Product Watch: EMC Teams Up With HP OpenView. AMD Releases x86-64 Simulator. MathSoft Introduces S-PLUS 6.

October 13, 2000

SHORT TAKES

EMC JOINS FORCES WITH HP OPENVIEW

Hopkinton, MASS. — EMC Corporation, a world leader in information storage, announced immediate availability of the EMC Fastrax data movement platform. An integrated member of EMC’s storage infrastructure, EMC E-Infostructure, EMC Fastrax performs the resource-intensive movement of backup and recovery data between EMC Symmetrix Enterprise Storage systems and industry standard tape libraries. The new Fastrax platform combines EMC and third-party software, open application programming interfaces (APIs) and data movement hardware. Fastrax enables customers using the HP OpenView OmniBack II backup/recovery application to leverage their existing investments in backup software, human resources and operating procedures.

Customers are facing technology choices to address the elimination of the backup window and the need for rapid data recovery. Disk-based solutions such as EMC TimeFinder and Symmetrix Remote Data Facility (SRDF) enable non-disruptive backups along with virtually instant data recovery. Traditional tape backup implementations facilitate non-disruptive backups, however, lack the robust capabilities for rapid data recovery. Fastrax fills the void by delivering the non-disruptive backup features of disk and tape solutions while providing faster data recovery times when compared with traditional tape backup offerings. EMC customers are able to choose between disk-based solutions, tape-based solutions or a combination of the two depending on their data recovery objectives. Customers can implement TimeFinder, SRDF, EMC Data Manager (EDM(TM)), Fastrax or integrated combinations of the above to meet their needs for instant recovery of data and to backup and restore less-critical information at high speeds.

The Fastrax data movement platform builds on the openness of the EMC offering and enhances customers’ ability to move information between EMC Symmetrix and non-EMC devices. Through cooperative development between EMC and HP OpenView Storage Management, Fastrax supports backup and recovery of database information via one of the industry’s premier backup and recovery applications and the first to support Fastrax, HP OpenView OmniBack II. Through its open E-Infostructure Developers Program, EMC over time will extend Fastrax support to additional data movement applications such as content distribution and data migration, and provide support for non-tape devices. Fastrax enables customers to perform tape-based back up of even the largest Oracle databases with no impact to host servers, applications or data access, a major source of conflict and complexity in backup operations. Fastrax backups take place while the same data is being used by production applications. The backup and recovery application determines what data needs to be backed up, then Fastrax and Symmetrix software transparently determine which data needs to move to tape and which needs to go to the active database. Additionally, multiple backups can be started and run concurrently. Information about HP OpenView and its solutions can be found on the World Wide Web at http://www.openview.hp.com and http://openview.hp.com/products/omniback// . Information about EMC’s products and services can be found at http://www.EMC.com .

PLUMTREE PORTAL TAPS PARALLEL PROCESSING

San Diego, CALIF. — Plumtree Portal 4.0 is finding a new use for an old concept, using parallel processing to solve big, compute-intensive problems more quickly. When the users multiply and their requests for different resources build up, the Plumtree core system unleashes the Massively Parallel Portal Engine to draw application services and content off of different servers and assemble them in response to user requests. Instead of doing so in a series of steps, however, it seeks the responses in parallel all at once. For example, if a portal page needs a list of e-mail messages that requires one second for an e-mail server to prepare, an inventory report from another server that takes a second to prepare and a one-second news item from another server, the Parallel Portal Engine can assemble all three in one second, not three, said Glenn Kelman, vice president of product management and marketing.

It does so by interacting with the servers through what Plumtree terms “gadgets,” or distributed portal components that embed key information and services from applications and other portal sites. The Parallel Engine communicates with the gadgets via HyperText Transfer Protocol and embeds gadget commands in the HTTP headers or coding at the top of the message. By allowing distributed computers to process the gadget commands, the Parallel Engine builds a page in response to a user request more quickly than could a server executing a series of commands, Kelman said. Handling portal processing loads in parallel “will become a requirement for large organizations” predicted Thomas Koulopoulos, president of the Delphi Group. The Plumtree Portal 4.0 was co-developed with Procter & Gamble, which executed parts of the design and testing of the new system. Procter & Gamble tested the system by forcing it to respond to 100,000 users from “dozens of conflicting security domains,” which would have tested the portal engine’s ability to process security clearances fast enough to handle a high traffic load, said P&G portal manager Dan Gerbus. The 4.0 version of the system is at the center of the company’s portal site, he said.

AMD RELEASES X86-64 TECHNOLOGY SIMULATOR

Frankfurt, GERMANY — AMD publicly released the x86-64 technology simulator, also referred to as the AMD SimNow! simulator, an application specifically designed to give BIOS vendors, tools developers, operating system manufacturers and application providers the ability to evaluate their x86-64 technology based software code prior to the release of AMD’s forthcoming 64-bit capable processors implementing x86-64 technology. These processors, codenamed “Hammer,” are planned to be announced at the end of 2001. The AMD SimNow! simulator is available free of charge at http://www.x86-64.org .

AMD’s x86-64 technology is designed to enable platform suppliers, developers, and corporations to transition to 64-bit computing environments while continuing to have leading performance on the vast installed base of existing 32-bit applications. Memory hungry applications such as large databases, CAD tools, and simulation engines that are currently limited by the 4GB addressing limitation benefit most from 64-bit computing. AMD is enhancing the x86 architecture to include a 64-bit mode that has both a 64-bit address space and a 64-bit data space. AMD’s 64-bit processors are planned to be designed to detect which mode is needed (32- or 64-bit) and compute accordingly. “SuSE believes that AMD has taken a revolutionary approach by involving the open source community so early in the design process of its 64-bit technology,” said Markus Rex, vice president of development, SuSE. “By enabling access to the simulator, all Linux developers can help ensure a fast, high-quality port of AMD’s x86-64 technology.” “I’m very pleased that AMD asked CodeSourcery to port portions of the AMD SimNow! simulator to GNU/Linux. This simulator is designed to make it possible for kernel developers and compiler engineers to start porting GNU/Linux to x86-64 without waiting for prototype hardware. This port is a clear sign of AMD’s commitment to GNU/Linux, and that’s really exciting,” said Mark Mitchell, chief technical officer of CodeSourcery, LLC. “The AMD SimNow! simulator release is an important follow-on to the earlier release of the x86-64 architecture specification; AMD’s open approach to their own hardware nicely parallels the openness of the GNU/Linux software model.”

STOREAGE SHOWS HOW SAN STORAGE CAN BE CONTROLLED

Scottsdale, ARIZ. — StoreAge Networking Technologies, a leader in the key field of Storage Area Network (SAN) Virtualization, announced its SAN Volume Manager (SVM) appliance will be featured in SanOne’s upcoming FCTC HOT SAN Demo Interoperability Lab at the FCTC Conference October 10-12 at the Park Center Plaza (Crowne Plaza) in San Jose, California. “We are pleased to be part of the SAN HOT Demo at FCTC, hosted by SanOne,” said Dan Schoffelman, StoreAge’s Director of US Operations. “Potential customers and partners will be able to see how easy it is use our SVM appliance to organize SAN storage into stripe sets, storage pools, and volumes, which can be assigned to servers and mission-critical applications based on business needs, and then grow as their needs grow. Using our storage virtualization technology makes it really possible to maximize the use of expensive storage resources. Seeing SVM in the HOT SAN Demo allows everyone to see our product in a setting which closely matches a customer production environment.”

