기: KTword World !!! (정보통신기술용어해설)

 

【   정보통신기술용어해설   】 귀하께서는 KTword (정보통신기술용어해설) 홈페이지의 용어해설 및 코딩, 테스트,
기술자료수집정리, 가이드북 집필 등을 위한   
개인연구실  로 접속하셨습니다 !!!

    정보통신용어검색    
알파벳 검색    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

우리말 검색   
     

Welcome to KTword world !!! Copyrightⓒ   since 2002   by Cha Jae Bok (cjb@kt.co.kr)


2007년 12월 21일 금요일

임베디드 - fpga4fun 업데이트

출처: http://www.ezdoum.com/stories.php?story=07/12/17/6547390


글쓴이:
EzDoum
글쓴날: 2007년 12월 17일 오전 10:55

fpga4fun을 오랫만에 들어가 봤더니 업데이트 됐군.

fpag와 usb인터페이스(fx2), arm processor를 넣은 개발 보드를 출시 했음.

FPGA + ARM development boards Getting started guide
http://www.knjn.com/files/Getting%20started%20with%20the%20Flashy%20boards.pdf
http://www.knjn.com/board_Xylo.html

FPGA + ARM development boards Getting started guide에서 참고로 할 자료가 많음
http://blogs.msdn.com/coding4fun/

2007년 12월 20일 목요일

Open-source EDA toolkit targets Unix

출처: http://www.scdsource.com/article.php?id=62


Open-source EDA toolkit targets Unix

By Richard Goering, 12/18/07

Using open-source EDA software today may require a difficult porting job to one's platform of choice. To ease that burden, startup ReyesSoft has launched the OpenEDA Toolkit 1.0, which ports popular open-source EDA software programs to several Unix-based platforms.

The toolkit is the first public offering from ReyesSoft (Fairfax, Calif.), a consulting firm run by Ed Immenschuh, founder and president. Immenschuh primarily has a software background, although he's worked with hardware engineers and written Verilog code. Through these experiences, he became aware of the high costs of commercial EDA software.

"I used to work for Silicon Graphics, and one of my friends tried to start up a network hardware company," Immenschuh said. "One problem was the cost of the tools. They were really, really expensive, anywhere from $20,000 to $40,000 per user."

While there are a number of open-source EDA programs, they're written mainly for Linux, Immenschuh said. If someone wants to run such a program on a Unix platform, they may have to download the source and do a "build" themselves. Not all Unix machines offer the full capability for doing so. "Some ports can be simple, while others can be hours long," he said. "There are always bugs, or something wrong with the compiler, or something else that prevents things from building."

The OpenEDA Toolkit ports open-source EDA software to Solaris Sparc and x86, Apple OS X 10.4, HP-UX 11i, Cygwin for Microsoft Windows, and Linux. The following open-source software packages are included:

  • Alliance CAD, which includes a VHDL compiler, simulator, logic synthesis, and automatic IC placement and routing.
  • Icarus Verilog, a command-line Verilog compiler for logic simulation and synthesis.
  • gEDA, an ongoing project that currently includes schematic capture, attribute management, bill of materials (BOM) generation, netlisting, analog and digital simulation, and printed circuit board (PCB) layout.
  • FreeHDL, a command-line VHDL-1993 compliant compiler.
  • Electric, which includes custom IC layout, a schematic drawing package, and a hardware description language specification.
  • GTKWave, a waveform output viewer.
  • PCB, an interactive PCB editor for the X11 window system.
  • Gerbv, a viewer for Gerber files generated from PCB CAD systems.
  • Gnucap, the Gnu Circuit Analysis Package, a general-purpose circuit simulator.
  • NgSpice, a mixed-level/mixed-signal circuit simulator.

Immenschuh said the toolkit is aimed at small companies that can't afford commercial EDA tools, as well as university students, some of whom may be using legacy Unix platforms. Following release of this toolkit, Immenschuh said, ReyesSoft hopes to go on to other open-source projects in areas such as 3D graphics and media streaming.

The OpenEDA Toolkit is available for $100 for the basic version. Purchasers can also buy one-year subscriptions to receive ongoing updates, and can buy a support package. For every purchase, ReyesSoft will donate 10 percent of the profits to open-source developers and projects, or to environmental organizations working to slow global warming.

Related articles
Do-it-yourselfer's EDA project wins open-source fans
Methodology drives verification framework development

2007년 12월 18일 화요일

전자 메일 보내기: [인기-추천상품] 한국쓰리젠, 인텔리나스 DT30

 

ETnews

[인기-추천상품] 한국쓰리젠, 인텔리나스 DT30
[ 2007-12-17 ]  
 한국쓰리젠(대표 이규용 www.3gendata.co.kr)은 자사의 주력 제품인 NAS(Network Area Storage) 기술력을 업그레이드 한 '인텔리나스 DT30'으로 추천상품에 올랐다.

 이 제품은 하드웨어(HW) RAID를 지원, 소프트웨어(SW) 기반 레이드를 지원하는 대부분의 데스크톱형 스토리지와 달리 장애시 데이터의 손실이 없는 것이 특징이다.

 또한 동급 제품으로서는 국내 최초로 2개의 기가비트 포트를 탑재해 빠른 네트워크 속도를 보장한다.

 한국쓰리젠 측은 인텔리나스 DT30이 웹하드 애플리케이션을 무상으로 제공한다고 밝혔다. 또 학교나 연구소, 기업들처럼 대용량의 파일이나 사용자가 많은 그룹이 협업할 때 적정한 솔루션으로 인터넷의 웹하드를 쓰는 비용보다 휠씬 저렴한 가격으로 전용 웹하드를 사용하는 장점을 누릴 수 있다.

 또한 이 제품은 파일서버, 서버통합, 백업 및 복구 전략 및 복제 솔루션에 모두 적합한 올인원 솔루션이다.

 부가적으로는 시점 파일 복구(Point in Time File Recovery)를 위한 스탭샷을 제공한다. 도메인 통합, 마이크로소프트 ADS, 유닉스 NIS를 지원하고, 유저, 그룹 관리를 통해 안정적인 접근 기능도 제공한다.

 인텔리나스 DT30은 동급 제품에서는 제공하지 않는 무중단 부품 교환이 가능한 것도 특징이다. '플러그&플레이' 형식의 네트워크 접속 및 설치 툴로 설치가 쉽고, 웹 기반의 관리툴을 제공함으로써 관리도 매우 쉽다. 또한 개인·팀별로 용량을 제한할 수 있고, 접속 제한도 가능하다.

 한국쓰리젠은 이 제품을 앞세워 내년부터 전통적인 주력 제품인 NAS 시장에서 고성장을 이룬다는 전략이다. 한편 이 업체의 NAS 제품은 국내에 이미 200여 군데 설치되어 성공적으로 운영되고 있다.
출력하기 닫기
Copyrightⓒ 2000-2005 ELECTRONIC TIMES INTERNET CO., LTD. All Rights Reserved.

2007년 12월 14일 금요일

[2007 IT히트상품] 디지털 시대, 디지털 리더 디지털타임스

전자 메일 보내기: [2007 IT히트상품] CPU 등 구성요소 신기술 채택, IA서버 한국HP '프로라이언트 DL380 G5' - 디지털 시대, 디지털 리더 디지털타임스


 
  
[2007 IT히트상품] CPU 등 구성요소 신기술 채택

IA서버 한국HP '프로라이언트 DL380 G5'
게재일자 : 2007/12/14

한국HP(대표 최준근)의 프로라이언트 DL380 G5 서버는 2006년 7월에 출시된 제품으로 DL380 서버의 후속기종으로, 기존 G4 서버와 비교해 CPU, 메모리, 네트워크, 디스크, I/O, 관리 등 대부분의 서버 구성 요소에 대한 신기술을 채택한 제품이다.

DL380 G5 서버는 2U로 구성돼 있으며 엔터프라이즈 시스템 급의 시스템 가용성과 관리성, 산업표준 제품이라는 점을 내세워 활발한 영업이 진행되고 있다.

DL380 G5 서버는 인텔의 가상화 기술을 탑재한 저전력 제온 5100 계열의 듀얼코어 프로세서 또는 제온 5300 계열의 쿼드코어 2개를 탑재할 수 있다. 여기에 인텔 벤슬리 플랫폼이 탑재되면 FSB가 최대 1333㎒로 증가돼 메모리 액세스 성능도 크게 개선됐다는 것이 HP의 설명이다.

DL380 G5 서버는 디스크, 메모리, 전원 공급기 및 쿨링팬 등 주요 서버 구성 요소에 대해 이중화를 제공하고 온라인 교체를 지원하여 가용성 및 편리성이 크게 증대됐다.

