Section 5.3. Component Considerations

5.3. Component Considerations

With our design criteria in mind, we set out to choose the best components for the gaming PC system. We took advice from our readers, because Barbara doesn't game at all and Tux Racer is Robert's idea of a challenging game. The following sections describe the components we chose, and why we chose them.

Advice from Jim Cooley

Cases made more cheaply than Antec are liable to use thinner steel or aluminum in the frame and side panels, so if portability is an issue double-check the case integrity. I've seen cases mangled and bent out of shape from a drop of less than two feet, something that wouldn't happen with an Antec.

5.3.1. Case

Antec P150 (http://www.antec.com)

We could have built the gaming PC in just about any case. For the LAN party PC we built for the previous edition of this book, we chose the aluminum Antec Super LANBOY for portability.

But portability turned out to be less important than we'd thought. When we went on a road trip to visit friends over the Labor Day weekend, Robert didn't think about the LAN party system. He took along his primary office desktop system instead. That system was a loaded mid-tower box, with a dual-core processor, 2 GB of memory, dual optical drives, and a 2.5 TB of hard disk space. It was much more capable than the LAN party PC, and, as it turned out, just as portable. So we decided to build our gaming PC in a standard mini- or mid-tower case, confident that it would be as easily portable as a system built in a special LAN party case.

With a blank sheet of paper, so to speak, we set out to choose the perfect case for our gaming PC. We wanted a case that was attractive, provided excellent cooling, and was as quiet as possible. After considering numerous alternatives, we settled on the Antec P150. The P150 uses a gloss white and brushed aluminum color scheme that looks good anywhere. Mike Chin of Silent PC Review (http://www.silentpcreview.com) helped Antec design the P150, so we were confident that it would provide excellent cooling at a very low noise level.

ALTERNATIVES: CASE

Nearly any mini- or mid-tower case. If the case includes a power supply, make sure that it is ATX v2.x-compliant and of sufficient wattage to support your configuration. If you plan to use dual video adapters, make sure the power supply is SLI-certified. If the case does not include a power supply, purchase a high-quality unit such as the Antec NeoHE separately. For SLI systems, we recommend the Antec NeoHE 550.

The P150 includes an Antec NeoHE 430 power supply, which is a premium unit. The only downside to the P150 is that the NeoHE 430 is not SLI-certified for use with dual video adapters. That wasn't a problem for us, because we don't intend to use dual video adapters, and the NeoHE 430 is otherwise a superb choice. If you plan to use SLI (or CrossFire), choose a case that doesn't include a power supply or one that includes an SLI-certified power supply. Otherwise, we think you'll also be delighted with the P150.

5.3.2. Processor

AMD Athlon 64 X2 4200+ (http://www.amd.com)

Intel's July 2006 introduction of its new-generation Core 2 Duo processor line wreaked havoc with processor pricing. To make room in the price list for the Core 2 Duo models, Intel took a meat-ax to the prices of their older Pentium D dual-core processors. Literally overnight, the price of performance was cut in half. AMD had no choice but to respond with similar price cuts, chopping the price of some of their Athlon 64 models by more than 60%.

In the past, a serious gaming system typically used a processor that cost $350 to $1,000, versus the $150 to $225 cost of mainstream processors. The most rabid gamers happily coughed up $1,000 for "extreme" processors like Athlon FX-series models. Happily, the processor price war kicked off by Intel means that even a serious gaming system can now use a $200 processor.

We considered three processor families for our gaming system, the Intel Pentium D, the Intel Core 2 Duo, and the AMD Athlon 64 X2. All of these are dual-core processors. Although dual core is of limited benefit for most games, there's really no alternative. Nowadays, only "value" processors are single core. Here are the issues we considered.

Intel Pentium D

The Intel Pentium D doesn't get much respect as a gaming processor, but that has more to do with pricing than any real problems with the processor itself. Intel had priced the Pentium D against the AMD Athlon 64 X2 for comparable general performance at a comparable price. But the Athlon 64 X2 architecture is more efficient for gaming. That meant that, for similar gaming performance, a less expensive Athlon 64 X2 matched a significantly more expensive Pentium D. Gamers abandoned Intel in droves. When Intel cut Pentium D prices dramatically, the equation shifted in favor of the Pentium D, dollar for dollar. Even AMD's July 2006 price cuts on the Athlon 64 X2 didn't completely close the gap, so the Pentium D suddenly became an excellent choice for a gaming system, particularly for a gamer on a tight budget.