“We selected StoreAge as it illustrates the dramatic enhancements now available in storage virtualization,” said Mike Flannery, general manager of SanOne, an Alanco Technologies Inc. Company. “Virtualization is a major component of the HOT SAN Demo as it offers powerful answers to storage migration, growth and control issues. The interoperability lab shows that this technology is now viable in a real-life multi-vendor environment.” “The HOT Demo is a `not to miss’ event for anyone challenged with the task of managing storage and considering SAN equipment, said conference chairman Dr. Lance Leventhal. “They can see what SANs can do, find out how much it costs, and what return they can expect on their investment.” SanOne’s SAN FCTC Interoperability Lab provides hands-on experience with the latest multi-vendor SAN solutions in storage consolidation, LAN-free and serverless backup, storage virtualization and enterprise storage resource management (SRM). The SVM is a SAN Appliance that provides virtual volume management in a heterogeneous environment with servers and storage from multiple vendors. The SVM features physical device monitoring, volume definitions, and virtual storage allocations for all servers and storage connected to the Fibre Channel loop or fabric. For further information, visit http://www.sanone.com .

MATHSOFT INTRODUCES S-PLUS 6 FOR UNIX

Seattle, WA. — MathSoft, Inc., Data Analysis Division, a leading provider of solutions for statistical data mining and predictive modeling technology, introduced S-PLUS 6 for UNIX, a major enhancement to the company’s statistical data mining software. S-PLUS 6 provides a popular software platform for business intelligence and decision support applications that leverage state-of-the-art statistical analysis and visualization methods. The release features a powerful new Java-based point-and-click user interface to simplify and accelerate access, analysis, and visualization of technical and business data. A new integration method, called CONNECT/Java, allows software developers to enhance the analytical and graphics capabilities within their applications by embedding the S-PLUS engine. S-PLUS 6 offers innovative visualization capabilities including a new form of Java-based graphic, called Graphlets(tm), which makes it easy to deploy interactive, drill-down graphics via Web pages. S-PLUS includes a suite of flexible and sophisticated data display techniques to transform quantitative information into easy to understand charts, using multiple tabbed pages. A new Java-based graphics device allows users to export graphs in file formats such as JPEG, TIFF, PNG, PNM, and Windows BMP formats, in addition to live Graphlets.

“S-PLUS 6’s object-oriented programming model is critical for building statistical tools for complex data analysis. In particular, its new interface to Java provides additional tools for integrating S-based analyses into a distributed environment where we can build Web-based reporting applications, access and analyze relational databases, and build other analytical systems,” said David James, Statistics Research, Lucent Technologies. “Java connectivity combined with powerful S-PLUS analytics accessible from our new user interface will allow our customers and our consulting group to build modern business intelligence systems that help users gain insight from their data,” said Shawn Javid, senior vice president and general manager, MathSoft Inc. “S-PLUS 6 offers new methods for handling data including missing data and robust statistics delivering more accurate results so decision makers and analysts can make smarter decisions.” S-PLUS 6 for UNIX will initially be available for Linux and Solaris. Versions for other UNIX platforms will be made released over the coming months. More information about MathSoft’s Data Analysis Division and its products and services is available at http://www.splus.mathsoft.com .

QUALSTAR ANNOUNCES LTO TAPE LIBRARY SHIPMENTS

Canoga Park, CALIF. — Qualstar Corporation announced that production shipments of its TLS-8000 Ultrium LTO (Linear Tape Open) Tape Libraries have begun based on LTO Ultrium technology from Hewlett Packard, a founding member of the LTO consortium. The TLS-8000 LTO Tape Library Family includes seven models, offering capacities from 1.1 terabytes (native) to over 52 terabytes using data compression. Aggregate data transfer rates can exceed 860 gigabytes per hour. Most models can be configured with Qualstar’s SANSmart Fibre Channel technology. “HP is delighted to have worked in close co-operation with Qualstar from the beginning of its LTO Ultrium program,” says Adam Thew, General Manager, OEM business, HP Storage Organization. “Qualstar’s partnership with HP offers customers previously unmatched levels of performance, capacity, and reliability in Qualstar’s industry-leading SANSmart TLS-8000 Series Tape Library Family.”

LTO technology establishes a new open format specification for high capacity, high-performance tape storage products for midrange and network server environments, and addresses growing customer need for improved data interchange across heterogeneous platforms. Ultrium is a high-capacity format that offers up to 100 GB of native capacity per tape cartridge, yielding a substantial increase in storage density, especially when used in libraries such as the TLS-8000 Family. Founded in 1984, Qualstar designs and manufactures automated cartridge tape libraries and 9-track tape drives, with an installed base of over 35,000 units. Qualstar’s tape libraries span capacities from 200 gigabytes to over 30 terabytes of native (uncompressed) storage capacity, in several popular tape formats. The TLS-Series library families are compatible with most networks, operating systems, and storage management software applications. Qualstar’s products are sold worldwide, primarily to value added resellers and original equipment manufacturers. More information about Qualstar can be found on its web site at http://www.Qualstar.com .

SYNOPSYS EXPANDS ITS LINUX PRODUCT PORTFOLIO

Mountain View, CALIF. — Synopsys, Inc., the technology leader for complex integrated circuit (IC) design, today announced the availability of a full tool suite for high-level design on the popular Linux operating system(OS). In addition to the complete RTL synthesis solution with Synopsys’ flagship product Design Compiler, the extensive list of Linux products includes PrimeTime, Synopsys’ industry-leading gate-level static timing analysis tool for System-on-a-Chip (SoC), Scirocco high-performance VHDL Simulator, and Module Compiler. Increasingly complex designs require design teams to rely on large networks of high-performance workstations and servers. Low-cost, reliable, Intel-based computers, running Linux, provide an ideal design and development platform because of their ability to fit easily into an existing UNIX environment. The availability of Synopsys’ synthesis and verification solutions, combined with the improved price/performance ratio of hardware, provides designers of complex IC and systems-on-a-chip (SoCs) with a high-throughput, low-cost design and verification environment.

“We’re in a very fast moving industry requiring a quick response to market needs. This can be very expensive when adding traditional high performance compute capabilities on a regular basis,” said Guntram Wolski, director, VLSI at Creative Advanced Technology Center located in Scotts Valley, California. “The availability of Synopsys synthesis tools in addition to the already ported simulation and verification tools running on Linux will enable us to meet these additional requirements at a lower total cost.” “We make some of the world’s most demanding chip designs that drive design cycle time. Using powerful Synopsys products on Intel-based Linux workstations and servers provides a new level of productivity that helps our engineering teams shorten design cycle times and stay on budget,” said Paul Chermak, vice president of marketing manufacturing industries at SGI. “We are proud and excited to contribute our Linux and application performance expertise in working with Synopsys to help bring its industry-leading product portfolio to the entire Linux community.” Additional information about Synopsys is available at http://www.synopsys.com .

