For example, I don’t have to buy a new case, nor do I have to buy a new optical drive. I also don’t need new audio speakers, mouse, keyboard or monitor. In my current setup I could also transfer over my video card as well.

When you add these things up it shaves off a lot of cash you would otherwise have to spend.

Many PCMech readers have built PCs before, but if you’re someone who hasn’t yet, I can say without question that your most expensive build will probably be the first one. And this is because you have nothing to transfer from a previous build. Everything must be bought new and yes, it can get costly - but only for the first time around.

What you can do is think of the future when building your first box so it will cost less when you upgrade later. The checklist is short and easy to remember.

Future-use things to consider when PC building

Computer Case

Your case should be a full-sized tower. The reason is so you can use standard ATX or MicroATX motherboards. This gives you more choice of what motherboard you want to use now and in the future.

The case you buy should be above average. You want something built well that will withstand the test of time. Being that your intent is to use the same case over and over, build quality matters.

Motherboard

Use a motherboard that supports several different types of processors. The better ones support at least four. This Intel motherboard for example does just that. For now maybe you want a Core 2 Duo processor but later on want to upgrade to a Quad-core. Using a motherboard that supports multiple CPU types will save you cash because you don’t have to buy another board when ready to upgrade.

Also bear in mind the new board you buy should support (at the time of this writing) a minimum of 8GB of RAM which the one linked above does. If you buy a new board that only supports 4GB, that’s your limit - and that’s not good. 8GB should stay current for at least 5 years, and you can incrementally update as needed.

Optical Drive

This is something you don’t have to put a ton of cash into because more people are moving away from the optical format in favor of flash-based media.

Concerning this particular component, the key is to avoid things you don’t need that add extra cost, such as:

• LightScribe - This is more for "cool factor" than anything else. You’re better off just labeling your discs with a Sharpie marker. It’s cheaper and does the same job. True, it doesn’t look as professional, but nobody cares.
• Super fast DVD burning - Not necessary. Why? Because getting blank media that supports the higher speeds costs more and probably isn’t available on the shelf where you buy discs. Most of the time the blank media you buy will be 12x max write speed for DVD and 30x max for CD, so the fact your drive can burn faster means nothing unless you’re willing to custom order blank discs that support it - which you probably won’t.
• Blu-Ray compatible - Again, not necessary. The only people truly interested in this format are those who have Blu-ray players for their television. And yes, you will get a big-big 25GB of storage on one of these BD-R discs, but, each disc at present costs $5 - and that doesn’t include shipping. You’re better off using flash-based media, if only for the fact that write/rewrite is much easier and faster compared to optical.

Input Devices (Keyboard and Mouse)

Concerning the mouse, make sure it connects via USB. With wireless, go for RF and not Bluetooth. For example, a really kick-ass wireless mouse is the Logitech G7. It’s got a ridiculous price tag (but worth it to some) and is USB/RF-wireless based. And don’t worry, there are plenty of RF-wireless USB mice that are much lower in price, like this one.

Note that there’s nothing bad or wrong about Bluetooth wireless, but you get more choice with RF. A lot more.

Concerning the keyboard, you want something that will stand the test of time like your computer case. There are many keyboard makes to choose from, but typically the best for long-term use are made by Microsoft and Logitech. Sure, other keyboards have lots of whiz-bang features on it, but what matters most is comfort and longevity.

With Microsoft and Logitech it’s the simple things that make them better, such as not having the lettering literally wear off your keys in less than six months.

Of course, neither beat the best computer keyboard of all time, the IBM Model M. But that takes a little bit of doing (i.e. a USB converter) to make the older ones work. And lets not forget they’re frickin’ loud. A resounding CLICK-CLACK happens every time you press a key on one of those. But it’s a good sound. At least I think so.

Audio speakers

The only rule to follow here is to use an audio speaker setup that is not overly dependent on software. And what I mean by that is you should not buy speakers that require software just to hear sound come out of them.

Whatever you buy should work like this: Open the box, take the speakers out, plug the AC adapter into the wall, plug in the audio cable to the sound card, done.