2m크기의 42U 랙에 최대 21개의 서버가 장착되며 탈부착 편의를 위해 퀵 릴리즈 레버가 달려 있어 관리자를 고려했다는 평가다. 또한 서버 및 CPU 제품의 당일 서비스 무상 업그레이드 지원을 통한 서비스를 제공한다.

한국HP는 x86서버 관리 소프트웨어 판매 극대화 등을 정하고, 고객사들에게 보다 안정적이고 산업 표준에 기반한 제품 공급을 위한 다각적인 마케팅 및 영업 활동 전개, NT서버 통합 및 가상화 시장수요를 발굴하고 판매량을 극대화한다는 전략이다.
[저작권자(c) 디지털타임스 무단 전재-재배포 금지]

2007년 12월 11일 화요일

Emailing: 매직아이, 533㎒급 고성능 SoC ‘폴룩스’ 양산


ETnews

매직아이, 533㎒급 고성능 SoC '폴룩스' 양산
[ 2007-12-07 ]  
  매직아이(대표 손해윤 www.mesdigital.com)가 533㎒ CPU와 3D 그래픽을 지원하는 칩 '폴룩스(Pollux·사진)' 양산에 돌입했다.

90나노미터(㎚) 공정으로 개발된 폴룩스는 고성능·고해상도 화면이 요구되는 카내비게이션·고해상도 전자액자·듀얼 판매시점관리(POS) 시스템 등에 사용하기 적합하다.

내장된 3D 그래픽 엔진은 생동감 있는 사용자 인터페이스와 듀얼 디스플레이 및 고해상도(1280x1024) 디스플레이를 지원한다. 폴룩스는 소형 메모리카드인 SD카드와 MMC카드, NTSC/PAL 인코더, USB2.0, 다층셀(MLC) 구조의 낸드 플래시 메모리, 박막트랜지스터(TFT), STN LCD 등을 모두 지원한다.

폴룩스를 전자제품에 적용하려는 기업들은 이달 말부터 매직아이 대리점과 아이지시스템에서 공식 개발툴을 예약 구매할 수 있다. 매직아이는 50만개를 공급하기로 계약, 폴룩스를 채택한 첫 번째 전자제품이 내년 1분기에 나올 예정이라고 설명했다.

손해윤 매직아이 사장은 "폴룩스를 고속 CPU와 3D그래픽을 하나의 칩으로 만든 제품으로는 획기적인 가격에 공급할 예정이다"라며 "이 칩을 장착할 내비게이션과 전자액자 등 소형 디지털 제품도 가격이 낮아져 저변이 확대되는 계기가 될 것이다"라고 전망했다.

정소영기자@전자신문, syjung@
출력하기 닫기
Copyrightⓒ 2000-2005 ELECTRONIC TIMES INTERNET CO., LTD. All Rights Reserved.

 
 

2007년 11월 28일 수요일

IC 설계에 지름길 제공해주는 HDL 코더

게재: 2006년 11월 01일
리차드 괴링

지금까지 IC 설계에서 당당한 행보를 해온 MathWorks사가 Simulink HDL 코더를 내놓았다. 이것은 자동으로 Simulink 모델과 Stateflow 다이어그램으로부터 합성 가능한 Verilog과 VHDL을 생성한다. 이러한 행보는 수천 Matlab 및 Simulink 유저들에게 FPGA 또는 ASIC 구현으로 가는 직접적인 경로를 제공한다.

MathWorks는 큰 사용자 기반을 가지고 이미 부상하고 있는 ESL 설계 시장의 경쟁자가 되었다. 회사에 따르면 새로운 링크로 유저들은 Matlab과 Simulink로 시스템 모델들과 알고리즘을 설계하고, 시뮬레이팅하며, 타당성을 입증할 수 있다고 한다. 또한, 자동으로 하드웨어와 소프트웨어 모두를 생성하며, 오리지날 시스템과 알고리즘 모델에 대해 HW/SW 구현물을 검증한다.

HDL Coder는 테스트벤치와 합성 및 시뮬레이션 스크립트들과 함께bit-true, cycle-accurate IEEE 1364-2001 Verilog과 IEEE 1076 VHDL을 생성해낸다. 옵션으로 ModelSim으로의 링크도 가지고 있어 Mentor Graphics 시뮬레이터 유저들은 HDL 코드를 다시 공동 시뮬레이션을 위한 Simulink로 되돌려 놓을 수 있다. 옵션인 Stateflow 제품으로는 유저들이 유한상태머신(FSM)으로부터 제어 로직을 생성시킬 수 있으며, 거기에 더하여 Simulink 모델들로부터 데이터 경로 로직을 생성시킬 수 있다.

신호 처리와 통신 부문 마케팅 디렉터인 Ken Karnofsky 씨에 의하면 고객들은 MathWorks 제품으로 설계를 시작하고 그리고 나서 HW/SW를 구현한다고 한다. "디바이스에 독립적으로 Simulink 모델들을 디자인의 HDL 레벨로 옮겨달라는 고객들의 요구가 꽤 있다. 지금까지의 기능들은 디바이스 벤더의 기술에 특정되는 경향이 있었다."

Karnofsky 씨는 그러나 Xilinx사가 지난1월 Matlab-to-HDL 공급업체인 Accelchip을 인수한 사실이 HDL 코더를 내놓는데 작용한 것은 아니라고 말했다. "우리는 수년 전부터 이러한 접근 방식과 전략을 준비해왔다"고 그는 설명했다. 그리고 MathWorks는 이미 Simulink 모델들로부터 디지털 필터들을 위해 RTL 코드를 생성시키는 성공적인 제품을 가지고 있었다고 지적했다. 고객들은 그러한 능력이 더 일반적인 방식으로 확장되기를 원한다고 그는 덧붙였다.

Karnofsky 씨에 의하면 HDL Coder 유저들에는 하드웨어 스펙에 간섭하지 않을 더 나은 방법을 필요로 하는 시스템 엔지니어들과 알고리즘 엔지니어들이 포함될 것이라고 한다. HDL Coder는 또한 이미 Simulink로 된 IP를 가지고 있는 이들과 FPGA 구현을 추구하는 DSP 엔지니어들에게도 호소력이 있을 것이라고 한다.

Karnofsky 씨는 Simulink에서 HDL로의 링크에 대해서는 ASIC 개발자들보다 FPGA 디자이너들이 더 많이 관심을 갖고 받아들이고 있다고 말했다. 그러나 그는 ASIC 커뮤니티의 시스템 아키텍트들이 Simulink를 이용하고 있으며, HDL Coder의 베타 테스터들은 FPGA 디자이너와 ASIC 디자이너가 거의 반반씩 섞여 있다는 사실에 주목했다.

HDL을 생성하던 이전의 Simulink add-on들은 고유의 블록셋과 IP 라이브러리들을 필요로 했다고 Karnofsky 씨는 말했다. 이와는 대조적으로 HDL Coder로는 유저들이 표준 Simulink 블록들을 가지고 설계한다. 이것은 하드웨어와 소프트웨어에 대해 개별적인 모델들을 갖게 되는 일을 피하고, 골든 레퍼런스 디자인을 대표하는 단일 모델을 가질 수 있도록 해준다고 그는 강조했다.

HDL 코드를 생성하기 위하여 디자이너들은 자신들이 원하는 구현을 명시하는데 GUI 또는 스크립팅 메커니즘을 이용한다. 한 콘트롤 파일이 파라미터들을 설정하고 직렬로 구현할 것인지, 아니면 종속 연결이나 병렬로 구현할 것인지를 지정할 수 있게 해준다. 유저들은 극성과 유형, 리셋 신호의 포트 이름과 더불어 Verilog이나 VHDL 등 특정 언어에 해당하는 옵션들을 선택할 수 있다.

이것은 행동 합성은 아니라고 Karnofsky 씨는 말했다. 그 이유는 Simulink가 시간의 관념을 가지고 있기 때문이다. HDL Coder는 untimed 디스크립션으로 시작하지 않는다. 그럼에도 불구하고 변환은 일어난다.

"Simulink는 자체 클럭을 가지고 있지만 이것은 Simulink 엔진을 통해 실행되는 샘플 속도에 상대적인 속도를 가지며, 칩의 클럭 속도와 동일하지 않다"고 Karnofsky 씨는 말했다. "변환은 Simulink 시간에서 칩의 실제 클럭 시간으로 이루어진다."

HDL Coder는 전체 칩이 아니라 Verilog이나 VHDL을 생성할 것이다. 이것은 데이터 경로나 알고리즘 집약적 디자인을 위한 것이라고 Karnofsky 씨는 말했다. 그러나 꼭 그러한 디자인에 한정될 필요는 없다고 한다.

