| HITACHI HOME | UP | SEARCH |

HITACHI TECHNOLOGY ’99Hitachi Review Special IssueJuly 1999

• Backbone Servers for the Era of Network Computing —the MP5800E and MP5600E Series—

• DVD Library System for Implementing a Massive Storage Network Warehouse at Low Cost

• Optical Wavelength-division Multiplexing Transmission System Capable of 320-Gbit/s Transmission

• M-700 Series Microwave Plasma Etching System for Handling Large-diameter 300-mm Wafers

• The World’s Largest Helical Fusion Experimental Device, Completed by the Ministry of Education’s NIFS

• Renewal of the Central Load Dispatching Center at the Kansai Electric Power Co., Inc. for the 21st Century

• An Environmental Management Support System to Help Users Gain ISO 14001 Certification

• Design of the West Japan Railway Company’s Series 500 Shinkansen Train —Winner of the 1998 JIII’s Inventions Award—

• A High-throughput, Nanometer-level Ultrathin-film Evaluator

• “PAM Refrigerator” Series: The World’s First Refrigerators with PAM Control Mode

• A 128-Mbit Single-electron Memory: The First-ever Super-highly Integrated Memory Operable at Room Temperature

top of this page

All Rights Reserved, Copyright © 1994, 1999, Hitachi, Ltd.

Backbone Servers for the Era of Network Computing—the MP5800E and MP5600E Series—

1

Network computing systems that enablecompanies to exploit the vast data assets storedon their mainframes through global access totheir databases via the Internet and corporateintranets and strategic use of core data throughdata warehousing are drawing increasedattention. This has led Hitachi to strengthen itslineup of M Parallel Series MP5800E andMP5600E model backbone servers andenhance its VOS3/FS operating system.

Seamless Integration with OpenSystems —Objectives in bolstering the MParallel Series

Against the backdrop of deregulation andthe financial industry’s “Big Bang,” we’veentered an era in which the ability ofbusinesses to rapidly respond to market needsand provide value-added customer services ismore important than ever before. This has ledcompanies to reassess the value of mainframesthat are clearly superior in terms of overall cost(deployment, operations management, andmaintenance), but that also ensure high-speedreliable management of large-scale on-linedatabase processing, large-scale batchprocessing, and backbone networking. It isinevitable that in the environment nowemerging in which various open systems aredistributed in a confused and complex way,mainframes will come to assume anincreasingly important role as backboneservers. Our intent, therefore, is to providestate-of-the-art processors and architecture thatwill serve as the nucleus of a networkcomputing system while building on traditionalstrengths and accommodating the latesttechnologies and leading-edge open-application environment.

Supports Java and DistributedObject Technology —Features of the new MP5800E andMP5600E models

Compared to the MP5800 that uses ACEtechnology to provide the world’s fastestuniprocessor performance, the MP5800Emodels offer 1.3-fold improvement inprocessing performance, a 15% improvementin price-performance ratio, and support one- toeight-way processor configurations. TheMP5800E models also provide built-in LANadapters and enhanced network connectivity.

Similarly, compared to the MP5600 that setnew records for compact implementation andcost-performance by adopting cutting-edge

seamless interworking between new and legacyapplications, it also provides a totalmanagement approach, and enhances themainframe’s functionality, reliability, andoperability as a backbone server. —Future development

Going into the 21st century, Hitachi plans toredouble its mainframe development efforts inorder to support the global competitivenessdemanded by businesses. While we arecurrently focusing most of our efforts ondeveloping a next-generation M parallel seriesmodel to succeed the ACE technology-basedMP5800, we also plan to commercialize a newaddition to the MP5600 line that fully exploits arange of state-of-the-art technologies.

CMOS technology throughout and enhancedparallel processing technology, the E modelversion achieved a 2-fold improvement inprocessing performance and a 15%improvement in price-performance ratio. Theresult is a range of scalability representing a100-fold increase in power across the samegroup. This provides a broad array from whichto select and expandability that enables acompany’s mainframe to grow as its businessexpands. —Evolution of the VOS3/FS operatingsystem

Even while strengthening the operatingsystem’s most basic capabilities of supportingcontinuous parallel processing of database,batch, and transaction processing 24 hours aday 365 days a year, it has also beensignificantly enhanced in other ways. The newVOS3/FS is designed to provide support forJava, WWW, distributed objects, datawarehousing, and other open applicationtechnologies, as well as better connectivity withopen networks. This not only supports

Akira Yamaoka (left) and MasayaWatanabe (middle) of theEnterprise Server Division andKazuo Imai (right) of the SoftwareDivision were instrumental indeveloping the MP5800E andMP5600E models.

Java is a trademark or a registered trademark of SunMicrosystems, Inc. in the U.S. and other countries.ACE is an acronym for Advanced CMOS-ECL, a Hitachiproprietary technology in which a CMOS (ComplementaryMetal-Oxide Semiconductor) and a bipolar ECL (EmitterCoupled Logic) element are monolithically integrated on thesame chip.

MP5600E Model

As the market for electronic data services andcontents providers using digital satellitebroadcasting and the Internet has continued toexpand, this has fueled an increasing demandfor systems that can accommodate massiveamounts of video, audio, and other multimediadata. At the same time, we are seeing anincreasing number of data warehousingdeployments and the sheer volume of datahandled by ordinary businesses is exploding.Under these circumstances, conventionaloptical and tape devices are giving out as theyreach their natural limits in terms of capacity,long-term storage retention, and operability. Itwas these considerations which led us todevelop a DVD library system that isimplemented with DVD-RAM (Digital VersatileDisc Random Access Memory), the newtechnology that is attracting widespreadattention as the next-generation large-capacityuniversal storage media.