SAMS PUBLISHING UNLEASHES RED HAT LINUX 7

Indianapolis, IND. — Red Hat Linux 7 Unleashed (Sams Publishing, ISBN 0-672-31985-3) continues the success of Red Hat Linux Unleashed, which has sold nearly 250,000 copies to date. This book offers complete coverage of the new features and enhancements in the highly anticipated Red Hat Linux 7 upgrade. In August, Red Hat Linux 6 Unleashed was awarded the second place honor in Linux Magazine’s Editor’s Choice Awards, the Tuxies. Red Hat Linux 7 Unleashed uses the same clear, example-driven style of the award-winning title. This book, like its predecessor, addresses advanced-level users with comprehensive reference material and extensive problem-solving examples.

Teaching the Red Hat Linux operating system from the inside out, Red Hat Linux 7 Unleashed describes how to install, configure and manage version 7. The book contains discussions of SMTP and protocols, Network Information Service, Samba, connecting to the internet and configuring and building kernels. Additional topics covered include new security issues, new XFree86 4.0 distribution and conversion of audio CD tracks to .mp3 format. Red Hat Linux 7 Unleashed contains the complete 3 CD distribution of Red Hat Linux 7 and has been fully tested against the final release product from Red Hat. William Ball, full-time Linux evangelist and consultant based in the Washington DC area, is a technical writer, editor and journalist. Ball has authored many Sams Publishing titles including Red Hat Linux Unleashed, Third Edition, Sams Teach Yourself Linux in 24 Hours, How to Use Linux and Que’s Using Linux. His expertise ranges from Linux server deployment to installing Linux on small devices such as laptops and portable devices.

INTRODUCING ARC TANGENT-A4 CONFIGURABLE PROCESSOR

San Jose, CALIF. — At the Microprocessor Forum this week, ARC Cores will announce its new ARC Tangent-A4 processor architecture, the next generation configurable processor from the company. The new CPU/DSP, expected to be available next year, is one of the key components of the ARC Tangent platform, a framework enabling integration of hardware/software IP and related tools from ARC’s subsidiary companies – MetaWare, Precise Software Technologies, and VAutomation – and third parties. Key objectives in developing the Tangent platform were to accelerate SoC (system on chip) development by providing verified hardware and software building blocks for systems development, including Ethernet, USB, MP3, DSP functions and a range of real-time operating systems like VxWorks and MQX. Hence the Tangent platform allows customers to concentrate on their value-add activities.

The Tangent-A4 processor, sitting at the heart of the Tangent platform, facilitates this approach with increased system-level capability, peripheral integration, and development tools that provide engineers with relevant program performance quickly without having to do too much work in searching for it. The Tangent-A4’s performance in real-world embedded applications has also been increased with the addition of a configurable data cache and enhanced DSP extensions. Jim Turley, VP marketing at ARC Cores, says, “In embedded systems design, customers are looking to develop solutions for real problems in real-time. That’s why we created the Tangent platform based around the Tangent-A4 processor. With software and hardware building blocks from ARC’s subsidiary companies and third parties, powered by the high performance Tangent-A4 configurable processor, we’re taking configurability to the system-level for applications requiring DSP and multimedia support.” Some of the key new features of the Tangent-A4 include a full-featured multi-way configurable data cache and a vastly improved configurable interrupt system. The latter is important in real-time systems with many peripheral devices requesting attention of the processor. Development tools are the other key improvement available for Tangent-A4. A signal visualization tool has been added, a powerful debug capability that allows the developer to see a two- or three-dimensional graphic of program data as it is being computed; visualization is invaluable for programmers dealing with complex algorithms. In addition, application code can be targeted at a new cycle-accurate simulator – enabling designers to work with HDL simulation accuracy but with greater execution speeds. The company’s website is located at: http://www.arccores.com .

JNI CORP LAUNCHES BUS ADAPTER FAMILY

San Diego, CALIF. — JNI Corporation, the first Fibre Channel company to introduce 2 Gigabit-per-second (2 Gb/s) technology to the market, today announced the availability of its newest 2 Gb/s products for beta testing, the 2 Gb/s Ready FibreStar FCE-6460 Fibre Channel-to-PCI and FCE-1473 Fibre Channel-to-SBus host bus adapters (HBAs), based on JNI’s new Emerald IV Fibre Channel controller. The 2 Gb/s Ready FibreStar HBAs have been tested and qualified on a variety of operating system platforms, including servers and storage subsystems. By the beta shipping date, the HBAs will be running on Sun Microsystems Solaris and Microsoft Windows NT. Subsequent driver releases will include Windows 2000, Hewlett Packard HP-UX, IBM AIX, Novell NetWare and Red Hat Linux.

JNI’s FCE-6460 and FCE-1473 2 Gb/s HBAs offer the same stability and reliability of the industry’s first 2 Gb/s Fibre Channel HBA, the FCE-6420D, and will also include a host of new features, including IP and SCSI protocols, FC-Tape support, and 1 or 2 Gb/s auto rate negotiation. This last, “on demand” feature will enable a smooth and automatic migration path to 2 Gb/s-enabled SANs by enabling the FCE-6460 and FCE-1473 operate in 1 Gb/s, 2 Gb/s or mixed 1 Gb/s and 2 Gb/s Fibre Channel environments. Widespread commercial deployment of 2 Gb/s Fibre Channel products is expected to begin in early 2001. The 2 Gb/s Ready FibreStar HBAs offer peak throughput performance of up to 400 MBytes/second (MB/s) in full-duplex mode, ~200 MB/s in half-duplex (PCI only). “As the 2 Gb/s pioneer, we’re very excited about our second-generation 2 Gb/s HBAs,” said Chris Humphrey, director of product marketing, JNI. “Our first 2 Gb/s board has been shipping to Fibre Channel manufacturers for more than a year, enabling JNI to set the standard for performance and interoperability in a 2 Gb/s storage area network. Our leadership ensures our second-generation 2 Gb/s HBA is a stable, enterprise quality product ready for the most demanding high performance environments.” Visit http://www.jni.com for more information.

PATHLIGHT AND PARTNERS DEMONSTRATE SAN SOLUTIONS

Ithaca, N.Y. — Pathlight Technology has teamed with leading storage and connectivity product vendors to demonstrate deliverable, complete heterogeneous Storage Area Network (SAN) solutions that integrate best-of-breed products to solve a broad range of data storage challenges facing today’s e-business. The demonstration was on display in booth 306 at the Fibre Channel Technology Conference (FCTC) being held October 10-12 at San Jose, CA. The demonstration will include solutions from Pathlight that solve important IT and business operations issues including Resource Consolidation, Data Consolidation, Decoupling Storage from Server, Resource Investment Protection, Data Availability, Disaster Recovery, Data Growth, Time to Deployment, Data Migration and Data Security. “This will be one of the most comprehensive SAN demonstrations ever presented,” said Pathlight Product Marketing Director Jim Geronaitis. “We have put together an environment that represents the complete array of technologies found in today’s enterprise to demonstrate how easy it can be to seamlessly tie these valuable resources together. More importantly, the demonstration illustrates available solutions to the very real challenges faced in today’s e-business.”