Intel Core 2 Duo

The first benchmark tests on the Core 2 Duo made it clear that this new processor family simply blew the doors off the older Pentium D and Athlon 64 X2. That might not have mattered if Intel had priced the Core 2 Duo models as premium products, but they didn't. Intel wanted to transition to the Core 2 Duo as quickly as possible, so they priced the Core 2 Duo models extremely aggressively. How aggressively? At introduction, the $316 Core 2 Duo E6600 matched or beat the overall performance of the $1,000+ AMD Athlon 64 FX-62, and even the entry-level $224 Core 2 Duo E6300 outpaced all but the fastest AMD processors. The Core 2 Duo also shifted the playing field in terms of gaming performance, with benchmark tests showing that Core 2 Duo was at least on a par with AMD's best. Core 2 Duo is a superb choice for a gaming system. But not the only superb choice, as we found.

AMD Athlon 64 X2

The introduction of the Intel Core 2 Duo knocked AMD down, but not out. At the very highest reaches of gaming performance, the AMD Athlon 64 X2 was no longer competitive with Intel Core 2 Duo, but very few systems, even dedicated gaming systems, use extreme processors. In the sweet spot for gaming processors, AMD could again be competitive with Intel simply by reducing its prices for the Athlon 64 X2, and that is exactly what they did. Suddenly, a fast X2, which not long before had sold for $400, now cost $200, and AMD was right back in the ballgame.

Athlon 64 X2 Power Consumption

AMD produces several Athlon 64 X2 models in two variants. The model designations are identical, but the more expensive variants use noticeably less power and produce less heat. At the time we built this system, the low-power version of the Athlon 64 X2 4200+ sold at a $60 premium. We didn't consider the benefit worth the cost, so we used the standard version. As time passes, we expect that AMD will improve its production processes and eventually begin producing only the low-power variants.

We were faced with making the choice among three excellent processor families. Talk about an embarrassment of riches. For $125, we could have chosen a Pentium D that offered sufficient performance for a mainstream gaming system. For $50 or so more, we could choose a Core 2 Duo or a fast Athlon 64 X2, either of which had noticeably better gaming performance.

If we had been on a tight budget, we wouldn't have hesitated to use the Pentium D, despite its somewhat lower performance and higher heat. We might have chosen the Core 2 Duo for our gaming PC except that availability was tightly constrained soon after its introduction, with Intel shipping every available Core 2 Duo processor to Dell and other large OEMs. Fortunately, the AMD Athlon 64 X2 was widely available, and provided excellent bang for the buck.

For our gaming PC, we chose the AMD Athlon 64 X2 4200+ in the new Socket AM2, which at the time was selling for about $185. Admittedly, it seemed odd to pay only a mainstream price for a high-performance processor like the 4200+, but we could get used to that. The simple truth is that any but the most extreme gaming PC no longer needs a $350+ processor. The Athlon 64 X2 4200+ is just as fast priced at $185 as it was when it sold for twice that much, and fast enough is fast enough.

We chose a retail-boxed Athlon 64 X2 4200+, which includes a decent CPU cooler. The stock AMD cooler appears to be an AVC Z7U7414001 heatpipe model, which is reasonably efficient and quiet. If you prefer to use a quieter or more efficient cooler, consider premium models from Thermalright (Ultra-90/K8, XP-90, or XP-120) or Zalman (CNPS7xxx- or 9xxx-series). Socket AM2 uses the same mounting arrangements for the CPU cooler as Socket 939, so nearly any Socket 939 cooler that fits your motherboard and case should work properly with a Socket AM2 processor.

5.3.3. Motherboard

ASUS M2N32-SLI Deluxe (http://www.asus.com)

Like Intel, AMD made a major change to its processor line in mid-2006 by introducing a new socket that will eventually replace Socket 754 and Socket 939. The announcement of Socket AM2 also presaged AMD's shift from DDR memory to DDR2 memory. Unlike Intel's shift from Netburst architecture to Core 2 architecture, AMD's shift from Socket 939 to Socket AM2 had little impact on performance. A Socket AM2 Athlon 64 processor with DDR2 memory is little or no faster than the same model for Socket 939 with DDR memory, and AMD makes no claims of increased performance.

