Volume 1, Issue 1

Downloadable Version (.pdf format)

August, 2002

• ATA-100 vs. ATA-133
• Windows Updates
• IEEE1394 & USB Info
• Wireless Shootout

• RAID Technology Info
• Single or Dual Processor?
• Intel i845G vs. SiS 650 Comparison

ATA-100 vs. ATA-133 / Performance vs. Price

One of the hot new products in the market is the ATA133 hard disk drive. Drive manufacturers haven't quite kept up with motherboard and chipset technology since almost every new motherboard now supports the ATA133 or the Ultra DMA Mode 6 standard, but there are only a handful of drives that are actually capable of using this standard.

ATA133 suggests that a capable hard drive can theoretically push 133MB/s in burst mode. This figure is somewhat artificial in the real-world as performance is what really counts in the end. In comparisons between ATA100 and ATA133 hard drives, the difference in performance between the two was negligible. In some tests the results were very close, while in others tests the ATA133 drive took the lead. In certain tests, the ATA133 drive was actually beaten by the ATA100 model. ATA133 should theoretically be 33% faster than ATA100, but the tests results basically didn't show much of an improvement at all.

Conclusion

If you're thinking about upgrading your drives from ATA100 to ATA133, you may want to reconsider; the slight increase in performance is far outweighed by the added expense of the ATA133 badge.

Windows Updates: What are you missing?

  On average, the current installation of Windows XP Professional requires 20+ updates from Microsoft to be completed. Updates are available for most Microsoft Operating Systems and applications. Many people do not run these updates citing lack of time, desire, or necessity to do so... but at what cost? Windows Update is the online extension of the Windows Operating System and includes the following:

After entering Windows Update, you are asked to allow Microsoft to scan your system for changes. This allows the Update server to find what patches you have, or may need for your system. After the scanning is completed you are presented with a list of available updates for your machine. You are free to choose which updates you would like to load, but we strongly recommend loading all updates listed in the "Critical Update" section at the top of the Windows Update page.

Making sure that your machine is kept up to date with the latest patches from Microsoft is the best way to insure that you have the best defense against bugs and other problems which might otherwise cause damage or corruption to your computer. 

Please visit: http://www.windowsupdate.com for more info.

FireWire and USB Technology

FireWire IEEE1394

Apple Computer first introduced "FireWire" in 1986 and later on became the open standard now known IEEE 1394.  This same standard was also referred to as "i.LINK" trademarked by Sony Corporation.  The IEEE 1394 unlike many other wired busses does not require a PC for functionality and it does not need to have digital data converted to analog.  For example, a file can be transferred between two digital video cameras or a picture from a digital camera can be sent directly to a printer.  Other advantages are Plug and Play and Hot Swap capability, ideal for portable devices

There are two cable specifications for the IEEE 1394 standard, which are 6-pin and 4-pin connectors.  The difference is the 6-pin connector includes power and ground pin, allowing a 1394 cable to provide power (up to 60watts) to a device. The 4-pin connector does not have this power pin, therefore 4-pin devices require their own power supply. 

USB (Universal Serial Bus)

The Universal Serial Bus (USB) was an idea developed to run multiple peripherals on the computers without physically

 installing boards, manually allocating system resources, and changing device without power up and down the computers. USB can be connected up to 127 devices to a host computer using a single interface and a system of USB hubs.  Many USB devices have a hub built into it like some keyboard having an additional USB port for a USB mouse. USB consists of two types of cables:  Series A and Series B.  Series B cables are use for low-speed (1.5Mbps) devices such as keyboard and mice and are limited to 3 meters in length.  Series A cables are use for high-speed (12Mbps or above) devices and can go up 5 meters in length.  A repeater hub can be used to extend the distances between the devices and the host or even further with special USB repeaters.  USB devices can draw power directly from the system, or from their own power supply.

  The recently released USB 2.0 is fully backward compatible with all older USB devices.  All PC operating system support USB 1.1.  As for USB 2.0, Microsoft has developed a driver for Windows 2000 and Windows XP.  Other operating systems require a third party driver to support USB 2.0.

Wireless Technology Shootout

Information on RAID Technology

What is RAID?

  R.A.I.D is a term short for Redundant Array of Independent (or Inexpensive) Disks.  RAID is when two or more hard disks are combined in way to either increase performance, add data protection (fault tolerance), or both.  It is a very useful technology for today’s high-demand storage subsystems, with practical and affordable configurations for everything from home desktop systems to high-end to workstations and servers.