Software that controls the sound card is fine (and necessary). Software that’s required just to make the speakers work is bad.

If you want an example of bad, Creative is notorious for having speaker setups that are too "attached" to software, so to speak. And without their proprietary software installed, the speakers either sound like crap or not work at all. Not good.

Final notes

Many people make the mistake of building a PC that only has "right-now" tech in it, meaning the end build will not stay current for more than 2 years at the most.

For example, let’s say you bought a brand new motherboard today but it can only use a Core 2 Duo processor as the fastest it will support. That’s a mistake. Yes, the Core 2 right now is fast - but it won’t be 2 years from now and you’ll be forced to buy another motherboard later when the time comes. What you should have done is bought a board that can support Core 2 or Quad-core. When the Core 2 becomes too slow, it’s not a problem because you can upgrade to a Quad-core and get another 2 or 3 years out of the same motherboard, thereby saving you money.

When building your first (or next) computer box, consider the future. Buying the right stuff the first time and spending a little extra means savings for the long-term.

Post from: PCMech. Helping Normal People Get Their Geek On And Live The Digital Lifestyle.

If you stock your built PC with parts that are the right choice the first time, your Windows installation will run better and faster.

Please Login or Register to read the rest of this article. Gold/Silver Membership required.

Post from: PCMech. Helping Normal People Get Their Geek On And Live The Digital Lifestyle.

For example, I have a home server running on a Biostar all-in-one motherboard with a Celeron 1.3 Ghz processor and 512 MB RAM. The machine has a RAID 3 card installed for redundant storage and runs Windows 2000 like a rock. This server houses all of our music and movies through a couple of shares.

One great benefit to using an older motherboard is the power requirements are typically a lot less than newer machines, which means the machine will not produce as much heat. On our server, I have it configured to automatically start up in the morning as well as automatically shut down at night so it doesn’t consume power while we are sleeping at night.

Anyhow, the point being if you are looking to build your own home server, give a good look at all-in-ones as they are typically abundant and cheap. On top of that, going with a second or third generation motherboard (look on eBay), could really make the cost appealing.

Post from: PCMech. Helping Normal People Get Their Geek On And Live The Digital Lifestyle.

The PnP/PCI Configuration section of the BIOS controls the settings for the motherboard’s PCI slots. This includes the plug-and-play capability of the BIOS. Lets look at those BIOS settings.

• PNP OS Installed
  If all your operating systems support Plug & Play (PnP), select Yes so that they can take over the management of device resources. If you are using a non-PnP-aware OS or not all of the operating systems you are using support PnP, select No to let the BIOS handle it instead. Some say that it is best to leave this option set to No regardless of whether your OS is PNP-capable or not. The reason is that when it is set to No, the BIOS will attempt to resolve any resource conflicts. If it is set to Yes, even if a conflice is detected, the BIOS will ignore it. So, setting it to Yes provide a bit of a safety net, and it will not affect the ability of the OS to perform PNP on its own.


• Reset Configuration Data (Force Update ESCD)
  ESCD (Extended System Configuration Data) is a feature of the Plug & Play BIOS that stores the IRQ, DMA, I/O and memory configurations of all the ISA, PCI and AGP cards in the system (PnP or otherwise). Normally, you should leave the setting as Disabled. If you encounter serious problems with the installation of a new PCI card, this settings can help bail you out. Such a conflict would be serious enough that the OS may not start. If this happens, you can go into the BIOS and enable this option. Next time the PC boots, the BIOS will go and re-configure the settings for all PNP cards. The BIOS will automatically reset this setting to DISABLED next time you boot.


• Resources Controlled By
  Normally, the BIOS controls the IRQ and DMA assignments of all of the boot and PNP devices in the system. When this option is set to AUTO, this is what happens, and the ESCD is the mechanism for doing it. If you set this option to Manual, you will be able to manually assign all IRQ and DMA information, usually via a sub-screen of the BIOS that will enable if you set this option to Manual.