But the processor and memory are not the sole determinants of performance. The motherboard and, more particularly, the chipset, can have a dramatic impact on system performance. Years ago, it was common for benchmark tests for a particular processor model to differ widely depending on the motherboard and chipset used for testing. Around the time Intel introduced the 440BX chipset for the Pentium III, those differences began to disappear. In recent years, the performance differences between motherboards and chipsets had become relatively minor, at most a few percent either way.

ALTERNATIVES: MOTHERBOARD

For an Intel Pentium D processor, choose any compatible Intel or ASUS motherboard based on an Intel 946-, 963-, 965-, or 975X-series chipset that provides a PCI Express video adapter slot. For an Intel Core 2 Duo processor, choose any compatible Intel or ASUS motherboard based on a 946-, 963-, 965-, or 975X-series chipset that provides a PCI Express video adapter slot. For a less expensive alternative to the ASUS M2N32-SLI for a Socket AM2 Athlon 64 X2 processor, choose the ASUS M2N-E or M2N-SLI Deluxe. Based on our performance testing, we do not recommend using a Socket 939 processor.

nVIDIA changed that when they introduced the nForce 590 SLI chipset. nVIDIA nForce chipsets had always been the premium choice for AMD processors, offering top-notch performance and stability. We expected nVIDIA's chipset for Socket AM2 to be more of the samefast and reliable, and little different from their nForce 4 series chipsets for Socket 939. Boy, were we wrong.

We benchmarked AMD Athlon 64 X2 4200+ processors in Socket 939 and Socket AM2 on similar nForce motherboards, expecting similar results. We were surprised to find that the Socket AM2 processor on an nForce 590 motherboard was faster than the Socket 939 processor on an nForce 4 motherboard. And not just a little faster. In some benchmarks, the nForce 590 SLI motherboard was as much as 30% faster, which is an incredible difference for a chipset to make. To put this in perspective, the $175 Athlon 64 X2 4200+ in an nForce 590 SLI motherboard gave faster performance benchmarks than we'd expect to see from an Athlon 64 FX-62 processor in an nForce 4 motherboard. nVIDIA really hit a home run with the nForce 590 SLI chipset.

Perhaps that performance difference isn't all attributable to the chipset. The motherboard itself may have something to do with it. At the time we built our gaming PC, Socket AM2 motherboards were still pretty thin on the ground. We were fortunate enough to get our hands on an ASUS M2N32-SLI Deluxe motherboard, which is what we used for benchmark-testing the Socket AM2 processor. We've always sworn by ASUS motherboards for their quality, performance, and reliability, so on that basis we chose the top-of-the-line ASUS M2N32-SLI Deluxe motherboard as the foundation of our gaming PC.

5.3.4. Memory

Kingston KVR667D2N5K2/2G PC2 5300 DDR2-SDRAM (1 GB x 2) (http://www.kingston.com)

If you want a stable system, install a premium power supply and top-quality memory. Using cheap memory almost guarantees frequent system crashes. We've used premium, name-brand memory in all of our systems for more than 20 years, and it has seldom let us down.

When determining memory requirements, it's important to remember that a dual-core processor like our AMD Athlon 64 X2 4200+ is effectively two processors. Each of those processors needs as much memory as a single-core processor does. We consider 1 GB the sweet spot for gaming on a single-core processor, so we decided to install 2 GB of memory in our dual-core gaming PC. Because the AMD Athlon 64 X2 4200+ has a dual-channel DDR2 memory controller, we decided to install two 1 GB memory modules.

As we always do when we're configuring a new system, we visited the Kingston and Crucial web sites and used their configurators to display lists of memory modules compatible with our ASUS M2N32-SLI Deluxe motherboard. As it happened, Kingston memory was a bit less expensive than comparable Crucial memory that day, so we opted for the Kingston modules.

ALTERNATIVES: MEMORY

Any compatible premium memory modules. We recommend a minimum of 2 GB of memory for a Gaming PC, and we recommend using only Kingston or Crucial modules.

Kingston listed three compatible 2 GB memory kits, each of which contained two matched 1 GB modules. The only differences in those kits were the speed and CAS latency (CL) of the modules and their price. At the time we built this system, a PC2 5300 CL5 memory kit sold for $161. We could instead have chosen a PC2 4200 CL4 kit that had slower access time but faster latency for about $190. That kit would have offered somewhat faster random memory access at the expense of slower sequential memory access, so we ruled out the CL4 kit. The third alternative was a 2 GB PC2 6400 CL5 kit for $346. That kit would have offered faster sequential memory access than the PC2 5300 kit, but at more than twice the cost. The PC2 6400 kit would have boosted performance by at most a few percent. That gain wouldn't be perceptible other than when running memory benchmark tests, so we decided to use the less expensive PC2 5300 kit.