합성 가능한 VHDL과 Verilog에 더하여 HDL Coder는 어떤 시뮬레이션 툴로도 읽을 수 있는 테스트벤치를 제공한다. HDL Coder는 또한 Synplicity의 Synplify Pro 합성툴과 Mentor의 ModelSim을 위한 스크립트들을 만들어낸다.

ModelSim 제품에 대한 MathWorks사의 링크로 유저들은 레거시 HDL 코드를 가져가 Simulink 환경에서 공동 시뮬레이팅 할 수 있다. 그러면 새로 생성되는 HDL은 레거시 HDL을 포함하게 된다. 그러나 이것은 ModelSim 시뮬레이터 유저들에게만 통한다.

모든 자동 생성 코드에 드는 질문 한 가지는 이것이 사람 손으로 만든 코드와 품질 면에서 어떻게 비견되는가 하는 것이다. "이 코드는 잘 구조화되고 해설되어있다"고 Karnofsky 씨는 말했다. "우리 고객 중에는 판독성과 품질에서 매우 만족하는 이들도 있다"고 그는 덧붙였다.

그러나 Karnofsky 씨는 "우리의 첫번째 제품이 전문적인 HDL 최적화와 동등하다고 주장하는 것은 아니다"라고 말했다. "나는 20년 이상 칩을 설계해 왔다"면서 MathWorks의 HDL 제품 매니저 Sudhir Sharma 씨는 "내 경험으로는 핸드 코딩으로 달성할 수 있는 품질에는 넓은 범위가 있다. 그러나 여기서는 모든 것이 Simulink 모델의 하이 레벨에서 캡처되기 때문에 RTL에 대해서는 별로 걱정하지 않아도 된다. Simulink 모델을 유지 관리하고 있더라도 시스템을 이해하기 위하여 코드를 판독하지 않아도 된다."

HDL Coder의 판매가는 1만 5,000달러이다. 유저들은 Matlab, Simulink, Fixed Point Toolbox 및Simulink Fixed Point를 이미 가지고 있어야 한다. 권장되는 제품은 Stateflow, Link for ModelSim, Signal Processing Toolbox, Signal Processing Blockset, Filter Design Toolbox가 있다.


Simulink는 HDL에 묶여있다. HDL Coder는 RTL Verilog과 VHDL을 생성시킨다.

- 출처: http://www.eetkorea.com/ART_8800439860_480103_NT_ac790059.HTM

2007년 11월 24일 토요일

초보자들을 위한 리눅스 가이드

게재: 2007년 09월 24일

Alexander Wolfe
InformationWeek


Love it, hate it, heard lots about it, but still don't have enough of a handle to form a firm opinion? Then we must be talking about Linux, the open-source OS that's alluring because it's heavy duty and it's free. Simultaneously, it's intimidating to newbies because it's typically more difficult to install and configure than Windows.

If you plan on giving the open-source OS a whirl, but, like the Georgia bride-to-be, got cold feet at the last minute, we've ferreted out six useful facts that'll ease your path when you decide to take the plunge.

1) How many versions of Linux are there?
Lots. At least 350, according to the list maintained by the enthusiast site DistroWatch.com. The site skews toward smaller distributions, with current flavor of the month Ubuntu listed as the most popular among the site's readers. Ubuntu has gained traction recently, garnering an endorsement from Sun Microsystems chief executive Jonathan Schwartz.

Ubuntu also appears to be gaining legitimacy via heavy grass-roots support. User-spawned Web resources include a blog devoted to the distro, a quick-start guide for dummies and a more advanced (how to install anything!) manual. (However, as What PC? points out, despite its funky name, Ubuntu is not noticeably simpler to get going than any other implementation of the OS.)

Ubuntu has a great back story: Its development was funded by South African Internet entrepreneur Mark Shuttleworth as an outgrowth of his efforts to offer improved educational opportunities to his nation's young people.

Another distribution much in demand is SUSE, available for free under the OpenSUSE.org program sponsored by Novell or in a for-pay version that comes with end-user support from Novell. Originally developed by German vendor SUSE Linux, the software has been heavily marketed to enterprise users ever since SUSE was acquired by Novell in 2004. Since that time, Novell has positioned itself as the main alternative to Red Hat, which is widely considered to be the leader in the enterprise Linux market. (In that regard, Novell CEO Jack Messman predicts that his company will emerge as one of the two dominant corporate suppliers of Linux, alongside Red Hat, as the market for paid open-source shakes out over the next two to five years.)

Other popular distros include Mandriva, Debian, and Fedora. (The latter is a free offering spun out of Red Hat. Don't forget Slackware, Knoppix, Gentoo Mepis, and others too numerous to mention.)

For those disinclined to deal with challenging installs, the easiest path may be Linspire. The eponymous company was founded by billionaire Michael Robertson, who made his money with the early Internet download service MP3.com. Robertson has positioned Linspire as consumer-friendly Windows alternative that costs a lot less—it's $50—and is bundled with many drivers and a bunch of applications.

2) What applications are available for Linux? And what the heck is LAMP?
The open-source app that gave Linux PCs a raison d'etre—and the one you need to set yourself up with if you expect to do anything useful—is OpenOffice.org.

Based on Sun's StarOffice, OpenOffice.org is a suite that's positioned as a free alternative to Microsoft Office (indeed, a Windows version is available). It's outfitted with word processing, spreadsheet, presentation (i.e, Powerpoint-style slides), and graphics programs. OpenOffice comes bundled with many Linux distros, including those from Red Hat and Novell. For roll-your-own types, some CD-ROM versions are available, but the easiest route is to just download the software.

While OpenOffice is the single most important Linux productivity package extant, the rap on the suite is that it has lingering issues regarding compatibility with MS Office file formats.

Other free, though far less popular, Linux office suites are GNOME Office and KOffice.

For the Web browser, Mozilla's Firefox and its companion Thunderbird e-mail client are a safe—and good—choice. They're even available on CD (for $6). Konqueror is another popular browser that's bundled with many distros.

The LAMP acronym that's kicked around so often refers to a "stack" of packages. Along with Linux, LAMP encompasses the Apache Web server and the MySQL database. The "P" is variously taken to refer to the PHP, Perl, or Python scripting languages. However, with the exception of the OS itself (and possibly MySQL), all those programs are of interest to developers, not average desktop users.

3) How can I listen to some tunes?
For many home users, once you get past word processing, the most important app is a music player. RealPlayer, famous for burrowing its way deep into Windows systems, is one of the few major players to offer a version for Linux. (It would hardly be fair to expect the same for programs named Windows Media Player and Winamp, though an open-source clone of the former is in the works.)

The Linux RealPlayer is based on the open-source Helix player, which offers downloads here. (However, since the Helix page also points to the Linux RealPlayer, it's easiest just to get that.) Another free player is amaroK from the KDE group.

Where to go to buy music is a tougher question. Neither iTunes, Yahoo Music nor Urge run under Linux. As for Rhapsody, users can get a subset of the service (basically, online music playing) but you can't buy songs online and you can't install the full version of Rhapsody on a Linux box. Most vexing is that, without persistent searching, it's hard to figure out precisely which pieces of the service work and which don't. A Rhapsody customer service answer attempts to explain; so does this Newsforge article.

One of the few operations that is set up to run under Linux is MP3tunes.com, the 88-cent-per-song online music store set up by the aforementioned Michael Robertson. If you're into artists off the beaten path, another service, called Mindawn, may be for you. While Mindawn doesn't have much music you've heard of, it does eschew DRM and offers its downloads in the FLAC format favored by PC audiophiles in the know.

Folks who store MP3s on their PCs are also often in charge of the family's digital photos. For them, there's word that Google's Picasa image management and sharing software has just been released on Linux.

4) What "desktop" environment should I use with my distro?
It's important to understand that when Linux people say "desktop," they don't mean your desktop. They're talking about your computer's user interface (UI). Given Linux's historical do-it-yourself culture, it's not surprising that the open-source UI was originally a separate element from the basic operating system. Today, nearly all distros come already packaged with one or both of the two main desktop environments for Linux: Gnome or KDE.

Conceptually, the desktops descended from the X Window system, a 1980s-era GUI from the Unix world.

At their current advanced stage of development, the differences between Gnome and KDE may be more political than technical. KDE is typically said to have more Windows-like bells and whistles, while Gnome is said to run faster. (Here's one user's perspective.)

Gnome benefits from its association with the Linux GNU Project, founded in 1984 by free-software advocate Richard Stallman. KDE, short for the K Desktop Environment, is considered a good choice for beginners.