• PCI/VGA Palette Snoop
  This option is only useful if you use an MPEG card or an add-on card that makes use of the graphics card’s Feature Connector. It corrects incorrect colour reproduction by “snooping” into the graphics card’s framebuffer memory and modifying (synchronizing) the information delivered from the graphics card’s Feature Connector to the MPEG or add-on card. It will also solve the problem of display inversion to a black screen after using the MPEG card.


• Assign IRQ for VGA
  Many high-end graphics accelerator cards now require an IRQ to function properly. Disabling this feature with such cards will cause improper operation and/or poor performance. Thus, it’s best to make sure you enable this feature if you are having problems with your graphics accelerator card.


• Assign IRQ for USB
  Assigns an IRQ to the USB controller. It enables or disables IRQ allocation for the USB (Universal Serial Bus). If you are using AGP, this shoudl be enabled. If you are not, you can disable this to free up an IRQ.

Post from: PCMech. Helping Normal People Get Their Geek On And Live The Digital Lifestyle.

System Speed Control

With some motherboards, all of the CPU and Voltage settings are controlled via the CMOS rather than via a series of jumpers on the motherboard. In fact, most boards on the market today do it this way. Only the older boards still have you set the processor settings via jumpers. the placement of these options varies from BIOS to BIOS. Sometimes they are spread out all around the setup utility. Sometimes they are all grouped together in their own menu. However it is laid out in your BIOS, let’s go over some of these settings:

• System Performance
  Some boards have an overall option to change the settings in a global fashion to increase speed. Normal would be the default setting, with FAST and TURBO being additional options.
  
• CPU Frequency Select
  This is where you set the front side bus speed for your system. The available options depend on the speeds supported by your motherboard, but 100MHz and 133 MHz are pretty standard options today.
  
• Frequency Stepping
  Most better motherboards allow you to increase the front side bus speed in increments. Commonly, you can increase the speed in 1 MHz increments. So, if you set the above setting to 100 MHz, then you can use this setting to increase it to 101, 102, 103, and so on. This is great for overclockers who require a fine-tuned control over the FSB speed in order to max out their available speed.
  
• DRAM Clock
  This allows you to control the speed at which the DRAM operates. The default setting is typically BY SPD. The SPD is the serial presence detect mechanism of the DIMM. This feature became standard on DIMMs after the advent of PC100. It is an EEPROM on the corner of the DIMM that holds all of the specs necessary to run the module. By having the BIOS determine memory settings BY SPD, it uses this EEPROM as the reference and automatically fine tunes settings to work to the specs demanded by the module. There are reasons why you might not want to just fall back on the SPD setting, including the fact that the manufacturers don’t always get it accurate when they program the EEPROM. So, many people override the setting with the manual options, and set the speed manually according to the bus speed.
  
• CPU Ratio Select
  This is where you select the multipler at which your processor operates. Most boards support a variety of multipliers.
  
• CPU Voltage Select
  Allows the user to control the voltage delievered to the CPU core. Most users will not fool with this, but overclockers very well may as they fine tune the voltage to aid in speed and stability of an overclocked processor.
  
• DDR Vcore Select
  Adjusts the voltage delivered to the DDR memory.
  
• AGP Voltage Adjust
  Adjusts the voltage sent across the AGP port to the video card.
  
• C.I.H. 4-Way Protection
  The actual defintion of this varies from board to board, but basically it is a protection machanism to protect your BIOS for being written by a virus or some other nasty program that is trying to corrupt your BIOS. If this is enabled, only a BIOS specific update utility will be permitted to write to the BIOS. This utility would be used to flash the BIOS.

DRAM Control

If the DRAM clock is set manually, you will have a few other options to take note of and define:

• CAS Latency
  CAS Latency is a measure of latency of a memory chip. CAS stands for Column Access Select. Basically, it is a measure of how long it takes from when an initial READ command is sent to memory to when the first piece of the resulting data is output. The measurement is done in clock cycles, so a CAS Latency of x means that a READ command sent to memory at clock cycle c will result in data output starting at clock cycle c + x. In short, CAS3 is the standard latency for memory modules, because it is cheaper to manufacture. If the CAS spec is not mentioned or defined, it is probably CAS3. CAS2, though, is a faster memory module. The latency is less, and this leads to faster application speeds. As you might expect, such memory is typically more expensive. If you are using CAS2 memory and it is enabled as such in the BIOS, you might notice a bit of a speed increase. Even with CAS2 memory, though, it could be necessary to choose CAS3 in your CMOS in order to make the memory stable in a higher bus-speed compter. Some BIOS versions default to CAS 2.5.
  