TECH HELP
Why "Performance" Memory Usually Isn't Worth Paying Extra For

Companies like Corsair and Mushkin sell "high-performance" memory to the enthusiast market. We're sometimes asked if it's worth paying more for such memory rather than using standard Kingston or Crucial modules. The short answer is that it's usually not.

Even nominally identical memory chips vary from one to the next. Some are faster than others, and performance memory packagers take advantage of that fact. They order large numbers of memory chips and use a process called binning to hand-select the fastest chips from that batch. After they've cherry-picked the fastest 5% or 10%, they resell the remaining chips to other memory packagers. They assemble those hand-picked chips into high-performance modules and test the finished modules to verify that they function at higher speeds and tighter memory timings than standard memory.

Many gamers happily pay substantial premiums for such memory, on the assumption that faster memory must translate to faster system performance. Alas, that's not necessarily true. If one type of memory is fast enough to keep up with the processor, or nearly so, substituting faster memory has very little effect on overall system performance.

Some might object that the benchmarks show the difference. Sure they do, when they test memory subsystem performance in isolation. But memory performance is only one aspect of overall system performance, and using faster memory helps only if memory speed is the bottleneck. For most gaming systems, it is not.

The one exception is overclocked systems. If you boost the bus speed to run your CPU at higher than nominal speed, which we do not recommend, you're also pushing other system components, including the memory, to speeds they were not designed to support. In such cases, it's a good idea to use hand-picked performance memory rather than depend on the tolerances built into standard memory modules.

5.3.5. Video Adapter

Pick one (or two)

Choosing a video adapter (or adapters) for a gaming system is the most complex decision you'll have to make. You have to weigh your budget against the minimum level of 3D graphics performance that is acceptable to you. You have to take into account the specific games you play, because some games are faster on nVIDIA adapters and others on ATi adapters. You have to decide whether it's better to install one expensive adapter or two midrange adapters. You have to weigh the advantages and drawbacks of buying an expensive adapter now and using it for a year versus installing a less expensive adapter now and upgrading every six months.

For our own configuration, we had no need of a new video adapter. Instead, we migrated a year-old nVIDIA GeForce 6800 Ultra from an older system to our new gaming PC. That formerly high-end card is now midrange in terms of performance against current models, but it's still more than fast enough for the games we play. We could have matched its performance with a current model that sold for $125, but there was no point to doing that. Instead, we'll wait until we actually need a faster graphics adapter, and then upgrade to what by then will probably be a midrange model.

When you choose a graphics adapter for your own gaming PC, we suggest using the following guidelines:

• Video adapters change in Internet time. Get the latest information and benchmarks from enthusiast sites such as AnandTech (http://www.anandtech.com), Sharky Extreme (http://www.sharkyextreme.com), and Tom's Hardware (http://www.tomshardware.com).

• ATi versus nVIDIA is a religious issue. Both companies produce excellent video chipsets, and overall performance is comparable between similarly priced adapters that use either company's chipsets. That said, we prefer to use nVIDIA adapters on motherboards that use nVIDIA system chipsets. Because we prefer nVIDIA-based motherboards for AMD systems, we generally use nVIDIA-based video adapters in AMD systems. For Intel-based systems, we've historically used mostly ATi adapters, although nVIDIA adapters work just as well. With AMD's buyout of ATi and the introduction of Intel's new-generation Core 2 Duo processors and new chipsets to support them, the graphics landscape will change significantly in late 2006 and into 2007. Only experience will tell which combinations are optimum in this new environment.

• Make sure the adapter you choose has sufficient onboard memory for the games you play. For casual gaming, particularly with older titles, 128 MB may suffice. A mainstream gaming adapter should have 256 MB, and if you play the latest, most intense games, you'll want 512 MB or more.

• Pay close attention to performance with the specific games you play. Some games play better on nVIDIA adapters, and others on ATi adapters. In the most extreme cases, a particular game may be faster on a midrange adapter from nVIDIA than on a high-end adapter from ATi, or vice versa.

• Unless your budget is effectively unlimited, give careful thought to your upgrade strategy. Quite often, you're better off upgrading every six months to the latest mid-range adapter than spending a lot of money on a high-end adapter initially and having to use it for a year or more.