Here's a list of distros that ship with KDE. Gnome is cagier and doesn't seem to provide a consolidated list. Many Linux distributions allow you to choose either one. That's the case for major Linux vendors Red Hat and Novell, even though both are members of the Gnome Foundation. Some distros do skew toward a single desktop. For example, Ubuntu ships with Gnome; Slackware packages KDE.

5) Linux on the server, Linux on the desktop—which is it?
Honestly? The server. Linux on the desktop hasn't taken off to the extent its adherents had hoped for. Perhaps it never will, according to noted analyst Rob Enderle, who told ComputerWorld that it's hard to out-compete Microsoft.

According to most authoritative estimates, Linux usage on the desktop hovers beneath 3 percent of all PCs.

However, that hasn't dimmed the expectations of the faithful, who predict wider desktop adoption is imminent. (Here are some presentations from the recent Desktop Linux Summit, which puts some technical meat on the bones of that optimism.)

Apart from the fact that most vendors can make more money with Windows, there are two major reasons more users haven't been convinced to take the Linux plunge. Most importantly, Windows is a one-stop operating system in a box. It ships complete with nearly all the drivers any user could every need, and can be up and running with several mouse clicks, a half-hour wait, and the entry of an annoying license code (which then has to be verified again online within 30 days to prove you didn't steal the thing).

However, for all its "free-ness," installing Linux usually requires more tweaking than most workaday PC users can handle. More of a stumbling block is the fact that drivers in the Linux world still aren't as widely available, nor are they as plug-and-play as their Windows counterparts. However, that situation continues to improve, which in turn accounts for the continued optimism of Linux pundits, who mostly believe that, as the driver and application-availability issues dissipate, users will come.

Perhaps surprisingly, Linux is seriously gearing up in one arena that usually escapes the attention of computer users. That's in the embedded sphere, where Linux is being used to power everything from smart phones to digital video recorders.

"The most visible example of Linux design wins in this area is TiVo and a range of television and video devices from Sony," said Bill Weinberg, senior technology analyst at the Open Source Development Labs (OSDL) in Beaverton, Ore.

Linux is appearing on some surprisingly diverse embedded platforms. For example, Sony's upcoming Playstation 3 will run Linux, in an apparent bid to encourage youthful developers to create games for its console.

Linux is also firmly entrenched as the operating system for point-of-sale terminals (aka cash registers). And here's a lightweight Linux computer for the wrist.

6) You've given me lots of facts, but not much advice. How do I get started?
One pain-free way to go (OK, it'll set you back $16, plus shipping) is by reading Test Driving Linux. The book, by David Brickner, includes a CD that allows you to boot Linux on a Windows computer without destroying the Windows install. On the downside, the book's Linux is, like the title says, a "test drive" that runs only off the CD; it won't permanently install the OS to your hard drive. (A further caveat is the CD is a bit fussy; it won't run if you can't get your PC to boot first from the CD drive. It didn't like my old Compaq desktop, for reasons unexplained, but it ran like a champ on an HP Pavilion laptop.)

If you're ready to give Linux a more permanent whirl, go back to Question 1, above, or to this list of distros. (The Wikipedia offers a "Which distro is right for you?" quiz.)

Alas, picking a distribution is easy compared to getting hold of the actual install. Parsing even the simplest Web page offering a free download of Linux is a major pain. The most effective way to download is to grab an ISO image, which is a file that you can burn directly onto a CD-ROM; most disk-authoring programs have an option to handle this. An ISO image to create an Ubuntu install CD is here. Scroll down almost to the bottom of this page and you can get an ISO image of an OpenSUSE boot CD. ISO images for 14 other distros, including Knoppix, Red Hat, Fedora, FreeBSD, and Slackware, are available here. (Many require multiple CDs.)

You can download an evaluation copy of Novell Linux Desktop 9 here.

If you want to avoid the download dance entirely, you can always spring for hard media, though disks are surprisingly difficult to find. A Debian distro is on CD for $10, here. The easiest route is to stop in at LinuxWorld Expo, where a spin through the show floor will net you dozens of free CD-ROMs.

When you're ready to do your install, the most important piece of advice I can provide—and one that you're unlikely to read upfront in most tutorials—is DON'T install Linux on the same hard drive on which your copy of Windows XP resides. Why? Because Windows is notoriously fussy about living alongside another OS. It might decide not to work. Plus, you risk erasing Windows entirely if your Linux CD engages in a session of drive formatting gone wild.

Better to dig up a second hard drive, and unplug the drive containing Windows for the duration of your Linux experiment (and, conversely, unplug the Linux drive when you're ready to return to Windows). Of course, if you're firing up Linux on an old machine that comes to you without an OS, this warning doesn't apply.

The other suggestion for prospective new users is to connect with others who are in the process of dipping their toes in the Linux waters. That's not as easy as it sounds. For some reason, Linux experts often can't seem to help themselves from adopting a schoolmarmish tone.

Of course, if you knew which darn distro to use, you wouldn't be trolling a newbie forum, would you?

- 출처: http://www.eetkorea.com/ART_8800480737_839585_NT_18e46c96.HTM

2007년 11월 23일 금요일

전자 메일 보내기: [글로벌 IT이슈 진단]스마트폰 인사이드

ETnews
[글로벌 IT이슈 진단]스마트폰 인사이드
[ 2007-11-21 ]
  '구글폰'이 세상이 공개됐다. 프로젝트 이름은 '안드로이드(Android)'. 이는 그리스어로 '인간을 꼭 빼닮은 물체'를 말한다. 흔히 로봇을 빗댈 때 사용한다. 로봇 중에서도 전자 두뇌와 인공 피부까지 갖춰 외관상 인간과 똑같을 정도로 진화한 로봇을 안드로이드라고 부른다. 우연의 일치일까. 지나칠 정도로 기술에 치중하는 구글의 철학이 여기에도 그대로 배어 있다. 구글은 구글폰을 2008년 중반에 내놓을 것이라고 공언했다.

구글폰이 공개되면서 '스마트폰'에 다시 관심이 쏠리고 있다. 과연 구글폰은 애플 아이폰만큼 휴대폰 시장에서 '태풍의 눈'으로 떠오를까. 아이폰이 최고의 스마트폰일까. 스마트폰은 실제 성능에서도 차이가 날까. 성능을 결정하는 주요 부품과 애플리케이션을 통해 스마트폰의 실제 경쟁력을 점검해 봤다.

'스마트폰의 대명사'로 불리는 아이폰, '멀티미디어 컴퓨터'로 불리는 최고급 사양의 노키아 스마트폰, 그리고 '사용자 애플리케이션'을 강조하고 저가폰·무료 통화를 내건 구글폰을 실험대에 올려 놨다. 비교 자료는 EE타임스·테크온라인·인포메이션 위크·로이터를 참조했다. 구글폰은 아직 출시하기 전임을 감안해 파트너로 언급한 부품업체를 중심으로 서비스와 기능을 따져 봤다.

# 애플 '아이폰' - 음악과 동영상에 최적화

애플이 지난 6월 선보인 아이폰. 신기술 평가에 인색한 얼리어댑터 사이에서도 스마트폰 역사를 새로 썼다는 찬사를 들을 정도로 화제였다. 아이폰은 실제 디자인과 유저 인터페이스 면에서 출시하자마자 '합격점(Cool)'을 받았다. 아이폰을 분해한 미국 기술 전문 사이트 '테크온라인'은 내부 구조는 휴대폰보다는 MP3플레이어인 '아이팟'과 유사하다고 진단했다. 실험팀은 분해하기 전 닌텐도 '위'와 마이크로소프트 'X박스' 설계 구조를 예상했으나 오히려 최신 아이팟 모델 쪽에 가깝다고 결론 내렸다. 한 마디로 음악·동영상과 같은 콘텐츠에 최적화했다는 것. 주요 부품 소싱도 아이팟 주요 파트너를 적극적으로 활용했다.

실험팀은 특히 애플이 핵심 부품을 자체 로고로 가려 이를 지우는 데 애를 먹었다. 애플이 애써 감추려고 했던 건 크게 세 부분이었다. 먼저 프로세서다. 프로세서는 예상대로 삼성전자 브랜드였다. 아키텍처는 ARM 플랫폼이었다. 삼성은 메인 프로세서와 함께 플래시 메모리도 함께 공급했다. 두 번째는 브로드컴 칩이었다. 정확한 기능은 파악하기 힘들지만 입출력 (I/O) 컨트롤러로 터치스크린을 위한 비디오 인터페이스 기능 정도로 내다봤다. 마지막으로 필립스가 디자인한 칩이었는데 세부 기능은 파악하기 힘들었다. 통신 기능과 관련해서는 인피니언이 크게 기여했다. 인피니언은 EDGE 기능을 지원하는 멀티미디어 엔진과 GSM RF 송수신기를 공급했다. 이 밖에 영국 CSR(블루투스), 마이크론(이미지 센서), 스카이웍스(전력 증폭기) 칩이 눈길을 끌었다.