• Bank Interleave
  This feature enables you to set the interleave mode of the SDRAM interface. Interleaving allows banks of SDRAM to alternate their refresh and access cycles. One bank will undergo its refresh cycle while another is being accessed. This improves performance of the SDRAM by masking the refresh time of each bank. A closer examination of interleaving will reveal that since the refresh cycles of all the SDRAM banks are staggered, this produces a kind of pipelining effect. Whether you set this option to 2-bank or 4-bank is determined by the type of DRAM you have and how many banks are on your DIMMs. Most DRAM in use today (sticks 64 MB of higher) are 4-bank, so setting this option to 4-bank is usually right. If you are unsure, you can look up the specs of your DRAM. Otherwise, just disable the option.
  
• DRAM PreChrg to Act CMD: Setup the minimum row precharge time.
  The Choice: 2T, 3T.
  
• DRAM Act to PreChrg CMD: Setup the minimum RAS pulsewidth.
  The Choice: 5T, 6T.
  
• DRAM Active to CMD: Setup the minimum CAS to RAS delay.
  The Choice: 2T, 3T.
  
• DRAM Queue Depth:
  The Choice: 4 level, 2 level, 3 level.
  
• DRAM Drive Strength: Setup the DRAM’s driving current strength.
  The Choice: Auto, Manual.
  
• DRAM Command Rate: Setup the timing at each cycle.
  The Choice: 1T Command, 2T Command.

Post from: PCMech. Helping Normal People Get Their Geek On And Live The Digital Lifestyle.

AT Form Factor

Within the AT form, we have regular AT and Baby AT. They basically differ in size. An AT board is about 12″ wide which means it can’t fit in many of today’s cases. AT boards generally are the older boards, 386 or earlier. Working inside the case was a lot more trouble with these because the size of the motherboard overlapped drive bays and such.

Baby AT is the form used by many 486 and Pentium boards.. Many Socket 7 motherboards and a few Pentium II boards used this form factor. A Baby AT board is roughly 8.5″ wide and 13″ long. The size varies a little from board to board. This reduced size makes it easier to work inside the case simply because there is more room. There are three rows of mounting holes to hold the board in the case.

AT form boards share common traits. They all have serial and parallel ports attached to the case in an expansion slot and connected to the board through cables. They also have a single keyboard connector soldered onto the board at the back of the board. The processor is still at the front of the board and can sometimes get in the way of expansion cards. The SIMM slots are in different places, although they are almost always at the top of the board.

There are some annoyances with the AT design. One is due to the layout. Since all ports are attached to the case and then connected to the motherboard via a cable, the board must have connectors for all of these: COM 1, COM 2, printer port, USB, PS/2 mouse, etc. Often these connectors are directly next to the IDE channel connectors and floppy drive connector. This leads to a severe cramping problem and makes working inside the computer more difficult. Secondly, the AT design is not conducive to efficient cooling of the system. Air is not blown over the areas that need it, namely the CPU. Also, the air flow draws in dust. Over time, the AT power supply will get dusty and the inside of the system will be coated with a layer of dust. For this reason, it is recommended you regularly remove the case and blow off the interior of the case.

You can see a diagram of an AT motherboard here: AT Socket 7 Motherboard. Here is some additional information on PC form factors: A+ Guide 2: Cases, Motherboards, and Processors.

Post from: PCMech. Helping Normal People Get Their Geek On And Live The Digital Lifestyle.

The history of this is quite simple. The early PC’s from IBM used DIP switches to set the system configuration. Later, though, they decided to transfer this information to bit-information stored in some static memory. This memory is low-power, obviously, and uses a form of static RAM called Complementary Metal Oxide Semiconductor (CMOS). Using a small, low-power battery, they could run just enough juice through this memory to get it to hold its content even when the PC is off.