The demonstration leverages Pathlight’s SAN Gateway and SAN Router to create an integrated SAN environment that combines products from the leading storage vendors including DLT devices from Overland Data, Adaptive RAID systems from Seek and Synchronix RAID storage from ECCS. In addition, Pathlight will integrate a multi-vendor chain of SCSI JBOD drives highlighting its support for all SCSI devices within a SAN environment. The heterogeneous demonstration will also feature a wide range of today’s most popular server and HBA combinations including a Dell Windows 2000 server using a Qlogic Fibre Channel HBA; a Compaq Windows NT 4.0 server featuring SCSI host attachment to the SAN; an IBM AIX server featuring an Emulex HBA; a Sun Solaris server featuring a JNI HBA; and an HP-UX server featuring an HP Fibre Channel HBA. Pathlight Virtual Private SAN access security technology will enable the attached systems to share channels to storage resources via SAN Gateway. Host systems will be connected to both Brocade Silkworm 2400 and Vixel Fabric switches to SAN Gateway. In addition, Virtual Private Map will extend access security to the SCSI-attached hosts in the demonstration enabling them to share storage as well. Visit the Pathlight web site at http://www.pathlight.com/ .

STRATUS FTSERVER IS FAULT-TOLERANT SERVER

Dallas, TEXAS — Stratus Computer, maker of reliable servers, announced today that its Stratus ftServer family provides the only fault-tolerant, symmetric multi-processor platform for running Exchange 2000 solutions. With the integration of collaboration-based applications and conferencing facilities with messaging, the need for continuous availability is becoming of paramount importance for organizational productivity and communications. Exchange 2000 with its increased reliability, combined with the fault-tolerant hardware, software, and integrated service technology of ftServer systems, provides end users with a powerful option for superior levels of availability.

“Exchange 2000 is the connecting fabric in many enterprises and, as its criticality expands within organizations, downtime is not an option,” said David Flawn, vice president of Stratus Industry Marketing. “Stratus ftServer models are ideally suited for Exchange 2000 environments, providing the highest levels of availability in the industry at a price that essentially gives users ‘FT for free’ relative to competitors’ solutions. ftServer systems deliver a continuously available platform, transparent to the application and simple to deploy, manage and maintain.” Announced in April, the Stratus ftServer family employs double and triple levels of modular redundancy to achieve up to ‘six nines’ of hardware availability in some models. The Intel processor-based ftServer uses current- generation Intel Pentium III Xeon processors and supports Windows 2000, including Advanced Server and Datacenter Server versions. Prices for fault-tolerant Stratus ftServer systems begin at $23,500. “With the launch of Exchange 2000, Stratus is prepared to provide a valuable component to customers implementing a reliable messaging solution,” said Stan Sorensen, group product manager, Server Applications at Microsoft. “Stratus has always been a company focused exclusively on fault-tolerant technology, and its ftServer systems deliver a continuously available server platform to a customer’s Exchange 2000 messaging and collaboration infrastructure.” Visit http://www.stratus.com for more information.

AMD WORKSTATION POWERED BY DUAL ATHLONS

San Jose, CALIF. — AMD reached a new milestone with the first public demonstration of a multiprocessor computer designed specifically to work with AMD processors. The demonstration, at the 2000 Microprocessor Forum, consisted of a computer powered by dual AMD Athlon processors, the AMD-760 MP chipset, and next-generation Double Data Rate (DDR) memory. The multiprocessing computer demonstration featured 3D Studio Max, a professional 3D design and modeling application capable of increasing system performance by using both processors simultaneously. “Today’s demonstration brings AMD one step closer to enable our customers to offer next-generation dual processor workstations and servers powered by AMD processors,” said Rich Heye, vice president and general manager of AMD’s Texas Microprocessor Division. “AMD’s dual processor platform is designed to take the extremely successful AMD Athlon processor into the enterprise markets that require multiprocessing workstation and server solutions.”

AMD is a global supplier of integrated circuits for the personal and networked computer and communications markets with manufacturing facilities in the United States, Europe, Japan and Asia. AMD produces microprocessors, flash memory devices, and support circuitry for communications and networking applications. Founded in 1969 and based in Sunnyvale, California, AMD had revenues of $2.9 billion in 1999. For more information about AMD products, please visit our virtual pressroom at http://www.amd.com/news/virtualpress/index.html .

DIGEX OFFERS SAN BASED ON COMPAQ ARCHITECTURE

Beltsville, MD. — Digex, Incorporated, a leading managed Web and application hosting service provider for businesses worldwide, today introduced its enterprise-class Managed Storage (SAN) solutions, strengthening its position as a leader and innovator in the rapidly growing storage market. Based on the Compaq StorageWorks MA8000 family of modular solutions, the Digex Managed SAN can store up to 1.5 terabytes (TB) of customer data per module, with the ability to configure multiple modules for even greater capacity and flexibility. With some 500 complex storage systems already under management, Digex is leveraging its deep expertise to offer the most scalable storage solution managed by any Web hosting provider. As with all of its managed storage offerings, Digex provides full services for the installation, configuration, administration, upgrades, security, and on-going management and maintenance required to ensure that clients’ data is secure and can be rapidly restored. Digex technical personnel are trained and certified on the system, which is tuned, tested, pre-engineered and fully integrated into Digex’s monitoring and back-up systems for the highest levels of availability.

Digex also announced it is implementing a state-of-the-art tape backup SAN solution based on the StorageTek PowderHorn 9310 automated tape library, a flexible, scalable backup architecture that provides extremely high capacity and throughput to meet Digex customers’ expanding storage requirements. The PowderHorn 9310 automated cartridge system is the world’s largest automated tape library, using the fastest precision robotics available for extremely fast data retrieval. Built-in redundancies in the system provide fail-over capabilities and maximum throughput. Industry analyst estimates on the storage services market range as high as $33 billion in 2002. With these additions, Digex is able to provide high-end comprehensive managed hosting services, including servers, software, storage and back-up solutions that can scale to accommodate even the largest enterprise e-business sites. The Digex Managed SAN incorporates Compaq’s SANworks storage management software and the Brocade Silkworm high bandwidth fabric Fibre Channel switch from Compaq.