 # 노키아 'N95' - 디지털과 아날로그의 조화

노키아가 '휴대폰 지존' 자존심을 걸고 아이폰 '대항마'로 내세운 스마트폰이 'N95'다. N95는 가격만 750달러로 노키아가 심혈을 기울인 프리미엄 제품이다. N95는 이 찬사에 걸맞게 위성항법장치(GPS)를 내장하고 HSDPA에서 WLAN·EDGE·WCDMA까지 지원하는 3세대 최신 모델이다. 업계 처음으로 '칼 짜이즈' 렌즈를 장착하고 DVD 클립 기능을 지원해 눈길을 끌었다. N95 내부를 뜯어 본(Under the hood) 'EE타임스'는 "신기술 플랫폼 경연장이었다"며 "휴대폰보다는 오히려 '멀티미디어 컴퓨터'에 가까웠다"고 강조했다. 최신 디지털 제품이지만 여전히 아날로그IC를 대거 사용한 점이 특이했다고 덧붙였다.

프리미엄 브랜드 'N80' 계보를 잇는 N95는 메인 통신 프로세서는 노키아 자체 아키텍처를 기반으로 ST마이크로가 설계했다. 노키아 칩에는 핀란드 말로 '욕심(Anheus)'이라는 마크가 있어 눈길을 끌었다. 무선 기술과 관련해서는 ST마이크로(802.11b/g), CSR(싱글 칩 블루투스), 텍사스인스트루먼츠(TI, GPS 통합 솔루션) 제품을 사용했다. 다른 스마트폰에 비해 아날로그IC를 많이 사용한 점이 독특했다. NXP의 FM 수신칩, 내셔널세미컨덕터의 dc/dc 컨버터, TI의 LED 드라이버, 아날로그 디바이스의 모터 드라이버 등이 디지털 부품과 자연스럽게 조화를 이루었다.

  # 구글 '안드로이드폰' - 가벼운 애플리케이션이 경쟁력

구글은 지난주 동영상을 통해 개발 중인 스마트폰을 선보였다. 창업자 세르게이 브린이 직접 시연자로 나선 이 동영상에서 구글은 "구글폰은 없다, 안드로이드만 있을 뿐"이라고 말했다. 이는 구글은 애플·노키아처럼 하드웨어 제조 기술이 없으며 대신에 사용자 인터페이스와 같은 애플리케이션에 더욱 초점을 맞추겠다는 의미다.

출시 카운트다운을 시작한 구글폰. 사실 제품이 나오기 전 성능을 정확하게 파헤치기는 불가능하다. IT 정보 사이트인 '인포메이션 위크'는 동맹군에 가담한 주요 업체를 기반으로 구글폰의 장·단점을 해부했다. 결론은 "애플리케이션에 주목하라"는 것. 먼저 유저 인터페이스. 스웨덴 TAT를 주목했다. TAT는 소니에릭슨·삼성전자의 파트너로 휴대폰 인터페이스 부문 최고 기술업체로 알려져 있다. 간편하면서도 단순한 인터페이스(Less is more)를 예상했다.

구글맵의 핵심 기능을 구현할 업체로는 SiRF 테크놀로지를 꼽았다. 멀티미디어 패키지 디자인과 플랫폼은 TI '몫'으로 돌렸다. 특히 TI가 3세대 '킬러 앱'으로 내놓은 'OMAP(Open Multimedia Application Platform)'가 탑재된다고 예상했다. 엔비디아(그래픽 칩), 시냅틱스(터치스크린), 누앙스(음성 인식)도 구글폰 핵심 기능이라고 설명했다. 특히 웹 브라우저와 관련해서는 동맹사는 아니지만 오페라의 최신 버전 '미니4'가 탑재되는 등 주로 (용량이) 가벼우면서 소비자가 사용하기 간편한 애플리케이션에 구글폰의 운명을 맡길 것으로 보인다고 예언했다.

강병준기자@전자신문, bjkang@
출력하기 닫기
Copyrightⓒ 2000-2005 ELECTRONIC TIMES INTERNET CO., LTD. All Rights Reserved.

전자 메일 보내기: 이파피루스, 편집 기능 내장 PDF 리더 무료 배포

 
ETnews
이파피루스, 편집 기능 내장 PDF 리더 무료 배포
[ 2007-11-23 ]  
 이파피루스(대표 김정희)는 자체 개발한 PDF 전용 리더 'e파피루스 리더'를 자사 홈페이지를 통해 무료 배포한다고 22일 밝혔다.

 이 제품은 PDF 파일보기 기능뿐 아니라 책갈피 추가와 메모 삽입, 그리기, 본문 검색, 형광펜, 밑줄긋기 등 다양한 문서 편집 기능들을 포함하고 있다. 특히 사용 PDF 편집 프로그램에서 제공하는 메모 삽입이나 책갈피 추가 등의 기능을 제공, 기업의 소프트웨어 구매 비용 절감에 도움이 될 것으로 회사 측은 기대했다.

 이파피루스는 e파피루스 리더를 개인·단체 구분없이 누구나 사용할 수 있는 완전 공개 프로그램으로 배포하고 있으며 지속적인 업그레이드와 기능 개발을 통해 단순한 PDF 리더 프로그램이 아닌 필수 문서 뷰어 프로그램으로 자리매김한다는 계획이다.

 이파피루스 서정호 상무는 "e파피루스 리더는 자체 PDF 엔진 기술을 바탕으로 개발, 성능과 안정성에서 다른 유사 프로그램보다 뛰어나다"고 말했다.

 e파피루스 리더는 이파피루스 홈페이지(www.pdfpro.co.kr)에서 받을 수 있다.

 한세희기자@전자신문, hahn@
출력하기 닫기
Copyrightⓒ 2000-2005 ELECTRONIC TIMES INTERNET CO., LTD. All Rights Reserved.

2007년 11월 11일 일요일

"나눔 배너, Nanum (Donation) Banner"를 달다...


꿈꾸는 아이들을 위한 Donors Camp :: 이벤트 "나눔배너 달면 공부방 아이들이 따뜻해 집니다!" 페이지를 보고 저도 참여했습니다... 블로그에 배너만 설치하면 아이들 공부방을 후원할 수 있는데 굳이 마다할 이유가 없죠...^^

2007년 11월 9일 금요일

전자 메일 보내기: FPGA 기반 프로토타입이 증가하는 이유


뉴스 및 동향

인쇄: 파일을 선택한 후 브라우저 메뉴에서 인쇄하십시오.

FPGA 기반 프로토타입이 증가하는 이유
게재: 2007년 11월 01일

By Andrew Haines
Senior VP of Marketing
Synplicity Inc.

ASIC과 SoC 비용의 급격한 상승 때문에 반도체 공급업체들은 각 디바이스들을 더 넓은 시장에 공급함으로써 만족스러운 투자 수익을 추구하고 있다. 소프트웨어 사용의 증가 추세는 그렇게 할 수 있는 효과적인 메커니즘을 제공한다. 왜냐하면 소프트웨어 내용물이 증가한다는 것은 기능이 늘어나는 것과 같기 때문이다. 따라서 소프트웨어의 변형은 시장별 제품 차별화를 제공한다.

따라서 ASIC 또는 SoC에 백만 줄의 소프트웨어 코드가 있는 것은 일반적이다. 게다가 멀티코어 사용의 증가는 이러한 장치에서 사용되는 소프트웨어의 엄청난 성장을 촉진하고 있으며 장치의 복잡성을 증가시키고 있다. 이러한 소프트웨어 사용의 증가는 전체 설계 프로세스에 어떤 영향을 미치는가?

소프트웨어의 복잡성이 증가함에 따라 적기 출시 목표를 달성하려면 설계 주기에서 매우 빠른 시기에 소프트웨어 개발을 시작해야 한다. 새로운 장치의 경우 소프트웨어가 매우 복잡하기 때문에 개발하는 데 18개월 이상이 걸릴 수 있으므로 칩 출시 일자보다 훨씬 앞서 소프트웨어 개발을 시작해야 한다. 하드웨어가 개발될 때까지 기다리면 너무 늦다.