An All-in-One System Integrat-ing Software and Hardware—Features of the DVD library system

Implemented using 100-by-150 DVD-RAMdiscs each having a memory capacity of 2.6gigabytes per side and featuring a built-in high-speed autochanger, the system provides amassive memory storage capacity of 260 by390 gigabytes. Since DVD-RAM discs supportread/write capabilities and of course can berandomly accessed, files can be flexiblyupdated or deleted. The system has excellentlong-term memory retention, and consideringits compact implementation and lower cost perbit compared with conventional optical media,this offers the ideal solution for long-termstorage of massive amounts of data.

Most significantly, the system provides an

all-in-one intelligent storage solution. It comesas an integrated system package with a fullrange of application-specific software, so thereis no need to purchase additional hardware orsoftware to deploy and run the system. Thesystem can also be brought on line very easilyby simply plugging it in as is to your existingnetwork, because it supports the openarchitecture industry standard TCP/IP protocoland Windows NT.

An Extensive Lineup Rangingfrom Home to Office Use—Management and maintenance

The DVD library system can be installedanywhere there is a network environment, andthus does not have to be installed near thesystem operator’s terminal as in the past. Whatis more, the system operator can set up and runthe system using a web browser on Windows95 or NT. A special system operator terminal is

therefore not required, and the status of thesystem can be easily monitored from a remotelocation. The system is implemented verycompactly, and introducing and maintaining thesystem is facilitated by the fact that it comes inan air-tight case and doesn’t require a keyboard,mouse, or display. —Who are potential users?

The system would be ideal for such media-oriented enterprises as electronic art museums,natural history museums, and electroniclibraries, but would also serve the needs ofbusinesses by providing massive capacitystorage for documents, records, CAD drawings,and so on. Especially with the recent revision ofthe tax code that allows national tax records tobe held in digital form, the DVD library systemoffers an excellent way to keep corporate taxrecords beginning with the first of the year1999, and to maintain those records for longperiods of time at low cost. —Outlook

Accepted as an international standard, DVD-RAM offers a high degree of flexibility and willprovide even greater capacity at lower cost inthe years to come. Hitachi plans to offer anextensive lineup of DVD library systems, notonly to serve the needs of the corporatemarketplace, but also tailored to the needs ofthe home market for storing digital satellitebroadcasts, images captured with MPEGcameras, and other storage needs.

Japan’s tax laws were revised in July 1998to allow tax-related records created by acomputer to be maintained in digital form.

DVD Library System for Implementing a Massive StorageNetwork Warehouse at Low Cost

2

Masahito Sumitomo (left)and Masao Nakamura(right) of the Digital MediaSystems Division wereinstrumental in developingthe DVD library system forlong-term storage ofmassive data.

DVD Library System

Windows 95 and Windows NT are registered trademarks ofMicrosoft Corp. of the US.

Optical Wavelength-division Multiplexing TransmissionSystem Capable of 320-Gbit/s Transmission

3

Demand is increasing at an accelerating pacefor larger capacity trunk communication linesthat can accommodate faster throughputsbetween Internet terminals and that will supportvideo and other broadband multimedia services.Hitachi is a key player in this emergingworldwide market as exemplified by the popularacceptance of its 10-Gbit/s optical transmissionequipment that was rolled out in 1997. Yetdemand for increased capacity is relentless. It isprojected that demand will increase 20 foldrequiring a throughput capacity of 200 Gbit/s bythe turn of the century. This motivated Hitachi’sTelecommunications System Group in 1998 todevelop an optical wavelength multiplexingtransmission system capable of combining upto 32 separate wavelengths, each carrying 10Gbit/s. The device is now being aggressivelymarketed around the globe along with the 10-Gbit/s optical transmission equipment.

Combining up to 32 DifferentWavelengths, Each Carrying a10-Gbit/s Optical Signal —Principle of optical wavelength-division multiplexing

Hitachi is proud of its achievement indeveloping a rugged practical opticaltransmission equipment that supports thefastest throughput that is currently available, 10Gbit/s. Even though this is equivalent to120,000 voice channels through a single fiber, itis still far short of the capacity needed tosupport the needs of next-generationmultimedia communications. This is whyoptical wavelength division multiplexingtechnology is employed, for it allows multiplelightwave signals to be carried on separatechannels or wavelengths, over a single fiber atthe same time. To illustrate how this works,imagine that data is transmitted by turning asignal on and off using red light. Now, withoptical wavelength-division multiplexing,separate signals can be carried over the samefiber using green and yellow in addition to redto turn their respective signals on and off. —Features of Hitachi’s opticalwavelength-division multiplexingsystem

The system is configured using two types ofdevices: End Terminals (ETs) connected tomultiple 10-Gbit/s optical transmissionequipment that sorts and routes the lightwavesignals carried on separate wavelengths, andline amplifiers (LAs) that amplify and stabilizethe signals while they are in route between ETs.Recently, we have succeeded in increasing the