• No matter which card or cards you install, play close attention to driver updates. ATi and nVIDIA both release driver updates frequently. Those updates may fix bugs, but often they are primarily performance tweaks for the most recent games. The performance delta between the current driver and an old version can be extraordinary.

• Consider carefully before you buy into the dual adapter concept. In theory, nVIDIA's SLI (Scalable Link Interface) and ATi's CrossFire are very attractive. You can install two less expensive adapters instead of one more expensive adapter, and get higher performance for less money. In practice, dual adapters may not work particularly well. Driver problems are common, and performance is not always as high as expected. If you do decide to use dual adapters, verify everything carefully. In particular, make absolutely certain that your motherboard is compatible with the adapters you choose and that the power supply you use is SLI-certified and can provide the required current to both adapters.

• Decide what is reasonable to spend, and then limit yourself to that amount. If you set a $150 budget, don't let the marketing hype convince you to walk out of the store with a $300 adapter. The 80/20 Rule definitely applies to graphics adapters. A $150 adapter provides 80% of the performance of a $750 model, give or take, and all but the most avid gamers will probably be happy with the performance of that $150 adapter.

• If you do decide to buy a high-end adapter, be aware that at the very high end the 80/20 Rule is replaced by the Law of Diminishing Returns. For example, a $750 adapter costs 50% more than a $500 adapter, but may be only 10% (or less) faster. Super-premium adapters are more often bought for bragging rights than for any perceptible performance benefit.

5.3.6. Sound Adapter

Integrated

When we sat down to design this system, we fully intended to install a standalone sound adapter. Then we started reading reviews of the ASUS M2N32-SLI Deluxe motherboard and its integrated ADI SoundMax HD Audio, and decided just to use the integrated audio.

Gamers install standalone sound adapters for two reasons. First, integrated audio can put a heavy burden on the main system processor, resulting in lower frame rates. Second, integrated audio typically lacks hardware acceleration for positional audio.

The integrated ADI codec answers the first objection easily. Frame rates with ADI audio enabled are only 2% to 3% lower than frame rates with ADI audio disabled (versus a 25% to 33% drop with many integrated audio solutions) and the audio quality is excellent. Unfortunately, ADI accomplished this feat by not including support for Creative's EAX positional audio. We don't care about EAX support, so the integrated audio is sufficient for us.

ALTERNATIVES: SOUND ADAPTER

If you need EAX support, the obvious choices are a Creative Labs Audigy 2 ZS, Audigy 4, or X-Fi sound card, and all of those are popular choices among gamers. Less obvious choices are the M-AUDIO Revolution 5.1 or 7.1. The Creative cards have somewhat superior gaming support, but at the expense of generally lower sound quality and more frequent driver problems. The M-AUDIO cards have decent gaming support, better audio quality, and generally fewer driver problems. If we wanted a standalone sound card, we'd install one of the M-AUDIO cards.

5.3.7. Hard Disk Drive

Seagate 7200.9 Barracuda SATA (500 GB x 2) (http://www.seagate.com)

We've used and recommended Seagate hard drives for many years, and have never had cause to regret it. Seagate Barracuda-series drives are fast, extremely reliable, quiet, and reasonably priced. So, when we configured our gaming PC, the only real decisions were which model of Barracuda drive to install, and how many.

Several of our gaming friends convinced us to try a RAID 0 once again, swearing that RAID 0 provided better gaming performance. We didn't really believe that. We've tested RAID 0 (and the similar RAID 0+1) many times over the years. While RAID 0/0+1 indeed boosts disk performance on a heavily loaded server, we've never seen much performance benefit on desktop systems. Weighed against the riskone drive failing in a RAID 0 causes all data on both drives to be lostusing RAID 0 on our gaming PC seemed like a sucker bet. Still, why not? It was worth testing again, and if the tests turned out as we expected, we could simply pull the second drive and use it in another system.

ALTERNATIVES: HARD DISK DRIVE

We chose the 500 GB Barracuda. That may be overkill for you, depending on your practices. (Many gamers fill their hard drives with rips of numerous DVD video and game discs.) If you don't need that much disk space, use a Barracuda of smaller capacity.

ADVICE FROM BRIAN BILBREY

In transitioning from one texture map to another, load times are significant. I might expect to see much better performance from hardware RAID 0 than from a JBOD configuration. Software RAID, not so much.