AGILENT ANNOUNCES PHYSICAL LAYER SOLUTION

Palo Alto, CALIF. — Agilent Technologies Inc., a leading developer of Fibre Channel solutions for storage area networks (SANs), announced three new communications components that together supply all the functions required for a complete physical layer solution for 2 Gigabit-per-second (Gb/s) Fibre Channel systems. The three new devices include a fiber-optic transceiver, a port bypass circuit and a serializer/deserializer (SerDes) chip. They are available as discrete devices and will be implemented in Agilent’s new 2 Gb/s Fibre Channel host adapters by the end of the year. Agilent’s new HFBR-5720L fiber-optic transceiver allows designers to implement a wide range of SAN solutions for Fibre Channel applications from high-end switching systems to high-speed peripheral interfaces. The HFBR-5720L transceiver’s “hot-plug” capability supports configuration changes or maintenance without having to power down the system. Designed for multimode applications up to 300 meters, the 3.3 V device complies with both the proposed specification for 2 Gb/s Fibre Channel and the existing 1 Gb/s standard. Agilent is currently providing samples of the transceiver with volume production scheduled for early 2001.

The Agilent HDMP-2630 and 2631 are the first SerDes ICs capable of supporting both 1 Gb/s and 2 Gb/s data rates. These chips convert 10-bit wide parallel data to and from a 2 Gb serial stream. Designers can set the transmit channel to a different data rate than the receive channel to support auto-negotiation. Both devices are immediately available in volume production. Agilent’s new quad port bypass circuit is the first chip of its kind targeted at 2 Gb/s SAN applications. Designed to minimize part count and reduce cost, the HDMP-0552 simplifies the migration from 1 Gb/s to 2 Gb/s designs by supporting both data rates with the same reference clock. The device’s “CDR (Clock and Data Recovery) Anywhere” technology gives designers the flexibility to place the CDR after any point in the Fibre Channel loop. Agilent is currently providing samples of the device with production expected by the end of the year. Information about Agilent Technologies can be found on the Web at http://www.agilent.com .

SUNBELT LAUNCHES SUPERBOOST-NT FOR EXCHANGE

Dallas, TEXAS — Sunbelt Software Inc. releases the first Windows NT bundled performance, enhancement and tuning tool suite for Microsoft Exchange Servers called SuperBoost-NT. SuperBoost 2000 is expected to be released in Q1, 2000. SuperBoost-NT will speed up deployment of your Exchange servers and optimize efficiency. SuperBoost-NT bundles Symantec’s Norton Speed Disk, a disk defragmenter and optimizer, SuperSpeed.com’s SuperSpeed-NT, a disk caching utility, and Sunbelt Software’s AutoPilot, a dynamic system-tuning package. Together, the three performance products create a software suite that speeds Windows NT applications, utilities and overall performance. The suite brings the user the convenience of purchasing and installing a single suite and offers the added benefit of three performance utilities that, combined, create a synergy that exceeds the individual performance improvements of each product.

“It’s a performance booster for your Exchange Servers,” said Stu Sjouwerman, president of Sunbelt Software. “In some configurations we have seen SuperBoost-NT create a system performance increase of up to 40 percent, we expect to see the same results from SuperBoost 2000 when it is released.” Norton Speed Disk 5.0 for Windows NT cleans up the Master File Table, disk free space, swap file, metadata and any open files, for maximum peak performance without rebooting. Furthermore, key processes, such as anti-virus protection, can remain enabled during the defragmentation process. SuperSpeed-NT is disk caching software that allows users to employ system memory as a hard drive substantially improving access time for frequently accessed files. SuperSpeed-NT is ideal for access intensive applications, such as database, mail, graphics and Internet-related applications, such as Web serving, browsing and searching. The third component in SuperBoost-NT is a tuning utility from Sunbelt Software called AutoPilot. AutoPilot is a highly advanced neural network that monitors the hardware, the operating system (Windows NT) and the applications running to improve performance in real-time. AutoPilot accomplishes this by analyzing the performance data, determining the current status of the system components and assisting the operating system in making scheduling decisions based upon the analysis. It automatically detects and eliminates system bottlenecks and makes symmetric multi-processor (SMP) Windows NT machines scale better.

GENSYM ANNOUNCES AVAILABILITY OF NETSLEUTH

Cambridge, MASS. — Gensym Corporation, a leader in adaptive software products for modeling, simulating and monitoring e-business infrastructure, announced that NetSleuth, network discovery and Internet Protocol reachability analysis (IPRA) software, is available immediately online at http://www.getnetsleuth.com . NetSleuth automatically constructs an accurate model of a network and all of its devices and performs IP reachability analysis to quickly reveal the root cause of a network outage. NetSleuth delivers extraordinary benefits to corporate network managers and network service providers. With NetSleuth, they can experience increased confidence in the stability of their network, handle larger networks without adding staff, save time and money by minimizing the effects of outages and provide higher levels of service.

The U.S. Army Communications-Electronic Command (CECOM) at Ft. Monmouth, N.J., uses Gensym technology to perform IP reachability analysis as part of its Network Management Decision Aid (NMDA) to help maintain battlefield networks. The Army’s accomplishments in this area clearly demonstrated the need for an effective IP reachability analyzer that is quick and simple to use in both the government and private sectors. Gensym used this kind of customer feedback in developing NetSleuth. “Information is one of a modern army’s most valuable assets,” said Michael Lombardi, Chief of CECOM’s Network Management Branch, which developed NMDA. “Tools that enable us to maintain the networks carrying our battlefield intelligence are very important. In corporate America, losing your network will cost you money. We’re playing for higher stakes.” NetSleuth reveals actual network topology by performing Layer 2 and Layer 3 discovery, which identifies electrical or optical connections as well as IP addresses. Most other products only perform Layer 3 discovery, which only identifies logical connectivity. NetSleuth is designed to work on any IP network and supports many transport types, including Ethernet, ATM and frame relay. It also represents higher-level networks such as WANs, LANs and VLANs, making it an extremely comprehensive auto-discovery and IP reachability analysis application.

WEBGAIN ANNOUNCES TOPLINK FOR WEBSPHERE SERVER

Santa Clara, CALIF. — WebGain Inc., the e-business application development company, announced the latest version of its award-winning Java object/relational (O/R) database mapping technology that features support for IBM’s WebSphere Application Server, Advanced Edition v.3.5. The new offering called TopLink for IBM WebSphere enables developers to quickly and easily integrate enterprise Java applications with relational databases, resulting in faster time-to-market. WebGain’s TopLink for IBM WebSphere offers faster development and runtime performance for EJB-based e-commerce applications that use relational databases. This allows enterprise development teams to build distributed Java applications that leverage the performance and scalability of proven enterprise databases without having to manually write and maintain code.

“WebSphere customers need the ability to easily integrate Java applications with relational databases and TopLink helps automate that task,” says Mike Milinkovich, vice president of TopLink at WebGain. “Developers want to use a single product that supports the leading J2EE applications servers on the market today such as IBM WebSphere.” “WebSphere is the first universal Internet software platform that supports any kind of e-business, from startup to mega-site,” said Paraic Sweeney, vice president, Business Transformation, IBM Software. “Working with the IBM WebSphere software platform, TopLink extends the application framework for customers, providing an easier, more efficient working environment for developing and customizing enterprise solutions integrated with relational databases.” WebGain’s TopLink also offers developers a powerful and scalable container managed persistence (CMP) solution, allowing EJB entity beans to be transparently stored in relational databases with no SQL coding. TopLink extends the EJB 1.1 specification with powerful features such as full inter-bean relationships, caching options, and a powerful querying system that permits complex object-level queries and dynamic finder methods. TopLink’s sophisticated features also include a flexible non-intrusive mapping mechanism, an easy-to-use visual mapping tool, and object level transactions.