Channel (OSC) between an ET and an LA formonitoring. We are already equipping the 10-Gbit/s optical transmission equipment withHitachi‘s proprietary network monitoring controlsystem called Network Manager (NM), that isfully compliant with the latest TMN(Telecommunication Management Network)international standard. But by implementingboth the OSC and NM in combination, this givesHitachi the ability to support throughputupgrading of trunk transmission links end toend. —Future development

We now have an Optical Add/DropMultiplexing System under development thatwill permit signals at the wavelength level to besplit off or combined at intermediate pointsalong trunk networks. By integrating this systemwith the optical wavelength-divisionmultiplexing transmission system, we canrapidly deploy lightwave communicationsnetworks that are flexibly tailored to the needs ofcustomers in much the same way that smallerarterials branch off from highways and highwaysare interconnected with other highways.

number of wavelengths from 16 to 32, for a totalcapacity of 320 Gbit/s over a single filament ofoptical fiber, by developing a very stable laserlight source that effectively suppressesfluctuations in laser wavelength and a moreprecise demultiplexer.

Over linear or ring topologies alike, thesystem permits users to increase the number ofoptical wavelengths carried in accordance withtheir needs. This provides a way of flexiblyincreasing throughput capacity at modest costwithout altering the fiber-optic lines that arealready in place. The Line Amplifiers are idealfor implementing direct city-to-cityinfrastructure facilities, because they supportrelatively long-hop relay intervals up to 400 kmthanks to their noise reduction and high level ofoutput.

Development of an End-to-EndNetwork Management System

—System reliability and maintenanceA network management system can be easily

implemented in the optical wavelength-divisionmultiplexing transmission system by installingan optical interface called Optical Supervisory

Hiroyuki Nakano (right) and Kazutaka Sakai (left) ofthe Telecommunications System Group wereinstrumental in developing the optical wavelength-division multiplexing transmission system.

Schematic of the 10-Gbit/s X 32 DenseWavelength DivisionMultiplexing System

Etching equipment to cope with shrinkingdesign rules as VLSI technology evolves towardsmaller features and new levels of density mustnot only provide superior processingperformance, but must also yield cost merits interms of return on investment. In addition, thereare many obstacles that must be overcome asthe industry shifts over beginning in 1999 tolarger diameter 300-mm wafer fab lines:automating wafer handling, reducing thefootprint of major equipment, and loweringcosts. Hitachi has considerable expertise indeveloping plasma etching equipment that iscapable of processing deep submicrometerrange features at lower operating pressures.Anticipating the need for even smaller designrules of 0.18-µm and faster throughput, Hitachideveloped the Model M-700 Series microwaveplasma etching system that already complieswith next-generation line standards before it ismandated.

Reconciling Larger DiameterWafers with 0.18-µm DesignRules —Technologies to accommodate largerwafers

While most semiconductor devices todayare fabricated on 200-mm wafers, this is nowchanging. In line with the need for new levels ofdensity and greater efficiency, we are nowembarking on a new era of more stringent 0.18-µm design rule processing on larger 300-mmwafers. To meet these requirements, the M-700Series etching system adopts a multi chamberapproach with two chambers each for etchingand for ashing, and an enhanced version of themicrowave plasma source used on the M-600Series system for processing 200-mm wafers.By optimizing the design of the microwaveplasma source for 300-mm wafers, the plasmauniformity reached the level of 6%.

It was a daunting task to reconcile largerdiameter wafers with 0.18-µm features. We met

the challenge by developing an Advanced-ECR(Electron Cyclotron Resonance) plasma sourcethat effectively controls the overall plasmauniformity, and by further refining the robustprocessing technology of the current M-600Series system that is already capable of massproducing devices with 0.18-µm scale features.

Compact footprint andCompliance with SEMI High Throughput

We applied a double-arm high-speedvacuum transfer robot that effectively doublesthe arm extension speed of the previoussystem, and this cut the time in half from 20seconds to 10 seconds to remove a finishedwafer and mount a new wafer for processing.The processing speed has also been reduced byintroducing a new type of electrostatic chuckthat only requires one plasma discharge. Thecurrent production system requires two plasma

discharges: one when the wafer is attached tothe chuck and the other when the piece isremoved from the chuck. In addition, a newsystem has been adopted for coupling high-frequency power to the ashing chamber, theprocessing speed has been increased three-fold, 10% uniformity is achieved, and variousother design enhancements that improvethroughput have been implemented. —Application to next-generation lines

The space taken up by major equipment incleanrooms-which are exorbitantly expensive tomaintain-is reflected in production costs. TheM-700 Series system was implemented verycompactly by optimizing the layout, so thesystem’s footprint is only 10% larger than theM-600 Series system. We also incorporated anumber of functions and maintenanceenhancements that will be required on next-generation lines in compliance with the US-based SEMI (Semiconductor Equipment andMaterials Institute), the de facto internationalstandard. These include an air-tight cassetteinterface, low particle levels, and long-waitcycles. —Future development

The quest for improved processperformance and smaller features sizes, as wellas the R&D to sustain these developments isnever ended. As the market for larger diameter300-mm wafers heats up, we plan to furtherupgrade and refine the performance andreliability of the M-700 Series microwaveplasma etching system.

M-700 Series Microwave Plasma Etching System forHandling Large-diameter 300-mm Wafers

4

Working out of the KasadoAdminisrative Division, KenYoshioka (left) and Yoshiaki Satou(right) were the primaries indeveloping the M-700 Seriesmicrowave plasma etching system.