RIGHT AGAIN

The tests turned out just as we expected. By choosing benchmarks that exercised the disk subsystem heavily, we could "prove" that RAID 0 gave better performance than a single drive. But those benchmarks hammered the drives with access patterns typical for a file server, not for a single-user desktop system. For real-world single-user use, including gaming, RAID 0 offered very little performance benefit. So, although we illustrate the build process using two drives, you can safely install just one drive in your own gaming PC.

The same objection holds true for using faster drives. We tried a Western Digital Raptor 10,000 RPM ATA drive and Seagate Cheetah 15,000 RPM SCSI drives, alone and in a RAID 0. While the faster drives made program loading faster and heavy disk operations noticeably snappier, they had little perceptible effect on game play once the game was loaded.

5.3.8. Optical Drive

BenQ DW1650 DVD writer (http://www.benq.us)

Many gamers install two optical drives in their systemsa DVD writer for general use and a DVD-ROM drive for faster read access to game DVDs. Fast DVD writers, such as the BenQ DW1650 model we chose for our gaming PC, make this strategy obsolete. In our testing, the BenQ DW1650 proved to have true random access times as fast as or faster than the several DVD-ROM models we also tested.

WHEN SLOWER IS FASTER

Take published random access times with a grain of salt. There are many ways to test random access time, and they are not standardized. A drive with a published 120 ms random access time may actually be faster than another drive that claims 85 ms random access. Other than testing specific drives with a utility like Nero CD-DVD Speed, there's no way to tell which claims are conservative and which are exaggerated.

For example, we tested the NEC ND-3550A (rated at 140 ms random access) and indeed found it to be slightly slower than the BenQ DW1650 (120 ms). However, the Plextor PX-740A and PX-716AL, both rated at 150 ms, were faster than the BenQ DW1650. We tested several DVD-ROM drives, with rated access times ranging from 85 ms to 120 ms, and found no correlation between rated access times and actual performance. One of the DVD-ROM models listed at 85 ms was the slowest drive we tested, and the 150 ms Plextor DVD writers were among the fastest.

ALTERNATIVES: OPTICAL DRIVE

Any DVD writer from BenQ, Lite-On, NEC, Pioneer, or Plextor. If you use DVD-RAM discs, make sure to choose a model with DVD-RAM support. If you decide to install a DVD-ROM drive in addition to or instead of a DVD writer, choose a current model from ASUS, Lite-On, Pioneer, Samsung, or Teac.

Advice from Brian Bilbrey

DVD for faster read access...I don't agree. I always load the whole game onto disk. The only need for the DVD is to validate that I am allowed to run the game.

5.3.9. Mouse and Keyboard

If you plan to take your gaming PC on the road, you'll probably want two mice and two keyboards, one set for home and one for away.

5.3.9.1. Home mouse and keyboard

Logitech cordless mouse/corded keyboard (http://www.logitech.com)

At home, we use cordless mice and corded keyboards on most systems. We have tested numerous cordless mice from Logitech and Microsoft. We much prefer the design and feel of the Logitech models, particularly those that include a receiver that doubles as a recharging cradle. Logitech cordless keyboards have reasonably long battery life, but we use them only when cordlessness is really important, because swapping cordless keyboard batteries is a pain in the begonia.

We particularly dislike recent Microsoft cordless mice, whose scroll wheels make it too easy to click accidentally while scrolling. Also, although battery life in the recent Microsoft cordless mice we have used is reasonably good, keyboard battery life is terrible. (In one test, we exhausted a fresh set of three AA heavy-duty alkaline batteries in less than one day of constant use.)

As to a corded keyboard, choose whichever model feels best to you. Again, we generally prefer the design and feel of the Logitech models, but Microsoft also offers several very good corded keyboards.

5.3.9.2. Away mouse and keyboard

Logitech MX-series corded optical mouse (http://www.logitech.com) Zippy EL-715 illuminated keyboard (http://www.zippy.com.tw)

We love the freedom of a cordless mouse for home use, but for away use corded devices are almost mandatory. Cordless keyboards and mice use radio frequency (RF) communications. Some provide an A-B switch to prevent conflicts by using different frequencies, but a choice of only two frequencies is wholly inadequate at a crowded LAN party. If you doubt that, just wait until the first time you watch the cursor moving across your screen in response to someone else's mouse movements. Until someone comes up with frequency-agile, stealthed cordless input devices, corded it is.