Like any battery, the battery in your PC has a finite lifespan. Eventually, it will become weak and will no longer be able to sustain the contents of your CMOS. When this happens, you will turn your PC on one day and get an error message. Sometimes you will get a checksum error, or it will notify you that the system configuration doesn’t match the CMOS information. Well, that’s because there is no CMOS information anymore because you need to replace your battery. And one problem is that users very seldom record their CMOS information or back it up.

This is usually a simple repair, but some oddball manufacturers actually soldered the batteries in, making this a much tougher job. If your battery is soldered in, you may want to take the whole thing to the shop. If you’re experienced with soldering, then you can tackle it yourself. The good news is that most manufacturers are now using easily removed batteries, the kind about the size of a nickel and can be removed by moving a small prong.

Before you do anything, though, you should record what your computer is supposed to know. If your battery is already dead, there’s nothing you can do. If your configuration is still there, though, record it. Go into CMOS and write down the info (HINT: You can write it real small and tape a copy inside your PC case). Easier yet, just go to those screens that are important and just hit the Print Screen button on your keyboard. There are also utilities out there that can backup your CMOS settings and record it to a diskette. This is quite convenient, especially when you don’t have a setup disk for the PC. After you remove the old battery, your computer will forget everything.

Let’s go through the battery replacement procedure:

1. Turn off the computer, unplug it, and remove the case.
   
2. Remove the old battery. Record which end faced what direction. Each end has a + or — on it. With skill and dexterity, the battery should snap out. Just study it, and you’ll figure out how to get it out. Don’t force it, though. It may be soldered in. If its a new board, you may only need to move a prong to take the battery out. Many newer boards have a small, flat coin-shaped battery which is a lot easier to remove.
   
3. Get a replacement battery. Take the old one to the store and match them up. It should be pretty easy to come by. The CR 2032 is a pretty standard size of battery for us on motherboards. Once you find the correct battery model, you may want to write it down somewhere.
   
4. Put the new battery in. Make sure the + and — face the same way as before. It should snap in. If your experienced at soldering, and your computer demands this, you may want to do that yourself.
   
5. Put the case back on and plug it in. When you turn it on, expect some type of error message like incorrect cmos. Don’t cry. This will happen. You just need to go into the CMOS and plug in all that info that you recorded before you started. If you didn’t do that, you’ll need to break out the manuals and find the info the hard way.

Good luck!

Post from: PCMech. Helping Normal People Get Their Geek On And Live The Digital Lifestyle.

Apollo Pro 133
This chipset, as you can tell from the name, supports the newer PC133 standard, meaning it can support 100MHz and 133MHz system busses. This helps to ensure that the chipset will be useful for a little while. The chipset is comprised of the VT82C693A system controller and the VT82C596B bus controller. It supports all Slot 1 processors as well as Socket 370 processors, and it supports UDMA/66 and AGP 2X, USB, ACPI, etc. Like other Via chipsets, it also supports the asynchronous memory bus, meaning that the memory can operate at a different speed than the main system bus. Overall, a great chipset, and one you no doubt see a lot of these days.

Apollo KM133
The KM133 is another Via creation designed for Athlon and Duron-based PC\’s. It is an offshoot of the KT 133, but has integrated video, in the form of S3\’s Savage4 and Savage2000 graphics cores. It provides AGP 4X support and ATA-100.

Apollo KX133
The KX133 is Via\’s solution to the Athlon-based high-performance PC. Obviously, it supports the 200MHz EV-6 front-side bus common to AMD Athlon systems. It also supports AGP 4X, PC133 and UDMA/66.

Apollo Pro Plus
This is a very flexible version of the Apollo Pro chipset, one that is meant to serve a variety of different functions, including mobile use. It supports Slot 1 and Socket 370 processors. It\’s borther does not support the Socket 370 Celeron processors. It uses the same VT82C693 north bridge, but uses the newer VT82C596A south bridge which incorporates power management features.