View of the M-700 Series Microwave Plasma Etching System

The World’s Largest Helical Fusion Experimental Device,Completed by the Ministry of Education’s NIFS

5

Nuclear fusion research aims to reproduce onearth the hydrogen burning process that servesas the energy source of our sun and other fixedstars, i.e. to establish “a second sun on theearth’s surface”. Research on this potentiallyideal energy source for humanity is advancingin countries throughout the world. As one of theexperimental devices developed in the courseof this research, this newly-completed LargeHelical Device (LHD) is being hailed as afurther step toward achieving controlled nuclearfusion.

Putting Japan’s Unique HelicalDevice to Use—What is NIFS?

The National Institute for Fusion Science(NIFS), an organization established in Toki City,Gifu Prefecture, Japan, in May 1989. Scientistsfrom a number of universities around thecountry jointly participate in it, with the aim ofconducting research on the nuclear fusionplasma principle and its applications. NagoyaUniversity, Kyoto University and HiroshimaUniversity are among the participants. It tookeight years to build LHD, the core researchequipment there. Plasma was first ignited therein March 1998, and it is now being put to verygood use.—How does LHD differ from othertypes of nuclear fusion experimentaldevices?

Many plasma confinement methods arebeing studied. Like the tokamak device, themethod used in LHD is to confine plasma in adoughnut-shaped magnetic field. But unlike thetokamak device, it generates the magnetic fieldwith helical coils and that’s where the namecomes from. Two coils wrap helically aroundthe outside of a vacuum chamber, and aconfining magnetic field called a heliotronmagnetic field is formed by running currentthrough the coils in the same direction. Thismethod, developed at Kyoto University, isunique to Japan.

different things with NIFS during the research,development, and production steps. We paid alot of attention to 3D design and productiontechnologies, how to get high precision,accuracy and reliability, high-volumesuperconducting technologies and so on, and alot of effort was expended on technologies thatnone of us has had any experience with before.Then, too, the scale of the device was about twodegrees of magnitude bigger than anythingpreceding it, so inevitably it became an on-siteconstruction project. At the height of theproduction period there were over 200employees working on the project, and so wehad to pay a lot of attention to personnelmanagement, too.—Hitachi’s involvement in the nuclearfusion field

We are the only manufacturer in the countrythat has a nuclear fusion experimental device,and we have participated in the design,development and construction of such devicesfor 40 years. Of course we are very proud ofthis, but to really make the dream of controllednuclear fusion a reality, we will have to not onlyapply the technologies we have developed todate, but also strive to achieve an even higherlevel of technology. We will have to make evenmore of an effort in the years to come. Whilereceiving assistance and cooperation from othernuclear fusion research organizations, we willhave to make even more of an effort in the yearsto come.

—Advantages of using a helical deviceWith the tokamak device, a strong driving

current is running through the plasma. This hasto do with the magnetic field configuration andbecause of it, the plasma can only be preservedfor a very short period of time. With the helicaldevice, however, the plasma rotates as if it wereentwined in the line of magnetic force, and thusthere is no need for a driving current and youget continuous stable operation as a result.

40 Years of History, StillStriving for Higher Technology—What things did you have to payparticular attention to or put a lot ofeffort into during the construction ofLHD?

In a big project like this one, it’s prettytough to follow the original developmentschedule that was laid out. A lot of changes hadto be made in the course of the project,changes in development factors and so on, andso we were constantly discussing many

Three persons who have beenworking on the developmentof LHD for the past eightyears. From left, YoshikatsuYasue of the Nuclear Fusion &Accelerator Project Division,Katsuhiko Asano of NuclearSystems Division, and SeiInomata of the Chubu AreaOperation.

The NIFS Large Helical Device, with whichplasma ignition was first successfully achievedon March 31, 1998. Outer diameter: 13.5m;weight: approx. 1,500 tons; plasma diameter:8m; magnetic field strength: 3T.

Interior of Vacuum Chamber That Confines theHigh Temperature (as high as 100 million°C)Plasma. Two superconducting helical coils wraparound the vacuum chamber, giving the chamberthe complicated appearance seen in thephotograph.

The renewal project of Central Load DispatchingCenter at the Kansai Electric Power Co., Inc. iscompleted. The central dispatching center is themainstay of controlling all the generators basedon the results of load forecasting. We havedeveloped various functions for this systemcombining the best technology of distributedsystems and optimization methods. And itcontributes to realize the advanced operatingenvironment such as reduction of fuel cost forgeneration, maintaining security control indexand supply control index. This system is expectedto be the new standard of central dispatchingcenter for the 21st century.

Advanced ELD (Economic LoadDispatching) SystemConsidering Dynamic RelationBetween the Fuel Cost and theGeneration— When did the project begin?

Every ten or fifteen year that the central loaddispatching center is renewed. This is a very bigproject. We had set about the research anddevelopment for the project since 1989. First, werenewed the basic function of monitoring powernetwork such as power flow of transmission linesand output of generators, and then, we developedand installed full system from 1995 to 1997. Itwas such a big and complicated system with thehigh expectation of customer that we had made alot of joint studies with customer. In the processof the joint studies, we developed new algorithms(the procedure to solve the problem or the seriesof solution) and made effort to implement the newsystem in cooperation with the researchlaboratory and the works.— What is the characteristic point ofthe new system?