For a corded mouse, we prefer a Logitech MX-series optical mouse. The $25 six-button MX 310 can be used with either hand, and is precise enough for most gamers. The $35 eight-button MX 518 offers higher tracking resolution and is specifically designed for gaming (although it's also excellent for general use). The MX 518 is available only for right-handers.

Although your regular corded keyboard may serve for away use, it's worth considering a specialty keyboard designed for use on the road, particularly at LAN parties. Our favorite travel keyboard is the Zippy EL-715. The EL-715 is a medium-size, notebook-style, 105-key keyboard with electroluminescent backlighting that is quite useful at dimly-lit LAN parties. At just over a pound and about 0.75" thick, the EL-715 is extremely portable. We find it a bit too cramped to use as our primary keyboard, but it's unsurpassed as a portable LAN party keyboard. The blue backlighting is bright enough to see what you're doing, but not so bright that it becomes intrusive in a dimly lit environment.

ALTERNATIVES: MOUSE AND KEYBOARD

Thousands. Well, hundreds anyway. Personal preference is the most important factor in choosing a keyboard, mouse, game controller, display, or other I/O peripheral. What we hate you may love, and vice versa. For a cordless or corded mouse or keyboard, we suggest you try a Logitech model first. If you don't like it, exchange it for a similar model from Microsoft. If you want the best gaming mouse and are willing to pay a premium price, the $45 Logitech G5 Laser Mouse is the best choice we know of.

5.3.10. Game Controllers

Choose your own

As with keyboards and mice, personal preference is the most important factor in selecting the best game controllers for your own needs. Cordless controllers are very nice for use at home or small LAN parties, but if you attend large LAN parties corded is the only way to go.

MAY THE FORCE BE WITH YOU

One downside to cordless models is that none we know of incorporates full force feedback, which apparently requires too much juice to run from batteries. The best cordless controllers can provide is limited vibration feedback. Full force feedback uses small motors within the game controller to provide tactile response to gaming actions. For example, as you maneuver your F-16 Falcon onto the six of a MiG-29 Fulcrum and begin hosing him down with your Vulcan rotary cannon, a force feedback joystick jitters and jerks to simulate recoil. Games that implement force feedback well are much more immersive than games that do not.

The first step is to decide which type or types of game controller you need, which is determined by the types of games you play. For example, we play mostly flight simulation and air combat games, so our primary controller is a joystick. For first-person shooter (FPS) games, most people consider a gamepad to be the optimum controller. For racing games, you'll want a wheel game controller. Although different types of controllers can substitute for each other to some extent, using an inappropriate game controller can put you at a severe disadvantage relative to players who are using controllers better suited to the type of game being played.

ALTERNATIVES: GAME CONTROLLERS

Many, some of which you'll love and some you'll hate. The trick is to figure out which is which before you pay for them. If you have no idea which controller(s) to buy, buy something inexpensive to start with and then try as many midrange and high-end controllers as you can get your hands on.

Once you determine which types of game controller you need, decide how to allocate your budget among them. If you regularly play games that require all three types of game controller, allocate your budget evenhandedly among them. That doesn't mean spending a third of your budget on each, because different types of controllers have different price points. It does mean you should buy all three high-end, all three midrange, or all three low-end, depending on your budget. When you play a racing game, for example, a $70 wheel controller is better than a $250 joystick. On the other hand, if you spend most of your time playing flight sims, play FPS games occasionally, and never play racing games, put most of your budget into the joystick, and spend whatever is left on a decent gamepad. And so on.

Before you buy a game controller, see if you can play with someone else's for at least a short session. If you ask nicely, most gamers are happy to let you try their rigs. Of course, the flipside is that if you do buy that $250 joystick, you can expect to be very popular at LAN parties.

ALTERNATIVES: SPEAKERS/HEADPHONES

If the $275 Z-5500 speaker set is a bit rich for your taste, consider the $125 Logitech Z-5300e 5.1 set. The Z-5300e has a combined RMS output of "only" 280W, but the sound quality is comparable to its more expensive sibling. If the Z-5300e is still beyond your budget, consider the $50 Logitech X-530 5.1 set. The combined output is a modest 70W RMS, and the sound quality is noticeably inferior to the better Logitech sets, but still reasonably good. (Ordinarily, we don't recommend inexpensive 5.1 speaker sets, but for gaming, support for surround sound deserves priority even at the expense of sound quality.)