Apollo Pro
The Apollo Pro is a popular chipset from Via that supports the Slot 1 processor. It comes in two configurations. (1) The energy-conscious version uses the VT82C691 north bridge with the VT82C596 south bridge. (2) high-performance users can use the VT82C691 with the VT82C586B south bridge. It offers support of PC100, UDMA/33, AGP2X and, of course, multiple memory configurations. It also supports using different kinds of memory, such as SDRAM, EDO, etc, so that users don\’t have to necessarily invest in new memory.

Post from: PCMech. Helping Normal People Get Their Geek On And Live The Digital Lifestyle.

Universal Serial Bus is a powerful external bus system for hooking devices to your PC and having ample bandwidth to perform the tasks today’s users are taking on. Originally developed in 1995, USB was developed by a collaboration of different companies in order to make connecting external devices to a PC as easy as plugging it in. Parallel and Serial ports got the job done, but installation was a bit more of a chore, and the bandwidth of these two interfaces was far from adequate for demanding environments.

USB not only solves the bandwidth issue, but also the expandability issue. Users of parallel and serial ports know that finding spare ports for hardware devices would be a pain in the rear if you already had the basic hardware like a mouse and printer. Add a few more peripherals to the mix, and manufacturers had to begin to devise work arounds like pass-through parallel ports for their scanners and ZIP drives. Serial ports were in use for modems, some mice, as well as PDAs and digital cameras. If a peripheral needed more bandwidth, it had to come with its own expansion card so that it could make use of the PCI bus. USB solves the problem by being expandable up to 127 devices, even through the use of USB hubs.

Today, USB is a widely popular interface and manufacturers all around the world are either developing their products to work exclusively over USB or are at least offering USB options. USB support is an afterthought today on all motherboards on the market. The ease of use, expandability and availability of the interface has led Intel and Microsoft to use USB as a key component of their Easy PC initiative.

How USB Works

USB is a plug-and-play external bus. It does this by requiring USB support in the underlying operating system (something all OSes in use by modern systems have). The host software then takes care of managing the attachment and de-attachment of perihperals while hiding all of the details of this from the system software. The phase at which a device is attached and identified is called enumeration (more below). Here, the system will communicate with the device and seek an indentification. It will then determine which driver to use for the system or if one needed to be installed. A unique address indentifer is assigned to each device for run-time ID and for data transfer. Each interaction with the peripheral is a transaction on the USB bus, the system originating and the perihperal responding appropriately.

Enumeration also identifies what type of data is to be transferred, or what mode of tranport the device will need:

• Interrupt - The device will just be assigned an interrupt on the bus. This will be used for mice, keyboards, things like that.


• Bulk - The mode used by devices who will be sending large blocks of data over the bus, like a printer.


• Isochronous - This mode is used for devices with a “live” connection, such as speakers. They need a direct line in and no error checking is necessary.

USB has the ability to be expanded with the use of hubs. The hub also has a role in the whole process. It manages the power delivered to each device and ensures each has adequate power during intialization. Since part of the USB standard allows for devices to get their power directly from their USB connection, if a device demands more power, the hub deals with the host PC software to ensure that it gets what it needs, delivering up to 2.5 W of power to each device. Obviously, peripherals requiring more power, like a printer or a scanner, could not be powered off of USB and will have their own AC adapters. Each hub is given its own ID on the bus, and hubs may be nested up to 5 levels deep. While sitting on the USB “network”, the hub will act as a bi-directional repeater, capturing transactions and repeating them onward to the target hub or device. It will catch all transactions addressed to itself and act accordingly.

The total bandwidth of the bus is 12 Mbps (1.5 Mbps for low-speed devices). During operating, the host is keeping track of the bandiwdth used by all enumerated devices. If total bandwidth comes up to 90% of the maximum, it will deny access to additional isochronous and interrupt devices. The extra 10% is used for bulk transfers and command packets (data packets carrying commands for one of the devices in the chain).