Most remarkable point of this system is theintroduction of distributed system andoptimization technique. First, in regard to thedistribution system, it was constructed fromindependent sub-systems that include severalcomputers for each function, and the customerscan add new functions as sub-systems accordingto their design preference at any time. Comparingwith the former systems that controls wholepower network with centralized large hostcomputers, new system provides the highlydependability for running whole systemcontinuously in case of emergency. Second, asfor the optimization technique, we give anexample of generating power at power plant. It isthe problem that should control many thermalgenerators for reducing fuel cost according to theload fluctuation. For this problem, we had been

applying traditional method considering onlystatic fuel cost curve to operate generatorsefficiently. But to meet the demands of the costreduction, we developed the technique ofadvanced ELD, which considers future loadfluctuations and can reduce the total fuel cost in aday. From the viewpoint of reducing further thethermal generation cost, we applied the techniqueof considering dynamical characteristics betweenfuel cost and generation power in this ELDfunction. This is the first system in the world thatapplied this advanced ELD technique.

Main Operation Control IsThermal and Hydro-pumpedGenerations— How the load should be forecasted?

The forecasting result can be revisedaccording to the forecasting target interval, yearly,monthly, weekly, daily forecasting. As the totaldemand of power depends on the weather andtemperature, first, we calculate the weekly roughresult of load forecasting, and then, calculatesuccessively the daily result of load forecasting indetail according to the exact data. If we forecast

daily load much higher than actual need, it willcost much more than usual, because the availablenumber of generators would be increased. On thecontrary, if we forecast daily load much less thanactual need, it will also cost much more thanusual, because the deficient power should besupplied from the generator or other powercompany with high cost. Therefore, correct loadforecasting is the most important work in thepower company.— What is the target of operationcontrol?

The output power of nuclear generators isalmost constant. It is called as base load. Targetof ELD detailed above is thermal and hydro-pumped generators. As for the thermalgenerators, it is important to decide whichgenerators should be stopped in the night, and inregard to the hydro-pumped generators, it isimportant to decide how much water should bepumped up in the night.— What is the future trend?

With the progress of time, more and moreinformation-oriented or automation-orientedsystem will be developed in the future.

Renewal of the Central Load Dispatching Center atthe Kansai Electric Power Co., Inc. for the 21st Century

6

Mr. Shigeru Tamura (left) of theOmika Administrative Division,and Mr. Chihiro Fukui (right) ofthe Hitachi Research Laboratorytook part in the renewal projectof Central Load DispatchingCenter at the Kansai ElectricPower Co., Inc.

Monitoring and Instruction Room at the Central Load Dispatching Center which realizethe human friendly interface for operators with large display. Information on themonitoring board is transmitted from the computers made by Hitachi.

An Environmental Management Support System to HelpUsers Gain ISO 14001 Certification

7

Environmental management is becomingincreasingly vital with the ongoing spread ofinternationalization and broadening concern forthe earth’s environment. In this climate, acquiringISO 14001 (the internationally recognizedenvironmental management standard)certification has been called the “passport” forfirms supporting the era of internationalization.The Environmental Management Support Systemfor ISO 14001 certification, developed byHitachi’s Omika Administrative Division, is a toolthat meets the needs of such firms. It has earnedhigh praise from users and affiliates of users. Forits development Omika Administrative Division inMay 1998 received the 25th Environmental Awardsupported by the Environmental Agency, jointlysponsored by the Environment InvestigationCenter and The Nikkan Kogyo Shimbun, Ltd.

Development PlannedIncorporating Advanced Know-how—What is ISO 14001 and what is itsobjective?

It’s an international common standard to helpcompanies and organizations do what they can tosolve the problems involved in preserving theearth’s environment. Taking environmentalprotection measures is one of the majormanagement issues in companies, and thisstandard aims to help them systemize andimprove the way measures are taken.—When was the standard established?

The ISO (International StandardsOrganization) standardized it in September 1996,and the month after that JIS (Japan IndustrialStandard) 14001 was determined. Actually, therewas an almost identical standard already inexistence in the U.K. at that time, British Standard7750, and some firms had already been certifiedunder that standard. Our Data Storage & RetrievalSystems Division was the third establishment inJapan to gain BS7750 certification.—What was the motive in developingthe Support System

Our Omika Administrative Division wassimultaneously awarded ISO and JIS certificationin October 1996, the eighth establishment withinHitachi to be so awarded. But it was pretty hardgoing for the people in charge of obtaining thecertification since all of the work was paper-based. This inspired us to convert know-howaccumulated up to then into digital data, not onlyto lighten our own load but also to create aproduct for use by others outside the company.

HOKI Publishing Co., Ltd., with their specializedknowledge of environment-related legislation.—What advantages does the SupportSystem provide to users?

It provides them with impact assessment,regulation checking, document handling, training& education management, and measurementrecording functions, and allows them, via theInternet, to get monthly updates of data onenvironment-related laws and regulations. TheSupport System is not merely a software package;users can freely use our know-how to make iteasier to make preparations for gaining ISO14001 certification. Some users have been able toget certification in only four months. Preparationsfor inspections after the second year can also bemade smoothly since the process is not a paper-based one. Finally, we also conduct seminars toassist those who want to be certified but do nothave much experience in PC usage or Internetaccessing.