We've found nothing we like nearly as much as Zalman ZM-RSF6F headphones for gaming. The bad news is that these are not general-purpose headphones. The bass is weak, although for gaming that is more than made up for by the excellent 3D imaging. Used for music and other stereo sources or for DVDs, the Zalman headphones boom and echo pretty badly. For general-purpose listening, buy a decent set of Grado or Sennheiser headphones.

5.3.11. Speakers and Headphones

Logitech Z-5500 5.1 speaker system (http://www.logitech.com): Zalman ZM-RS6F 5.1 surround headphones (http://www.zalman.co.kr)

A gaming system needs good, high-power speakers, and the Logitech Z-5500 5.1 speaker system is the best we know of for the purpose. With four 62W RMS satellite speakers, a 69W center-channel speaker, and a 188W subwoofer, the Z-5500 produces a wall-rattling 505W RMS. You don't just hear the bass, you feel the bass vibrating your internal organs. Nor is the Z-680 limited to gaming. At lower volume, it's also excellent for anything from listening to background music to playing DVDs.

As nice as the Logitech Z-5500 is, it's overkill for a road trip, not to mention for playing games at home while the spousal unit is trying to sleep. For those situations, you need headphones, and the Zalman ZM-RSF6F 5.1 surround headphones are the best we know of for gaming. They're light, durable enough to stand up to LAN party use, fold up into a self-contained unit, and their 3-meter cord gives you plenty of slack to move around.

The Zalman headphones use three separate drivers per ear to produce a surround sound field comparable to that provided by a 5.1 speaker system. The importance of positional audio may not be apparent at first glance, but it can mean the difference between winning and losing. When you play an FPS, knowing where your opponents are is critical. Stereo headphones just don't cut it. By the time you figure out where shots are coming from, you're dead meat. The Zalman headphones make it easy to discriminate not just left and right but front and rear. The difference is amazing.

5.3.12. Display

Samsung 930BF 19" FPD (http://www.samsung.com)

In the first edition of this book, we specified two displays. For road trips, a CRT monitor was simply too bulky and heavy to take along. On the other hand, the flat-panel LCD displays of the time were too slow for gaming. So we compromised, specifying a good CRT for use at home and a light, easily portable LCD display to take on the road.

Nowadays, although CRT monitors still have a few advantagesprimarily low price and off-axis color fidelitythere's no real reason not to use an LCD display as the only display for a gaming system. Slow LCD display response time and the associated ghosting were the main problem with early LCD displays. Many current LCD display models are more than fast enough to use for gaming. Robert uses a 19" Samsung 930BF on his own system.

ALTERNATIVES: DISPLAY

Any 17" or larger LCD display from NEC, Samsung, or ViewSonic that has response time fast enough to suit you. For most gaming, a 12ms BWB response time is adequate, and 8ms or faster ideal. For a 17" model, either VGA (analog) or DVI/HDMI (digital) input is sufficient. For 19" and larger models, we recommend digital input. (Make sure your video adapter provides an output that's compatible with the display you choose.)

Some models are available with built-in speakers. Although their sound quality isn't as good as standalone speakers, many LAN partiers choose one of these models to eliminate the need to carry separate speakers. You can always use the built-in speakers on the road and a better set of speakers at home. We prefer to use standalone speakers at home and headphones on the road.

Table 5-1 summarizes our component choices for the gaming PC.

Table 5-1. Bill of materials for gaming PC

Component	Product
Case	Antec P150
Power supply	Antec NeoHE 430 (bundled with case)
Motherboard	ASUS M2N32-SLI Deluxe
Processor	AMD Athlon 64 X2 4200+
CPU cooler	AMD (bundled with processor)
Memory	Kingston KVR667D2N5K2/2G (two 1 GB PC2 5300 DDR2 DIMMs)
Video adapter	(See text)
Sound adapter	(Integrated)
Hard drive	Seagate Barracuda SATA 7200.9 (two 500 GB in RAID 0)
Optical drive	BenQ DW1650 DVD writer
Home mouse and keyboard	Logitech cordless mouse/corded keyboard
Away mouse	Logitech MX-series corded optical mouse
Away keyboard	Zippy EL-715 illuminated keyboard
Game controller(s)	(See text)
Speakers	Logitech Z-5500 5.1 speaker system
Headphones	Zalman ZM-RS6F surround headphones
Display	Samsung 930BF 19" LCD