RAID Levels

 Today there are many choices for RAID configuration, and you must know the abilities of each type of RAID to choose the right one.  It is best to choose the type of RAID level you need first before choosing a RAID product.

Hardware or Software RAID?

  Most forms of RAID can be implemented in either software or hardware.  Hardware RAID just means you use a PCI RAID controller card or a controller chip built onto your motherboard.  Hardware RAID uses its own BIOS and configuration utility for setup of the arrays - a utility you must access during system boot process.  Software RAID is a function of the disk management of your operating system. Hardware RAID is the preferred configuration since it performs independent of the operating systems installed and does not impact system performance by using CPU time.  Hardware RAID is also more reliable than software RAID and gives you more options.  Software RAID does have a use when you do not want to spend money buying RAID controller card.  

IDE or SCSI CHANNEL RAID?

 You now have the choice to build a RAID system using either IDE/ATA hard drives or SCSI hard drives.  Standard IDE RAID uses any two IDE hard drive to create RAID 0 (striping) or RAID 1 (Mirroring) disk array, or more recently you have the option of RAID 0+1 (uses 4 drives to combine benefits striping and mirroring).

  Some newer IDE RAID controllers can do RAID 5 or handle more than 2 to 4 drives, such as the Promise SuperTrak SX6000.  In general, IDE RAID arrays are easier to set-up than SCSI arrays and just as reliable.  IDE RAID provides a good low cost alternative to SCSI RAID.  Promise www.promise.com and Highpoint Technologies www.highpoint-tech.com are leaders in the field of IDE controllers and IDE RAID cards.  You can even get a motherboard with built-in IDE RAID controller option, giving you two IDE channels separate from the IDE channel already built into the chipset.

Still, in some cases IDE RAID is just not enough.  For Corporate/Enterprise Level or Web Servers, very fast data IO throughput is critical, as is 24/7 reliability. SCSI RAID is the solution as it offers more options (RAID 0, 1, or 5), more drives that can be supported per channel or array, faster input/output (especially using Ultra 160 technology), and hot-swap drive bays, and therefore is best suited to critical server environments where even shutting down the system for 10 minutes is a major problem.  Adaptec www.adaptec.com and Intel www.intel.com are the primary manufacturers for top-quality SCSI RAID cards.  Yes, there are IDE hot swap and some higher-end IDE RAID cards that can do RAID 5 but in most situations SCSI would be your best option fro hot swap.  Some server motherboards now have the option of ZCR (Zero Channel Raid) that allows for a motherboard with onboard SCSI to be upgraded to SCSI RAID at a lower than normal cost by adding in special card in the zero-channel slot.

When is using a dual processor system a good idea?  This is a question often asked by system builders, especially when configuring a high-end workstation or server.  Unfortunately there are several misconceptions about what a dual processor system can or cannot do.  Let us review the basics.

  First, there are currently several types of dual processor systems (also known as SMP = Symmetric Multi-Processing) on the market, primarily supporting Dual Pentium III/Tualatin, Dual XEON (P3/P4), and Dual Athlon MP configurations.  These will sometimes require special power supplies.  Second, in order to take advantage of the SMP functions, you must use an operating system that takes advantage of SMP (that is "SMP aware").

  The real question is performance.  Will running two processors make the system twice as fast?  Does 2 x 1.0 GHz CPUs = 2.0 GHz speed?  Unfortunately no, because both CPU's cannot be utilized 100% of the time.  Normally there is between a 10-40% gain in overall performance depending how many applications are being run and whether they can take full advantage of SMP.  Uses such as Email Server, Database or Application Server, Web Servers, CAD/CAM, and Video Editing are prime candidates for multiprocessor configuration.  Dual processor systems will cost more to build and upgrade, but can provide a performance edge in high-demand environments.  In many other cases a single processor system with a lot of memory will more than meet your customer's reliability or performance.  The best way to decide is to know exactly how the system will be used, now and in the future.

Intel and SiS have both released low cost, integrated graphics solutions to the P4 market. These chipsets share many features including FSAA support, AGP 4x slot support, SMA graphics memory and have optional DVI Support. Neither chipset supports hardware T&L, or hardware vertex or pixel shading. A key difference between the two chipsets is in the way they handle memory. Both chipsets share system memory, currently up to 64MB, but the Intel chipset has an advantage. The Intel i845G Chipset uses Dynamic Video Memory Technology, (DVMT), which borrows between 0MB – 64MB (Increments of 8MB) from the system memory as it is needed for the graphics operations.