The USB bus has two types of connectors for peripherals, the “A” and “B” connector. Each is shaped differently so that users will never get confused what kind they should be using in a certain situation. The ports will be shaped in either A or B and the shapes are not interchangable, as you can see to the right. The wires for the USB connectors themselves contain two wires for delivering power to the peripherals (+5V and ground) as well as a twisted pair oif wires for actual data transfer. These wires are, of course, shieled with an overall sheath, with the connectors on either end.

The USB protocol defines that the bandwidth is to be split up into data frames. Each frame contains 1.5 bytes and a new frame starts each millisecond. Each frame gives space to the interrupt and isochronous transfers first so they always get bandwidth. The other devices consume the remaining space. The host, of course, controls this.

USB devices, also, are completely hot-swappable, meaning you can plug them in and unplug them without powering down the PC.

Post from: PCMech. Helping Normal People Get Their Geek On And Live The Digital Lifestyle.

Although Intel got into the Socket 7 chipset business and basically took over, there were many great chipsets out there that didn’t boast the Intel name. Also, Intel decided to move onto sixth generation chipsets after its release of the TX chipset. TX was their last Pentium-class chipset, leaving the market wide open for their competitors to come in and further the field. Companies like VIA, ALi and SIS all began producing inexpensive and high performing chip sets. These vendors were giving Intel a hard time since their chip sets offered Pentium II performance from the simpler, low-cost Socket 7 motherboards including AGP and 100 MHz bus. These advancements led to the evolution of the Socket 7 platform into what is called “Super 7″.

VIA

Via made the chipset that many consider the best non-Intel alternative you could buy. They had a reputation for being on top of the market and driving Intel to improve their technology, but Via chipsets did not hold much of the market at the time. Lets look at some common Via chipsets built for the Socket 7 platform:

• Apollo VP1 - the chipset that marked the entrance of Via into the mainstream market. Although now, VP1 is considered slow and meager, in the times of the 430FX, VP1 packed a punch. Via offered support for options such as EDO, BEDO and SDRAM as well as UltraDMA well before Intel ever got around to it. It lacks power saving features, but it was the design of this chipset that was used in the more popular VP2 as well as the VPX.
  
• Apollo VP2 - The VP2 hit the spotlight with a bang, easily hitting the top of the list for Socket 7 chipsets and bumping head to head with the Intel competition. The VP2 chipset combines the best features found in all Socket-7 chipsets, from BEDO and SDRAM support to UMA and UltraDMA support. VP2 offers support for 512 MB of RAM, all cacheable, as well as up to 2MB or L2 cache. The VP2 is indeed a powerful chipset. The only downer is that it does not support AGP. This support was offered in the VP3. The VP2 was used with all Pentiums, Cyrix 6×86’s, and AMD’s K5 and K6.
  
• VP3 - With Intel’s release of the 440LX chipset, which supports AGP, consumers were limited to the Slot 1 architecture if they wanted AGP. But, then, BAM! Via comes out with the first Socket 7 Pentium chipset that supported AGP, the VP3. Combine the support of the VP2 with larger CPU-DRAM write buffers, support for 1 gig of cacheable RAM, and AGP support, and you’ve got yourself a VP3 chipset. While it indeed was a powerful chipset, it’s biggest letdown was that it did not support the 100MHz bus speed yet.
  
• MVP3 - the Mobile VP3, really a misnomer because it was used in both notebooks and desktops. This chipset offered many important advancements over the original VP3, the most important being the 100MHz frontside bus support. One of the more interesting features of the MVP3 chipset is the ability to run the memory clock at 66MHz while running the rest of the system at 100MHz. This asynchronous bus design gives the user the ability to run at 100MHz frontside bus while still using some of that old EDO or SDRAM lying around.
  
• VXPro - A chipset for the affordable, entry-level system. It was mostly found on cheap motherboards such as those by PCchips. The VXpro supported UltraDMA and SDRAM, although it did not offer the fast memory timings. The VXPro was much like the original VP1, but it was cheaply built. It only supported 64MB of cacheable memory, like the 430TX. Really, the only good thing about this chipset was the price. 

Post from: PCMech. Helping Normal People Get Their Geek On And Live The Digital Lifestyle.