Monthly Updates of RegulationData via the Internet—What is it that’s particularly hardabout getting certified?

Two things, mainly. First, related regulationsare modified or updated very frequently, and it’shard to keep up with the changes. Second, wedon’t know what method will be used to assessenvironmental impact in the certification process.Also, even if you do become certified, there is anannual check called ‘surveillance’ you have topass, and it is necessary to be re-certified afterthree years. So documents, records and so forthhave to be carefully managed. This means that itis essential to have specialized knowledge of lawsand regulations, the environmental effects ofchemicals and so on. In our case, we started thecertification process by compiling anenvironmental database, using informationobtained from Hitachi research labs. We also gota lot of help and cooperation from the CHUO

Co-coordinators of theEnvironmental ManagementSupport System developmentprogram, Yoshiaki Ichikawa (left)of the Omika AdministrativeDivision and Industrial SystemsDivision, and Seiten Takahashi(right) of the Power & IndustrialSystems R&D Laboratory.

Window of Access Environmental Impact

At the National Commendation for InventionsAwards Ceremony sponsored by the JapanInstitute of Invention and Innovation in June1998, the West Japan Railway Company (JRWest) and Hitachi, Ltd. received the prestigiousInternational Trade and Industry Minister’sInventions Award for the creation of its Series 500“Nozomi” Shinkansen, the super-sleek bullet trainthat links Tokyo with Hakata. At the sameceremony JR West and Hitachi, Ltd. were jointlypresented with the Meritorious Invention Awardfor their development of the train. Highly praisedfor its combination of design and technology, witha top speed of 300 km/h that ranks among theworld’s fastest, the train features a unique blend ofideas combined into a forum that goes beyondanything preceding it.

Main Development Objective:Minimize Hard Pressure Wavesand Sound in Tunnels—When and how did the idea for yourinvention come about?

In November 1991, our Design Center and theTransportation Systems Division jointlysponsored a ‘Transportation Vehicle Fair’ toshowcase Hitachi’s railway vehicle technology.This was a first-time experiment for the industry,and one of the models we had on display caughtthe eye of JR West, which was then looking toachieve higher speed in its trains. It was thatmodel that eventually led to the development ofthe 500 Series.—When did the development actuallybegin?

JR West began doing aerodynamic field testson the WIN350, a prototype train for the 500Series, in 1992. In these tests they measured andanalyzed factors such as hard pressure waves andsound in tunnels. We also participated in thetests. Actual development started in June of 1993

and we got an order for full-scale production inDecember 1994.—What do you mean by “hard pressurewaves in tunnels”?

Well, when a high-speed train goes into atunnel, the air in the tunnel compresses, andwhen the energy from some of the pressure wavesis forced out of the exit at the other end of thetunnel it makes a loud noise. Same principle as apopgun. Now, there are 142 tunnels on the SanyoShinkansen track, and they cover almost half thelength of the entire track. So the main point indeveloping the new cars is to reduce the noise asmuch as possible by further streamlining of thehead of the lead car and the cross sections of allthe cars.—So what sort of car shape came outafter the development stage?

We conducted tests jointly with the RailwayTechnical Research Institute, and on the basis ofthe results obtained held discussions with users,planners and designers to determine the kind of

shape we would use. In the end we decided to gowith a 15-meter nose length for the front car and acylindrical cross section for the cars. To ensureboth lightness and rigidity we decided on analuminum honeycomb structure, and to reduce airresistance we went for curved-glass windows andplug doors (doors whose surfaces are flush withthe surfaces bordering them). And we tried toachieve a uniquely beautiful shape so that thecars would have long-term appeal.—What results did you get incombining design and technology?

Even at 300 kilometers per hour, we were ableto achieve a lower track noise level and fewer hardpressure waves in tunnels than had been achievedpreviously at 270 kilos per hour with the 300Series introduced in 1992. Compared with the300 Series, we reduced running resistance by 30percent and energy consumption by 15 percent.Finally, we were able to reduce the vibrationgenerated when trains pass each other or runthrough a tunnel, which makes the ride morecomfortable.—How do you evaluate the speedresults you achieved?

First, the 300 kilometers per hour reached onthe Sanyo Line is the highest ever for a train.Also—and not many know this—the train set twoother records that made the 1997 Guiness Book,an average speed of 242.5 kilos per hour fromstarting station to terminal station, and an averagespeed from one station to the next of 261.8 kilosper hour. So the speed records of the 500 Seriesare there for all the world to see.

Design of the West Japan Railway Company’s Series 500Shinkansen Train—Winner of the 1998 JIII’s Inventions Award—

8

Recipients of NationalCommendation for Invention (ThePrize of the Minister of Ministry ofInternational Trade and Industry)for their creation. From left,Morishige Hattori of theTransportation Systems Division,and Hirofumi Tanaka of the DesignCenter.

Series 500Shinkansen Train

A High-throughput, Nanometer-level Ultrathin-filmEvaluator

9

In the field of research on next-generation devicesand new device materials, and in the devicefabrication process, there is a growing need foraccurate controlling of the multilayer structurecharacteristics of nanometer-level ultrathin films.To meet this need, Hitachi has developed the newHD-2000 Ultrathin-film Evaluator, which employsa new electron optics system to observe, analyze,and measure electron-level structures with highresolution and high contrast.