FPGA 기반 프로토타입은 에뮬레이션과 시뮬레이션 같은 대체 검증 방법보다 최소 10배 빠른 속도로 소프트웨어를 실행시킬 수 있기 때문에 이상적인 소프트웨어 개발 플랫폼이다. 각 검증 방법은 검증 흐름에서 자신의 기능이 있지만 FPGA 기반 프로토타입만이 매우 복잡한 소프트웨어 응용 프로그램과 운영 체제를 실행하는데 충분한 빠른 속도로 실행된다. 최적 시스템 통합 및 소프트웨어 테스트는 10~50MHz 이상 또는 최대 100MHz의 속도가 필요하다. 그리고 FPGA 기반 프로토타입이 이러한 속도를 얻을 수 있는 유일한 솔루션이며, 그렇기 때문에 그 사용이 증가하고 있다.

FPGA 기반 프로토타입의 사용이 증가하는 또 다른 중요한 이유는 낮은 비용이다. 요즈음 100 ~ 200명의 소프트웨어 개발자가 복잡한 SoC에 대한 작업을 하는 것은 드문 일이 아니다. FPGA 기반 프로토타입은 저렴하고 배치 가능하기 때문에 여러 소프트웨어 개발 팀원들에게 광범위하게 배포된다. 따라서 대부분의 소프트웨어 개발 팀원은 장소와 관계 없이 다른 방식보다 여러 달 먼저 작업을 진행할 수 있다.

어려운 프로세스

소프트웨어와 하드웨어를 통합하는 것은 어려운 프로세스이다. 설계 팀은 모든 사양이 계획대로 구현되었는지 확인해야 한다. 이러한 작업은 보통, 하드웨어 설계를 철저히 테스트하고 버그가 없는 것을 확인한 후에 실행된다.

그러나 세밀한 검증 후에도 틀림없이 소프트웨어와 하드웨어를 결합할 때 발견하기 어려운 버그가 늦게 발견된다. 복잡한 SoC의 버그는 깊이 숨어 있다가 복잡하고 예측할 수 없는 하드웨어와 소프트웨어 간의 상호 작용 때문에 비결정론적으로 나타난다. 이러한 버그를 찾기만 하는 작업도 매우 길고 시간이 오래 걸리는 테스트 시퀀스가 필요하며 느린 검증 방법으로는 불가능한 일이다. 게다가 문제가 발견된 후 실질적인 설계 디버깅 또한 상당한 시간과 노력이 필요하다. Synplicity의 TotalRecall 완전 가시성 기술과 같은 프로토타입 디버그에 대한 새로운 방식은 찾기 어려운 버그에 대한 테스트 벤치를 확보하고 시뮬레이터에 다시 표시할 수 있어 이러한 작업을 단순화한다.

FPGA 기반 프로토타입에는 설계 검증과 관련이 없는 흥미롭고 강력한 기능이 있다. 즉 설계 초기에 아키텍처 조사에 FPGA 기반 프로토타입을 사용할 수 있다. 예를 들어, 설계자가 알고리즘이나 구현을 달리할 때 성능이 어떻게 변화하는지 알아보기 위해 DSP 또는 그래픽 알고리즘을 조사해야 할 경우가 있다. 이러한 경우 아키텍처 변경이 비디오 그래픽 알고리즘에 미치는 영향을 평가하려면 하드웨어 성능이 필요하다. FPGA 기반 프로토타입을 사용하여 설계자는 설계 초기에 이 조사를 수행할 수 있다. 그럼으로써 실제 SoC의 성능을 보기 전에 효과적인 부품이 선택되었는지 확인할 수 있다.

고객의 반응, 고객 조사, 공급업체 반응을 기준으로 Synplicity는 SoC와 ASIC의 90퍼센트 이상이 오늘날 FPGA로 프로토타입이 만들어지고 있다고 추정하였다. ASIC 및 SoC 프로토타입의 수요는 두 자리 비율로 증가하고 있으며, 시장 조사 회사인 Gary Smith EDA의 멀티코어 사용 예측에 따르면 앞으로도 수요는 지속적으로 증가할 것이다.

이러한 장점에도 불구하고 FPGA 기반 프로토타입은 다른 검증 방식을 대체하지 않는다. 하지만 FPGA 기반 프로토타입은 오늘날 사용할 수 있는 중요한 기법 중 하나이다. 사실 검증 방법과 통합의 향상은 향후 개발의 주요 수단이다.

프로토타입 실행 속도를 높이고 변경을 용이하게 하기 위해 많은 노력을 할 수 있다. 긍극적인 목표는 설계자가 신속하게 작업에 착수하고 신속하게 디버깅하며 지체 없이 변경할 수 있는 환경을 만드는 것이다. 이 새로운 시대에 FPGA 기반 프로토타입은 다른 검증 방법과 함께 중요한 위치를 점하게 될 것이다. "at-speed"로 실행하는 기능은(예: 하드웨어가 작동하는 속도로 테스트를 실행함) SoC 검증의 필수적 요소이다. 그래야만 설계자가 장치와 소프트웨어 복잡성의 증가와 관련하여 새로 발생하는 문제를 처리할 수 있을 것이다.



<이번호 저널 2007년 11월 1일~15일>자에서 이 기사 및 다른 기사들도 찾아볼 수 있습니다.


본 기사는 에 있는 전자 엔지니어 기사에서 인쇄한 것입니다. http://www.eetkorea.com/ART_8800486712_839585_NT_547d6045.HTM

이전 기사로 전자 엔지니어

CSR enables $15 Wi-Fi Internet radio design


New Products

To print: Select File and then Print from your browser's menu.

CSR enables $15 Wi-Fi Internet radio design
Posted: 08 Nov 2007

Wireless chip specialist CSR is readying a Wi-Fi Internet radio reference design that it says has an electronics BOM of just $15 in volume, and thus changes the economics of designing products for Internet radio reception.

Dubbed the RadiPro, the design incorporates CSR's UniFi single-chip Wi-Fi technology, a multimedia applications processor that combines the company's proprietary Kalimba DSP and RISC cores and a high quality 16bit codec, all based on CSR's BlueCore-5 architecture.

According to Ronen Kenig, product marketing manager at CSR's consumer business unit, products using the RadioPro will be available in shops next month.

He adds CSR believes that by 2009, 40 percent of MP3 and portable media players will offer Internet radio capabilities, as would half of all DAB and satellite radios and a third of home Hi-Fi systems. Other applications are likely to include iPod docking stations and set top boxes.

The low power design of RadioPro boasts up to 25hrs of active streaming time running on a 1,500mAH battery. RadioPro provides wireless streaming of internet radio via Wi-Fi, without the need for a PC, and can support over 10,000 radio stations through a dedicated internet radio portal. CSR has teamed with specialist company vTuner to provide the portal with its designs, but this portal link can be exchanged if necessary.

The design also includes all the hardware drivers and streaming protocols needed, as well as MP3, WMA and RealAudio codec support at launch. Also included is over-the-air field upgradeability, allowing operators to fix bugs, add new software features including further codecs, and new services.

CSR's roadmap for the part suggests additional codecs, for instance ACC, will be available by the middle of next year, as well as FM receiver and Bluetooth capabilities, support for Asian fonts, and streaming from Digital Media Servers.

Figures published by RAJAR (Radio Joint Audience Research) show that almost a quarter of the U.K. population listened to radio via the internet in March 2007 alone. However, as a relatively new entrant in the home entertainment arena, the initial cost for specific internet radio products has so far come with a hefty price tag of no less than $200.

CSR says with its BOM of $15, ODMs would be able to bring internet radio products to consumers for much less, typically $60 to $70 for a complete product according to Kenig.

- John Walko
EE Times Europe


This article was printed from EE Times-Asia located at: http://www.eetasia.com/ART_8800488056_499488_NP_e9041e6c.HTM

Back to Article Home

2007년 10월 30일 화요일

D/A 컨버터에 대해 이해하는 날이 오려나...?


DAC(Digital-to-Analog Converter)는 2진수를 그 값에 비례하는 전압으로 변환시킨다. DAC를 만들기 위해서 필요한 외부 회로는 Low-Pass Filter를 위한 1개의 저항과 1개의 커패시터만 있으면 된다. FPGA 자원의 사용량도 적다. Delta-Sigma DAC는 디지털 기법을 이용하기 때문에 온도 변화에 무관하고, 프로그램 가능 논리로 구현이 가능하다. Delta-Sigma는 사실 고속의 단일 비트 DAC이며, 디지털 피드백 을 사용하여 펄스열이 생성된다. 펄스열이 평균 Duty_cycle은 2진 입력 값에 비례한다. 펄스열이 Low-Pass Filter를 지나면 아날로그 신호가 만들어진다.

Delta-Sigma DAC의 Block Diagram이다. Delta는 산술적 차이를 뜻하고, Sigma는 산술적 합을 뜻한다. 여기서는 이진 가산기를 이용해서 차이와 합을 연산한다.