INSTANTIATIONS BOOSTS PERFORMANCE OF JOVE

Portland, ORE. — Instantiations, Inc. announced today that it has begun shipping JOVE 2.0, Optimizing Native Compiler for Java Technology. JOVE sets the standard for Java technology performance by empowering developers with the tools necessary to create and deploy today’s most demanding e-business systems. Dramatic system performance gains, which would normally take weeks or months of engineering work, are achieved in minutes with JOVE 2.0. Instantiations has also significantly reduced the price of the product and has removed all run-time fees. There is now no fee for unlimited deployment of customer applications that have been created with JOVE. “For the past year our JOVE product has provided developers with a state-of-the-art tool set for creating and deploying high performance Java applications,” said Mike Taylor, Instantiations’ CEO. “With the release of JOVE 2.0, we are now providing a more powerful product at a fraction of the previous price.”

The JOVE 2.0 Optimizing Native Compiler for Java Technology includes a unique and highly sophisticated software optimization engine, a native Java compiler, an intuitive console and scaleable runtime platform that enables complex Java applications to be deployed as high-performance, robust executable files. “Building and deploying next generation Internet applications can be quite complex and demanding. When it came time to deploy our Java solution to customers, we choose JOVE,” said Jon Hancock, CEO, patternWare, Inc. “The server scaling and performance benefits enabled by JOVE are worth the price alone. When you add to that the ability to deploy a native executable without the worries of Java deployment dependencies, we raised the bar of competitive advantage in the client realm. The bottom-line is that we could not go into day one of production without JOVE.”

TURBOLINUX SUPPORTS IBM ESERVER

San Francisco, CALIF. — TurboLinux, Inc., the high-performance Linux company, announced its support of IBM’s entire eServer line. IBM’s new eServer line offers TurboLinux customers a way to manage end-to-end growth, risk and expense, while offering all-new application flexibility and enhanced server performance, and delivering mainframe-class scalability, reliability, security and capacity on demand. Linux will be the first operating system to run on all IBM servers, supporting a wide range of e-businesses. “TurboLinux has taken a leadership role in delivering Linux-enabled solutions with enterprise-rich features such as clustering and high-performance computing farms,” said T. Paul Thomas, president and CEO of TurboLinux. “Working with IBM to offer TurboLinux from network appliances up to mainframe computers is a real breakthrough for customers who want to develop, deploy and manage applications and solutions across their enterprise.”

“We are very pleased to offer our customers the power and flexibility of TurboLinux across our complete eServer line,” said Dick Sullivan, vice president, Solutions and Integration Marketing, IBM Software. “TurboLinux has consistently focused on meeting the demands of large customers with requirements for scalable Linux solutions.” TurboLinux and IBM will be showing offerings from the xSeries and zSeries, as well as other IBM/TurboLinux solutions for the enterprise during the Mission Possible roadshow visiting six cities across the United States in October and November. To register for the Mission Possible roadshow visit: http://www.turbolinux.com/missionpossible.html

INRANGE ANNOUNCES IN-VSN FC/9000 DIRECTOR

Mount Laurel, N.J. — INRANGE Technologies announced today that IBM will resell INRANGE’s IN-VSN FC/9000 Fibre Channel Director for IBM eServer z900 FICON networks as well as Open Systems Fibre Channel Storage Area Networks (SANs). The general availability of the FC/9000 from IBM has been set for December 8, 2000. The 64-port FC/9000 is the industry’s highest capacity fibre channel director. It is part of the IN-VSN family of Virtual Storage Networking solutions, which lays the infrastructure for building large-scale storage networks that can be extended globally.

INRANGE Technologies designs, manufactures, markets and services networking and switching products for storage, data and telecommunications networks. Our products provide fast and reliable connections among networks of computers and related devices and are used in Fortune 1000 businesses and other large enterprises that operate large-scale systems where reliability and continuous availability are critical. Our products are designed to be compatible with various vendors’ products and multiple communication standards and protocols. For more information, visit http://www.inrange.com or visit http://www.ibm.com .

VITESSE INTRODUCES HEX PORT BYPASS CIRCUIT

Camarillo, CALIF. — Vitesse Semiconductor Corp. expanded its wide offering of Fibre Channel products with a new IC for the growing 2.125Gb/s Fibre Channel storage market. The VSC7147 is the industry’s first six Port Bypass Circuits (PBC) that supports both 1.0625Gb/s and 2.125Gb/s Fibre Channel data rates. The device enables Fibre Channel Disk Arrays, JBODs (Just a Bunch of Disks) and storage subsystem equipment designers to integrate more ports and disk drives and maintain consistently high signal quality using fewer components. In addition to the PBCs that steer serial Fibre Channel signals to disk drives and bypass faulty ports, the VSC7147 features dual FibreTimer Repeater/Retimer cells. Each FibreTimer cell contains an all-digital clock recovery unit (CRU) that can be configured as either a Repeater or a Retimer.

In the repeater mode, recovered data is retransmitted to a recovered clock, allowing for improved jitter attenuation and low latency. In the retimer mode, recovered data is retransmitted synchronously to a local reference clock with no jitter transfer, ensuring compliance with Fibre Channel signal quality specifications. The digital design of the FibreTimers produces a device that is well controlled over voltage, process and temperature variations. The VSC7147 provides two Signal Detect Units (SDU) to detect the presence of valid Fibre Channel encoded serial data, user selectable on-chip receive and transmit termination at either 100 or 150 ohms and low-power 2.5V operation. The device is packaged in a 100-pin TQFP to minimize board size, cost, high frequency routing and jitter in price/performance sensitive JBOD and disk array designs. “The FibreTimer technology, developed by Vitesse, has been phenomenally successful at 1Gb/s. The VSC7147 brings this technology to the emerging 2Gb/s market in a product optimized for the disk array market,” stated Tim Hornback, director of Datacom marketing at Vitesse. Company/product information can be found on the Web at http://www.vitesse.com .

NSTOR INTRODUCES 2U RAID SUBSYSTEM

San Jose, CALIF. — The Fibre Channel Technology Conference (FCTC) being held in San Jose, California, October 10-12, was the venue selected by nStor Technologies, Inc. to introduce the latest in its NexStor 2U (3.5″ high) RAID technology line. Introduced at the show today was the NexStor 3250, a new Fibre Channel RAID system that features active/active controllers and twelve 1-inch disk drives in a compact 2U enclosure. The 3250 is designed for users who require more spindles for performance and is a companion product to the 3150 which utilizes eight 1.6 inch drive in the 2U enclosure for the higher capacity applications. “The trend toward smaller, more dense packaging for all system components is quite clear,” commented Larry Hemmerich, nStor President and CEO. “We have seen it with switch technology, and more recently servers being packaged in 1U and 2U enclosures. nStor has taken the leadership position in the storage arena with enclosures and RAID systems in ultra-dense, 2U packaging. The NexStor 3250 is the latest innovation from our design team that adds tremendous momentum to this push.”