A New System Combining theAdvantages of SEM and TEM—Objective in developing the HD-2000evaluator

Year by year semiconductor-based electronicdevices are becoming both thinner and moremultilayered. So it has become important toachieve film structure and composition where theelectron level is on the order of a few nanometers(a nanometer is one-millionth of a millimeter). Itis essential to obtain high-resolution images ofthese thin films so that they can be observed andevaluated. This will be necessary for thedevelopment of new semiconductor materials andimprove the yield of the semiconductor devicefabrication process. But it was becoming difficultto do this with conventional electronmicroscopes.

For example, users can operate a SEM(scanning electron microscope) while watchingthe monitor even in a brightly lit room. But themaximum resolution of SEMs is only 1.5nanometers. With a TEM (transmission electronmicroscope), on the other hand, resolution is ashigh as 0.2 nanometers, but TEMs are quite bulkyand considerable expertise is required to operatethem. Furthermore, obtaining results with them isa time-consuming process because images haveto be printed out on paper.

The objective, then, was to develop a tool thatcombined the advantages of SEMs and TEMs,one that would enable users to observe very thinfilms with high resolution even in brightly-litspaces, produce results quickly, and do highlysensitive analysis of the films. This will be arequirement for next-generation devices. This wasthe rationale behind the development of thisunique Hitachi concept, the HD-2000 Ultrathin-film Evaluator.—Main features of the HD-2000

A newly designed objective lens and X-raydetecting element, which raises analysissensitivity to about 2.5 times that of aconventional TEM. This makes nanometer-orderelement analysis possible. Moreover, alloperations—from setting test object samples to

gate technology could only be applied in limitedcases. We then developed a FIB (focused ionbeam) device which could automatically cutsamples with ions and make samples in abouttwo hours, and succeeded in linking it to HD-2000 with the use of a joint folder. After gettingclear images of a sample’s inner structure with theHD-2000, we used highly accelerated voltagesecondary electron images to obtain the regionsto be analyzed with 0.1-µm accuracy, then againused FIB to make high-precision thin films. In thisway we were able to fabricate samples more easilyand more quickly, in fact with processing speedmore than 20 times greater than that of theconventional method.—What’s the next step in your work?

The X-ray mapping of HD-2000 makes itpossible to clearly observe nanometer-levelelement distribution. Here, too, it providesheretofore unattainable precision in analysis. Butwe would like to go further, to achieve even higherresolution so that element distribution could beshown as more realistic images.

observing, analyzing and measuring them—canbe done in brightly-lit spaces with high resolutionof 0.2 nanometers. And images can be easilyobserved on the GUI (graphical user interface) ofa Windows 95 operating system. Of course, aswith TEMs no film is necessary. And since allimages can be saved as digital data, analysisresults can be quickly processed and sent tofactories and research laboratories linked by acomputer network.

Fast, Pinpoint SampleProduction by Linking HD-2000With FIB—I hear you were also able to achieveimproved throughput in sampleproduction.

In making samples, one must be able toaccurately cut parts to be observed, and grindthem down to micrometer-order thin films. In thepast one had to rely on highly skilled techniciansto do this work, and even so it might take as longas 25 hours to complete the job. And pinpoint

Shigeto Isakozawa (left) andOsamu Tanaka (right), bothmembers of theInstruments, are co-developers of the HD-2000Ultrathin-film Evaluator.

Ultrathin-film Evaluator“HD-2000”

In October 1998, Hitachi unveiled its “PAMRefrigerator” series, nine refrigerator modelsemploying PAM (Pulse-Amplitude Modulation)control, the first-ever refrigerators with thisfeature. Following on the heels of the world-standard “Vegetable-centered Refrigerators”(vegetable compartment in the middle) introducedin January 1996, they represent a wholly newconcept in home refrigerators. PAM controlsimultaneously enables both energy saving andhigh power, earning high praise from industryanalysts. In February1999 Hitachi received theInternational Trade and Industry Minister’s EnergyConservation Award from the EnergyConservation Center for the development of theseries.

PAM Control Provides BothEnergy Saving and High Power—What is PAM control?

‘PAM’ is an acronym for Pulse-AmplitudeModulation. PAM control is the technique ofcontrolling the compressor’s DC motor tomaintain operational efficiency throughout theentire revolutions area, from low speed to highspeed. This is the first time this technique hasbeen applied to refrigerators. It was necessary todevelop a number of new technologies in order todo this, and so we ended up applying for 18separate patents during the development period.The idea behind developing PAM refrigeratorswas to combine energy conservation with therefrigerator’s ability to preserve the texture andflavor of the foodstuffs it stores. The PAMrefrigerator’s “fast-freezing” function, whichreduces the freezing time of conventionalrefrigerators by some 40 percent, also helps topreserve food flavor.—How were you able to keep therefrigerator’s noise level down?

A PAM refrigerator’s compressor runs atbetween 2,000 and 4,300 revolutions per minute,and we tried to obtain energy-saving operation atthe lowest possible revolutions area. Furthermore,mechanical resonance is avoided by using a six-

step compressor, whose revolutions mode varieswith the steps. In this way we can keep noise to aminimum.

Simple Operation: Two Buttonson the Door—How do PAM refrigerators differfrom conventional inverterrefrigerators?