데이터 입력은 8비트 헥사값으로 입력이 이루어진다. DeltaB는 Sigma Latch에 저장된 값으로 10bit이다. 그래서 최상위 비트 L[9]를 2개 복사해 binary로 표현되고 나머지 8bit 는 0 값으로 변환되어 DeltaB로 입력 된다. 각각 Adder와 Latch에서 계산되고 저장된 값을 나타내었다. Delta Adder의 입력이 무부호이기는 하지만 두 개의 가산기의 출력은 유부호로 간주된다. Delta Adder는 DACin과 현재 DAC 출력의 차이를 계산하다. 이 값과 Sigma Latch에 저장되어있는 이전 출력의 값이 Sigma Adder에서 더해진다.


- 출처: http://cafe.naver.com/plduser/607

CMRR과 PSRR에 대해서 자세하게 설명해주세요.

[질문] ----------

CMRR이 Common Mode Rejection Ratio 인것과, PSRR이 Power Supply Rejection Ratio 인것은 알겠는데..

정확히 어떤 상황에서 어떻게 되는건지 이해를 하지 못하겠습니다.

좀 자세하게 설명해주실뿐 없으신가오\;ㅅ;

[답변] ----------

CMRR 은 동상제거비 라고하고 하는데 OPAMP의 경우 두개의 입력단자가 있습니다.

반전입력단자와 비반전 입력단자입니다. (여기서 반전,비반전의 의미는 입력에 대해 출력의 위상을 의미합니다.) 이상태에서 출력은 G x (비반전입력 - 반전입력) 가 됩니다.

따라서 입력에 똑같은 신호를 넣어주면 출력으로는 아무것도 나오지 말아야 하죠.

그런데 아주 미미하지만 출력이 나오는데 이때와 정상상태에서의 신호비를 CMRR 이라 합니다. 보통 80~120dB정도 되므로 똑같이 1V입력을 넣어주면 출력으로 약 1uV~0.1mV정도가 나온다는 것이죠.

PSRR은 전원신호에 대한 제거비입니다.

전원이 아주 깨끗하면 좋겠지만 실제 우리가 받는 전원은 노이즈와 전원리플등이 포함되어 있습니다. 또한 다른 기기에서 유기되는 신호도 있는데 이렇게 전원에 포함된 신호가 출력으로 나오면 안되지만 이또한 미미하게 출력으로 나옵니다. 이것에 대한 신호비를 PSRR 이라 합니다.

위의 CMRR,PSRR은 둘다 높을수록 좋지만 일반적으로 주파수가 높아지면 위의 두가지 값도 악화가 됩니다.

참고하세요.

[출처] ----------

http://kin.naver.com/db/detail.php?d1id=11&dir_id=110209&eid=VdMKK4mTTdQlkpo6FgXJ1qXD8bZHf/5A&qb=cHNycg==

2007년 10월 22일 월요일

Start~!!!


출처: http://munizone.tistory.com/entry/Start

성공적인 스트럭처드 ASIC 설계

성공적인 스트럭처드 ASIC 설계
게재 : 2007년 09월 03일

By Frank McMillan
Senior Applications Engineer / ChipX Inc.
E-mail: frankm@chipx.com

딥서브미크론 스탠다드 셀 ASIC에 수반되는 비경상 엔지니어링 비용이 증가하고 설계 기간이 늘어남에 따라 주문형 IC 디자인에 스트럭처드 ASIC을 이용하는 것은 점점 더 매력적인 옵션이 되어가고 있다. 스트럭처드 ASIC은 경쟁력 있는 단가로 우수한 실리콘 성능과 훨씬 낮은 NRE를 제공한다. 스트럭처드 ASIC에 광범위한 사이즈가 이용될 수 있다는 것은 주요 시스템 칩으로도 사용될 수 있고 작고 경제적인 보조칩으로도 사용될 수 있다는 것을 의미한다.

스트럭처드 ASIC 슬라이스 디자인에서 많은 물리적 설계 문제들이 이미 해결되었다. 따라서 백엔드 레이아웃 작업에 들어가는 시간이 줄어들 수 있으며, 이는 사인오프와 시제품 전달을 앞당긴다. 그러나 ASIC 슬라이스들은 사전 정의된 스트럭처를 가지고 있고, 따라서 설계자들은 원하는 성능을 얻기 위해서 칩 리소스의 배열을 세심하게 고려해야 한다.

FPGA는 ASIC에 대한 또 다른 대안이 될 수 있다. 이것들은 보통 룩업 테이블과 컨피규러블 로직 셀들을 기반으로 하며, 이에 견줄만한 ASIC 기술에 비해 면적 효율이 떨어지고 전력은 더 많이 잡아먹는다. 연간 생산량이 일단 5,000 피스를 넘으면 스트럭처드 ASIC을 이용하는 것이 훨씬 경제적이다. 그러나 아직 많은 어플리케이션에서 FPGA가 낮은 업프론트 비용으로 빠른 턴어라운드를 제공하는 이상적인 시제품 제작 툴이 되어주고 있다.

디자인의 시제품이 FPGA로 제작되었다면 스트럭처드 ASIC으로의 이동을 계획하는 것이 중요하다. 이상적으로는, FPGA와 ASIC 라이브러리를 이용한 동시 개발이 일찍 이루어질 수 있다. 그러나 그렇게 하지 않았다고 하더라도 조금만 계획하면 마이그레이션이 훨씬 고통없이 이루어질 수 있다.

해야 할 일

- 미스매치를 피하는 합성툴 전략을 채택하라. 보통 설계자들은 벤더 네트리스트이든 아니면 RTL에서의 HDL 디스크립션이든 자신들이 선택한 프론트엔드 환경을 사용할 수 있다. 그러나 FPGA 합성이나 ASIC 합성에는 서로 다른 툴이나 동일한 툴의 서로 다른 버전이 사용될 수 있다. 코드 체커와 least-common-denominator 코딩 방식을 이용하여 결과의 불일치를 피하라. 그렇게 하여 서로 다른 툴들이 RTL 코드를 서로 다르게 해석하지 않도록 한다.

- 상세한 타이밍 정보를 전달하라. ASIC 벤더는 합성, 타이밍 주도 레이아웃, 포스트 레이아웃 정적 타이밍 분석을 수행하는데 그러한 정보를 필요로 한다. 좋은 시스템 클럭 정보와 I/O 시스템 타이밍 예산, FPGA 합성 스크립트 상의 잘못된 경로 및 멀티사이클 경로 정보를 제공함으로써 레이아웃 속도를 올릴 수 있다.

- ASIC 공급업체와 DFT 요건에 대하여 빨리 논의하라. FPGA 프로토타입에 테스트 회로가 포함될 필요는 없지만 ASIC내에 테스트할 수 없는 회로가 있다면 디바이스의 결함 커버리지를 낮출 뿐만 아니라 불량 부품이 테스터를 통과하게 할 수 있다. 몇몇 ASIC 벤더들은 자신들의 NRE에 테스트 삽입과 자동 테스트 프로그램 생성을 포함시키기도 한다. 그러나 전용 테스트핀 또는 다중화된 테스트핀이 필요할 수도 있으며, 추가적인 테스트 회로가 필요할 수도 있다.

- 리셋과 초기화 로직을 추가하라. FPGA가 필요로 하지 않는다고 할지라도 리셋과 초기화 로직을 추가하여 완성된 프로토타입에 사용된 테스트 벡터들이 시뮬레이션에서 동일한 결과를 얻을 수 있도록 해야 한다.

- 패키지 스타일과 핀아웃을 일찍 결정하라. 스트럭처드 ASIC 벤더들은 비용과 보드 면적을 절약하면서도 설계 요건에 정확하게 맞는 넓은 범위의 패키지를 제공할 수 있다. FPGA에서는 패키지 선택이 제한된다. FPGA 프로토타입과의 핀 호환성이 요구된다면 ASIC 벤더와 빨리 핀아웃과 패키지 선택에 대하여 논의함으로써 요건을 충족시킬 수 있도록 해야 한다.

해선 안될 일

- ASIC에 맞지않는 프로토타입 FPGA RAM 기능을 이용하지 마라. 스트럭처드 ASIC의 RAM은 보통 구성 능력이 뛰어나기 때문에 서로 다른 RAM 사이즈가 문제가 되지는 않는다. 그렇지만 전체 RAM 용량과 이용 가능한 인스턴스의 수를 넘어서면 안 된다. 비동기식 액세스와 서로 다른 워드 크기를 갖는 비대칭 read/write 포트를 갖는 것은 가능하지 않거나 여분의 wrapper, 그리고 로직의 추가를 필요로 할 수 있다.