“We have already installed the NexStor 3150, a 2U, 8-drive RAID system that utilizes this same technology, into the market with excellent results. Customers have indicated that the performance of the units is exceptional and, as a result, sales of our new 2U-based products has started to ramp.” nStor has created a powerful RAID system in the new NexStor 3250. Fully configured with twelve 1-inch high, 36 Gigabyte drives, the NexStor 3250F can hold 432GB of data and two active/active RAID controllers all in a unit that is only 3.5″ high. When 73GigaByte drives become available early next year, the unit will hold 864GigaBytes of storage. A NexStor 3250 with a complement of NexStor 1202F expansion units, also announced today, can support up to 3.5TeraBytes of RAID-protected data in a 28-inch high space with today’s 36GB, 1″ drive technology. Additional information about nStor can be found by visiting http://www.nstor.com .

CONNECTCOM ANNOUNCES GBIC VERSION OF ADAPTER

San Jose, CALIF. — ConnectCom Solutions, Inc., a technology leader providing SCSI, 1394 Firewire and Fibre Channel system on chip, board and software connectivity solutions, announced that its 1 Gigabit (Gb) Fibre Channel Adapter is now available in configurations to support both copper-wire interfaces and Gigabit Interface Converter (GBIC) optical modules. The new CSBFC1000G supports the GBIC module system for both optical and copper interfaces for maximum flexibility in the SAN environment. ConnectCom entered the Fibre Channel market this spring with the announcement of its original CSBFC1000 1Gb Fibre Channel host adapter board designed to bring low-cost, high-performance connectivity to PCI-based platforms. ConnectCom’s Fibre Channel adapter provides a turnkey solution for SAN configurations with both computer-to-computer (FC-IP) and computer-to-storage (FCP-SCSI) protocols. Together with data zoning and file sharing software, this Fibre Channel adapter, when used in a SAN configuration, provides users the ability to communicate with one another and to access storage devices without the overhead of a server. The configuration has a much higher data throughput – up to 200 MB/second – than that of the Local Area Network (LAN) configuration commonly used today. ConnectCom’s turnkey solution – both hardware and software – provides for a seamless installation.

The CSBFC1000 family supports all commonly used Fibre Channel physical links. Various optical links can be supported with the new CSBFC1000G, which supports popular GBIC optical interface modules for both long-wave, single mode as well as short-wave, multi-mode links. For low-cost, copper-only configurations, the CSBFC1000 provides direct connections up to 30 meters with a standard HSSDC connector. The ConnectCom adapters are based on an advanced-technology single chip Fibre Channel controller IC that integrates all low-level Fibre Channel protocol functions, high-level (FC-4) protocol assist hardware, advanced RISC processor, host messaging unit and PCI interface to provide the highest performance at the lowest cost. The 64-bit, 66MHz PCI interface supports today’s fastest PCI host platforms, and allows burst data rates of up to 528MB/second to be achieved. Other PCI configurations, including 32-bit data or 33 MHz clock rates are also supported. Visit http://www.ConnectCom.net for more information.

NQL RELEASES NETWORK QUERY LANGUAGE FOR TESTING

Santa Ana, CALIF. — In the first step to extend its market reach beyond Microsoft Windows environments, NQL Inc. announced beta availability of a Java version of its Network Query Language that has been qualified for the Linux operating system. Linux and Unix versions of the core technology offer the same benefits of the five-year development effort that went into Network Query Language for Windows and contain essentially the same deployment features as the Windows editions. According to International Data Corporation, in the hotly contested server market, Linux has the second largest market share (24 percent) after Windows NT (38 percent). The many versions of Unix combined hold a 15 percent share. “With the growing popularity of open platforms, we see this Java support as an important strategic move. Support for these major platforms should help us capitalize on our opportunity to become a standard for content engineering,” said Douglas J. Tullio, chairman, chief executive officer and president of NQL Inc.

Targeted to desktop and server platforms where Java 2 has been implemented, NQL Java Edition is also in the process of being qualified for Macintosh and Sun operating systems. For the short term, the company is providing NQL Java technical support and an installation process for Linux customers who adopt the technology. NQL provides implementation support for e-business solution providers, independent software vendors and enterprise IT departments through its Professional Services Solutions Group. NQL has been shipping a Windows version of Network Query Language in three editions – workstation, server and enterprise – since the second quarter of this year. The NQL Enterprise Edition for large-scale clients was released last week. NQL Java is available in the same editions and the company believes it is a stepping-stone toward future development plans to have NQL running directly on Java-powered mobile devices. For more information, please visit the company’s web site at http://www.nqli.com .

VA LINUX INTRODUCES DEBIAN-BASED 2U SERVERS

Fremont, CALIF. — VA Linux Systems, Inc. announced the immediate availability of Debian GNU/Linux software on its 2200 series Internet servers, complete with technical support, services and free updates via the Internet. Customers ordering VA’s popular 2200 series 2U (3.5-inch-high) rackmount Linux servers will now have the option of choosing between Debian- and Red Hat-based Linux software for their systems. VA Linux is the first major server vendor to pre-install, deliver and support Debian GNU/Linux – the most popular non-commercial version of the Linux operating system and one of the first major Linux distributions, created in 1993. With over 3,800 software packages maintained by more than 500 volunteer developers worldwide, Debian GNU/Linux is highly regarded for its technical maturity and exceptional level of quality maintained by the Debian Project’s extensive network of developers and users. Debian 2.2 is fully compliant with the Linux Development Platform Specification (LDPS 1.0), which allows standardization of application development on Linux.

VA Linux customers who choose Debian 2.2 on the VA Linux 2200 series servers will be able to take advantage of automatic software updates via “apt-get” – Debian’s Internet-based software update management tool. “Apt-get,” an integral Debian feature since March 1998, saves time by automatically identifying required security updates, software dependencies and desirable enhancements. A system administrator can quickly and efficiently download and install the appropriate software packages with a few simple commands, for a single system or across an entire network. VA Linux customers will be able to install the latest Debian GNU/Linux software updates as they become available – according to their customized specifications – from a VA Linux-maintained Internet site at no additional cost. “Many of our Internet customers have asked for Debian-based servers because of the high quality, security and seamless update functionality offered by Debian GNU/Linux,” said Dr. Larry M. Augustin, president and CEO of VA Linux Systems. “Many of our OSDN Web sites, including Slashdot and Linux.com, run Debian GNU/Linux, and we applaud the Debian Project for its long, outstanding contribution to the Free Software and Open Source communities. We are proud to begin providing Debian on our 2200 series servers and look forward to offering the Debian option on more of our systems – and eventually through our Build-to-Order Software Selector (BOSS) as well.”