With inverter control, the input voltage to thecompressor motor is fixed at 250 volts, but withPAM control it is 170 volts for low motor speedand 280 volts at high motor speed. Because ofthis, the revolutions area is about double that fora conventional refrigerator, and high-efficiencyoperation can be obtained throughout the entirerevolutions area. Also, with PAM refrigerators it

is easy to switch between energy-savingoperation (with the lowest power consumption inthe industry) and high-power operation (with theshortest freezing time in the industry, 25minutes); all one has to do is push the twobuttons on the door.—What are the main features of the R-S50PAM, the top model in the series?

It’s inner capacity is 495L even though it hasthe same external dimensions as a conventional450L refrigerator. It consumes less electricalpower than any refrigerator in its class. Also,besides the central vegetable compartment andfreezer, it has a “variable temperature freezer” inwhich the temperature can be switched from sub-zero chilled (about -1°C) to full freeze (about -18°C). Finally, it has a separate icebox that makesvirtually odor-free ice in only 55 minutes, thefastest in the industry.—Are PAM refrigerators what youwould call ‘environmentally friendly’?

Yes. In inverter-controlled refrigerators, thepower factor (a measure of how effectively electricpower is used) is only about 75%. We haveimproved that to 95%, which greatly reduces theload on the unit’s electrical equipment. The R-S50PAM also uses a more environmentally-friendly packing material than the usual vinylchloride.

“PAM Refrigerator” Series: The World’s First Refrigerators with PAM Control Mode

10

Kazuhisa Ichimoto (left) andAkinobu Takemoto (right), bothmembers of the Refrigeration & AirConditioning Division, are co-developers of the PAM Refrigerator.

PAM Refrigerator “R-S50PAM”

A 128-Mbit Single-electron Memory: The First-ever Super-highly Integrated Memory Operable at Room Temperature

11

A single-electron memory is one in which themovement of individual electrons can becontrolled, thus allowing one bit of information tobe stored by only a few electrons. At thebeginning of the 1990’s, the possibility ofachieving such a memory was but a distantdream, but by 1993, Hitachi had achieved thefirst-ever single-electron memory that could beoperated at room temperature. In 1995 itscapacity was increased to 64 bits, and 1998 sawthe first-ever development of a 128-megabitsingle-electron memory suitable for application intoday’s memory semiconductor devices.

Overcoming MemorySemiconductor DeviceLimitations—Could you explain the background tothe development of single-electronmemory devices?

Today’s semiconductor memory devices, suchas DRAMs (dynamic random access memory)and flash memories, are basically achieved bycombining a capacitor and a switch. Here, thecapacitor stores and discharges some 100,000electrons at a time to record ‘o’ or ‘1’ binary data.Devices using this system, however, cannotincrease memory capacity without limitation, anddevices cannot be made smaller than a certainfixed size. Thinking, therefore, how great it wouldbe to have memory that could record with only afew electrons, we came up with the idea of single-electron memory.

Actually, the first proposals for single-electronmemory devices were made more than 10 yearsago. But these devices could only operate atextremely low temperature of -273°C. It wasHitachi that in 1993 first developed a device thatcould operate at room temperature. In 1996 an IC(integrated circuit) level device was developed.And then we succeeded in making a big jump in1998 to the LSI (large-scale integration) levelwith the 128 megabit device. The capacity of thisdevice is two million times greater than the onethat preceded it.

Amassing Electrons in NaturallyFormed Minute Pores—What are the technologies withwhich the impossible can be madepossible?

By making memory scale on the order of 10nanometers. This is an extremely small scalesince it takes 100,000 nanometers to make onemillimeter. But since the minimum scale with

temperature. Then we improved the procedure intwo ways by adding 0.25-µm processingtechnology to it. First, we obtained perpendicularcurrent flow instead of the conventional parallelflow in the memory elements on the wafer surface.Second, we obtained a two-level, top-and-bottomstructure. Finally, by applying CMOS(complementary metal-oxide semiconductor)technology for reading and writing, we couldconstruct circuits that automatically controlledwriting time in response to memory elementvariability. It was these technologies that enabledthe development of the 128 megabit device.—Advantages of single-electronmemory devices

First, because of their simple structure, theyare only one-fourth the size of DRAMs withequivalent capacity. Second, if we can getmemory structure down to 10 nanometers, weshould be able to get memory capacity as high asone trillion bits, instead of the 10 billion withcurrent semiconductor technology. This would bethe same as getting all the digital data from 20full-length feature films onto the tip of your finger.—What will be the next step in yourwork?

To make an actual product out of the device,there are a number of choices we will have tomake regarding the chip specifications. We willhave to decide whether to aim at high accessspeed, or to put emphasis on production cost, orto maximize memory capacity. At the same timewe will have to achieve high reliability in thedevice and improve it so that it can be suitablymass-produced. We are aiming at putting thedevice to practical use by the year 2005.

available processing technology was only 100nanometers, we then thought of the possibility ofapplying a natural structure. We fabricated a verythin silicon film, and used the minute pores thatformed naturally in it as memory electrodes. Bychance the first film formed had a thickness of 3nanometers, and by good fortune the pores thatformed at that thickness turned out to be mostsuitable for amassing and controlling themovement of electrons at room temperature. Andthrough our unique process for controlling filmthickness we achieved operability at room

Kazuo Yano (left) and Tomoyuki Ishii (right), bothresearchers at the Central Research Laboratory,are co-developers of the 128-megabit single-electron memory.

A 128-Mbit Single-electron Memory