- ASIC 라인업에서 사용할 수 없는 FPGA I/O를 선택해서는 안 된다. ASIC 라이브러리를 확인하고 거기서 이용할 수 있는 것을 매치시켜라. 그렇게 함으로써 ASIC을 시제품 보드에 끼워넣을 때 깜짝 놀라는 일을 피할 수 있을 것이다.

- FPGA 벤더들이 공급한 고유 IP를 이용하지 마라. 가급적 이름있는 업체에서 만든, FPGA와 ASIC 모두에 쉽게 이용될 수 있는 합성 가능한 IP를 사용하라. PLL 같은 아날로그 IP가 사용되는 곳에서 타깃 ASIC 벤더가 디자인이 요구하는 주파수 생성이나 디스큐(deskew) 요건에 맞는지 확실히 해야 한다.

- 긴 인터커넥트를 사용해서는 안 된다. 백엔드 툴들은 연결된 회로들을 가까이 배치함으로써 인터커넥트 지연을 감소시켜준다. 그러나 스트럭처드 ASIC에서는 RAM 위치가 고정된다. RAM을 다이 하단의 I/O에 수반되는 회로와 함께 다이 상단에 연결하면 지연이 증가할 것이다. 임계 타이밍 인터페이스를 가지고 있다면 레이아웃 작업 후 긴 인터커넥트를 만들어내는 핀아웃과 RAM 배치를 피하라.

- 가능한 한 비동기식 로직을 사용해서는 안 된다. 스트럭처드 ASIC에 FPGA 비동기식 로직 경로를 재생성시킬 수 없을 것이다.


ASIC 개발에 수반되는 NRE 비용과 설계 시간의 증가로 스트럭처드 ASIC이 더 매력적인 옵션이 되고 있다.



출처: http://www.eetkorea.com/ART_8800477935_839575_NT_d9b5c8da.HTM

2007년 10월 11일 목요일

Low-power portable product design with FPGAs

출처: http://www.pldesignline.com/202400981;jsessionid=M52MFF2H1ZE0MQSNDLOSKH0CJUNN2JVN?printableArticle=true

Low-power portable product design with FPGAs

Flash-based FPGAs featuring sophisticated low-power operating modes allow designers to quickly develop products that maximize battery life.

By Mike Thompson, Actel Corporation

October 10, 2007

The past decade has seen massive growth in portable products. When combined with time-to-market pressures and the increasing need for flexibility, this growth makes new low-power field-programmable gate arrays (FPGAs) ideal platforms for the development of these consumer and industrial applications. Historically, FPGAs and programmable logic devices (PLDs) have been notoriously power hungry. Fortunately, this paradigm is changing. Advances in FPGA design have dramatically lowered the power consumption of new devices, making them an ideal solution for battery-powered applications.

At the same time, the embedded market continues to move toward 32-bit processing to handle the increasing computational needs of today's cutting-edge designs. It has been difficult to find a broadly used, industry-standard processor that can be implemented efficiently in the course-grained architecture of FPGAs. This has changed with the availability of the FPGA-optimized ARM Cortex-M1 processor. When coupled with power-efficient, flash-based FPGAs, the 32-bit Cortex-M1 offers designers a flexible system construction platform for building portable products that offer maximum battery life.

The new power paradigm
The relevant physics of integrated circuit power consumption is changing as process geometries shrink. In the past, dynamic power dominated and the power supply could be lowered with every successive process shrink. Lower voltages meant less dynamic power, and the trend continued in the right direction, but lowering operating voltages is no longer possible. Additionally, the physics of semiconductors at smaller process geometries has dramatically increased static power related to leakage. Process technology has shrunk to the point that static power is becoming a greater issue for portable applications than dynamic power. This is especially true when maximum temperatures are considered where leakage currents can increase by an order of magnitude or more.

Due to the increasing impact of leakage and ongoing efforts to increase power efficiency, non-volatile flash-based FPGAs have been able to approach and, in some cases, beat application-specific integrated circuit (ASIC) and application-specific standard product (ASSP) power efficiencies. These technical changes and innovations, such as new power optimization modes, have enabled FPGAs to demonstrate dramatically lower static power consumption. This makes them an ideal solution for portable applications that must also balance flexibility and the ability to accommodate the ever-changing standards for end products.

Combined with reduced dynamic power, their surprisingly low static power numbers enable flash-based FPGAs to provide lower total system power in many cases than ASIC and ASSP solutions. As a result, the perceived FPGA power penalty is much greater than what is seen today in actual designs, and when the flexibility of FPGAs is taken into account, it is a wonder that ASICs and some ASSPs still find a place in the market.

The new generation of low-power FPGAs
A new generation of single-chip, nonvolatile flash-based FPGAs eliminates the power up current spike associated with the configuration of traditional FPGA technologies. As a result, programmable logic can achieve new levels of power efficiency. Now, sophisticated low power and sleep modes that were only previously available on ASSPs are showing up on flash-based FPGAs.

Traditional SRAM-based FPGAs lose their configuration while in sleep mode and therefore require reconfiguration, which takes hundreds of milliseconds and consumes hundreds of milliwatts of power, so low power modes and especially sleep modes are superfluous. Flash-based FPGAs, on the other hand, offer a variety of low power modes and can be put to sleep and woken up just like ASSPs, because the live-at-power-up feature enables immediate operation of the device when power returns. Furthermore, new technologies (like Flash*Freeze from Actel) are enabling easy entry and exit from the ultra-low power modes while retaining SRAM and register data.

Another important low power feature of flash-based FPGAs is that they don't require additional support devices to function properly. Often, SRAM-based FPGAs need an external memory to store the configuration bitstream, a CPLD to act as a configuration controller, and a device to trap brownouts and power glitches so the FPGA will be properly reset and reconfigured. In addition to the added board space and cost, each extra component adds to the overall power profile of SRAM FPGA solutions, these added components make system power-up more complex. Flash-based FPGAs do not require external components to operate, enabling them to provide more than five times longer battery life in portable applications than other low power programmable logic devices.

ARM Cortex-M1 – the FPGA processor
Efficient low-power design requires more than just the consideration of which FPGA to use. The need for processing power is increasing, and the decision as to which processor to use can have significant impact on overall power usage.

ARM has a reputation for offering the industry's lowest power consumption with its processors. When combined with nonvolatile flash-based FPGAs, the small and fast ARM Cortex-M1 processor offers a number of low-power benefits. Designed for implementation in FPGAs, the processor offers an ideal balance between size and speed enabling maximum power efficiency with good embedded performance. Based on the ARM architecture, the Cortex-M1 processor is supported by a large ecosystem of tools, but – unlike other industry-standard processors – it can be implemented effectively in the coarse-grained architecture of FPGAs.

Derived from ARM's three-stage Cortex-M3 pipeline, the Cortex-M1 provides a good balance in terms of power efficiency between hardware and software resources. The processor is designed to offer good embedded performance and a complete instruction set, but lesser-used, high-end processor features have been left off to keep size and power consumption to a minimum.


그림1. The Cortex-M1 is tailored for FPGA-based implementation.

The Cortex-M1 runs a subset of the new Thumb-2 instruction set. It also features support for tightly coupled memory and a sophisticated low-latency interrupt controller to improve embedded performance and maximize power efficiency. The Thumb-2 technology uses 31 percent less memory reducing cost and power, while providing up to 38 percent higher performance than existing high density code, which can be used to prolong battery life or to enrich the product feature set.

One of the benefits of Thumb-2 over previous ARM instruction set architectures is that 16- and 32-bit instructions are executed in the same mode. It is a big advantage to be able to freely mix 16- and 32-bit instructions, as this reduces interrupt latency and power consumption. Interrupts can be serviced in 16-bit mode, for example, so the additional clock cycles required to get to 32-bit mode are no longer needed. To further reduce interrupt latency, ARM included a configurable Nested Vectored Interrupt Controller giving designers greater control to manage the power used to service interrupts and exceptions.

The portable product solution
The market for portable products will continue to grow, forcing designers to find ways to make their applications run longer with more features from finite battery resources.


그림2. Battery life experiment – SmartPhone.


그림3. Battery life experiment – Handheld Radio.

The trend toward portable applications will continue. Low-power, flash-based FPGAs coupled with FPGA-centric embedded processors like the ARM Cortex-M1 provide designers with a flexible system constructions platform for building cutting-edge portable products.

Mike Thompson is senior manager, IP product marketing, at Actel Corporation where he is responsible for the development and infrastructure support of new microprocessor IP cores for use in Actel FPGAs. Mike has more than 25 years of experience in both design and support of IP, microprocessors, microcontrollers and the development of embedded applications working previously for MIPS, ZiLOG, Philips/Signetics, and AMD. He has a BSEE from Northern Illinois University and an MBA from Santa Clara University. You can contact Mike at mike.thompson@actel.com.