ILOG TECHNOLOGY UNIFIES OPTIMIZATION METHODS

Mountain View, CALIF. — ILOG announced the release of ILOG Concert Technology, a C++ optimization modeling library and interface that enables increased flexibility for application developers and will provide a foundation for new optimization applications while making existing applications run faster. ILOG Concert Technology is integrated into the latest versions of ILOG’s two optimization solving engines, ILOG CPLEX 7.0 and ILOG Solver 5.0 and upgrades of the ILOG Optimization Suite. ILOG Concert Technology unifies the two primary optimization methods, constraint programming (CP) and math programming. Using the new technology “bridge,” developers can employ the best of both optimization methods to solve new classes of optimization problems – without having to learn different software techniques. Also announced today is a 60 percent performance improvement in ILOG CPLEX 7.0.

Optimization software helps businesses in different industries make the best decisions given their resources and a wide range of possible choices. Decisions such as the most efficient way to sequence an automotive assembly line, the best way to bottle beer or the earliest real delivery date for a customer order – have all been made after using optimization software. Optimization is also being used on the web to enable customer self service applications. ILOG Concert Technology is expected to help bridge the gulf between operations research (OR) professionals and software development teams. Leading companies frequently employ OR professionals to develop planning models, but must rely on software developers to integrate these models into a software application. ILOG Concert Technology allows optimization models implemented in C++ to be expressed in a format familiar to OR professionals, shortening deployment cycles. This new technology also provides a building block for blending long-term planning, traditionally the domain of the OR realm, and short-term decision making, which characterizes many uses of CP technology. The traditional divide between CP’s scheduling orientation and MP’s planning orientation presents very different approaches to problem solving. The Web’s e-business environment is blurring the lines between these two areas, driving the need for tools that offer benefits to both.

============================================================

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industry updates delivered to you every week!

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion XL — were added to the benchmark suite as MLPerf continues Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing power it brings to artificial intelligence.  Nvidia's DGX Read more…

Call for Participation in Workshop on Potential NSF CISE Quantum Initiative

March 26, 2024

Editor’s Note: Next month there will be a workshop to discuss what a quantum initiative led by NSF’s Computer, Information Science and Engineering (CISE) directorate could entail. The details are posted below in a Ca Read more…

Waseda U. Researchers Reports New Quantum Algorithm for Speeding Optimization

March 25, 2024

Optimization problems cover a wide range of applications and are often cited as good candidates for quantum computing. However, the execution time for constrained combinatorial optimization applications on quantum device Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at the network layer threatens to make bigger and brawnier pro Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HBM3E memory as well as the the ability to train 1 trillion pa Read more…

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing po Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HB Read more…

Nvidia Looks to Accelerate GenAI Adoption with NIM

March 19, 2024

Today at the GPU Technology Conference, Nvidia launched a new offering aimed at helping customers quickly deploy their generative AI applications in a secure, s Read more…

The Generative AI Future Is Now, Nvidia’s Huang Says

March 19, 2024

We are in the early days of a transformative shift in how business gets done thanks to the advent of generative AI, according to Nvidia CEO and cofounder Jensen Read more…

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, codenamed Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from Read more…

Nvidia Showcases Quantum Cloud, Expanding Quantum Portfolio at GTC24

March 18, 2024

Nvidia’s barrage of quantum news at GTC24 this week includes new products, signature collaborations, and a new Nvidia Quantum Cloud for quantum developers. Wh Read more…

Alibaba Shuts Down its Quantum Computing Effort

November 30, 2023

In case you missed it, China’s e-commerce giant Alibaba has shut down its quantum computing research effort. It’s not entirely clear what drove the change. Read more…

Nvidia H100: Are 550,000 GPUs Enough for This Year?

August 17, 2023

The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…

Shutterstock 1285747942

AMD’s Horsepower-packed MI300X GPU Beats Nvidia’s Upcoming H200

December 7, 2023

AMD and Nvidia are locked in an AI performance battle – much like the gaming GPU performance clash the companies have waged for decades. AMD has claimed it Read more…

DoD Takes a Long View of Quantum Computing

December 19, 2023

Given the large sums tied to expensive weapon systems – think $100-million-plus per F-35 fighter – it’s easy to forget the U.S. Department of Defense is a Read more…

Synopsys Eats Ansys: Does HPC Get Indigestion?

February 8, 2024

Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…

Choosing the Right GPU for LLM Inference and Training

December 11, 2023

Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…

Intel’s Server and PC Chip Development Will Blur After 2025

January 15, 2024

Intel's dealing with much more than chip rivals breathing down its neck; it is simultaneously integrating a bevy of new technologies such as chiplets, artificia Read more…

Baidu Exits Quantum, Closely Following Alibaba’s Earlier Move

January 5, 2024

Reuters reported this week that Baidu, China’s giant e-commerce and services provider, is exiting the quantum computing development arena. Reuters reported � Read more…

Leading Solution Providers

Contributors

Comparing NVIDIA A100 and NVIDIA L40S: Which GPU is Ideal for AI and Graphics-Intensive Workloads?

October 30, 2023

With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…

Shutterstock 1179408610

Google Addresses the Mysteries of Its Hypercomputer 

December 28, 2023

When Google launched its Hypercomputer earlier this month (December 2023), the first reaction was, "Say what?" It turns out that the Hypercomputer is Google's t Read more…

AMD MI3000A

How AMD May Get Across the CUDA Moat

October 5, 2023

When discussing GenAI, the term "GPU" almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word "GPU" is assumed to mean "Nvidia" products. (As an aside, the popular Nvidia hardware used in GenAI are not technically... Read more…

Shutterstock 1606064203

Meta’s Zuckerberg Puts Its AI Future in the Hands of 600,000 GPUs

January 25, 2024

In under two minutes, Meta's CEO, Mark Zuckerberg, laid out the company's AI plans, which included a plan to build an artificial intelligence system with the eq Read more…

Google Introduces ‘Hypercomputer’ to Its AI Infrastructure

December 11, 2023

Google ran out of monikers to describe its new AI system released on December 7. Supercomputer perhaps wasn't an apt description, so it settled on Hypercomputer Read more…

China Is All In on a RISC-V Future

January 8, 2024

The state of RISC-V in China was discussed in a recent report released by the Jamestown Foundation, a Washington, D.C.-based think tank. The report, entitled "E Read more…

Intel Won’t Have a Xeon Max Chip with New Emerald Rapids CPU

December 14, 2023

As expected, Intel officially announced its 5th generation Xeon server chips codenamed Emerald Rapids at an event in New York City, where the focus was really o Read more…

IBM Quantum Summit: Two New QPUs, Upgraded Qiskit, 10-year Roadmap and More

December 4, 2023

IBM kicks off its annual Quantum Summit today and will announce a broad range of advances including its much-anticipated 1121-qubit Condor QPU, a smaller 133-qu Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire