At the “heart” of every computer lies the CPU, or central processing unit, which is responsible for carrying out arithmetic and logic functions as well as executing instructions to other components. The components of a CPU work together, and depending on how they are made, determine exactly how fast these operations can be carried out along with how complex the operations can be. Each of the separate components of a CPU on their own are relatively simple. Some of the primary components of a CPU, also known as a microprocessor, are the arithmetic logic unit (ALU), the control unit and the registers

To begin with, the arithmetic logic unit is the part of the CPU that, as its name implies, carries out the mathematical functions of addition, subtraction, multiplication and division. It is often thought that these functions are all the CPU does in a computer, but this is not true. The ALU works along with, and as a major part of, the other components of a CPU to run many complex processes. A CPU can contain more than one arithmetic logic unit, and these ALUs can also be used for the purpose of maintaining timers that help run the computer.

The control unit is another fundamental part of the CPU. Essentially, it regulates the flow of information through the processor. The functions that a control unit performs can vary based on what a particular CPU was built to do. Mostly, this component receives, decodes, stores results and manages execution of data that flows through the CPU. More complex control units need to schedule when and how this great amount of information is to be processed and make sure that the data is sent to the correct components of the computer.

More components of a CPU that are vital to its operation are the registers, which are very small memory locations that are responsible for holding the data that is to be processed. The most important of these registers is known as the instruction pointer, which directs the CPU to the next memory location from where it is to receive information. Another type of register is the accumulator, which is responsible for storing the next values that will be processed by the CPU. Together all of these components of a CPU are becoming faster, more compact and more powerful as time goes on and technology advances.

source: http://www.wisegeek.com/what-are-the-components-of-a-cpu.htm

If the processor market was a game of King of the Hill…no, that’s too easy.

The fact of the matter is that Intel currently sells the fastest CPUs. Its quad-core Core i5 and Core i7s are unmatched in the desktop space—and that’s precisely the reason you pay more for them than the dual-core, Clarkdale-based Core i3 and Core i5 processors (not to mention AMD’s entire Phenom II lineup, including its flagship X4 965 Black Edition).

Of course, “being the fastest” assumes that you’re looking at the right benchmarks. Intel’s dominance is most evident in video encoding apps, threaded compression/decompression workloads, and content creation tools like 3ds Max. If you’re gaming, AMD is perfectly capable of standing right up to Intel.

Now, sure, you’ll see those folks who run their benchmarks at 640×480 in an effort to demonstrate the differences between one processor and another. Truth be told, even as high as 1920×1200, there are measurable performance gaps between CPUs. But at the end of the day, your graphics card, more than anything, determines how well your games run.

Why the rambling sidebar on gaming? Because Intel sees its brand new Core i7-980X Extreme Edition—previously referred to as Gulftown—as a sweet gaming processor. And it will become the fastest processor you can buy (it’s technically not available yet), so it’d naturally be great in a gaming box. However, at $1,000, you’re spending an extra $800 or so that’d be better invested into a pair of Radeon HD 5870s. As a result, before we show you any benchmarks, I’ll say that this probably isn’t the processor you need for a solid gaming experience. A Phenom II X4 or Core i7-920 is still plenty potent there. With that said, if money is no object and you want a six-core CPU and a pair of high-end graphics cards, you certainly can’t go wrong.

Six Cores Of Fury

The real reason to give Core i7-980X a long, hard look is that it’s a beast in the applications truly able to lean on its scaled-up architecture—and there are many. It leverages technologies first introduced on the Bloomfield generation of Core i7-900-series chips, like Turbo Boost and Hyper-Threading. But a recent shift to 32nm manufacturing results in transistors with decreased oxide thickness, reduced gate length, and, ultimately, less leakage current.

Consequently, Intel was able to increase complexity without pushing its design over the 130W TDP established by Bloomfield, giving us a six-core CPU featuring 12MB of shared L3 cache and capable of dropping into the same LGA 1366 interface you already know. The real question is whether the Core i7-980X is as Extreme as its price.

source:
http://www.tomshardware.com/reviews/core-i7-980x-gulftown,2573.html#xtor=RSS-182

Introduction

I’ve written a couple of articles on multi-core technology recently; one was on multi-core CPUs and one was on multi-core programming. Seeing as how I have got multi, multi-core thoughts floating around in my head I thought I would write another article on the topic. This article will discuss cache coherency as it relates to multi-core processors.

What is cache coherency? In the context of multi-core processors, cache coherency refers to the integrity of data stored in each core’s cache. But why is cache coherency necessary? To answer this question I will refer to the multi-core processor shown in figure 1. Imagine that there are two threads running through the processor; one in core 1 and one in core 2. Now imagine that each core accesses, from the main memory, variable ‘x’ and places that variable in its cache. Now, if core 1 modifies the value of variable ‘x’, then, the value that core 2 has in its cache for variable ‘x’ is out of sync with the value core 1 has in its cache. This is an important issue with multi-core processors. Actually, this problem is not very different from multi processor (multiple chips) cache coherency problems.

Figure 1: Diagram of a multi-core processor (courtesy of csail.mit.edu)

In multi-core processors, cache coherency is solved by utilizing an inter-core bus. This inter-core bus is a bus which is connected to the cache of each core; see the cache bus in figure 2.

Figure 2: Multi-core chip with inter-core bus (courtesy of csail.mit.edu)

Cache Coherency Protocols

There are two basic methods to utilize the inter-core bus to notify other cores when a core changes something in its cache. One method is referred to as “update”. In the update method, if core 1 modifies variable ‘x’ it sends the updated value of ‘x’ onto the inter-core bus. Each cache is always listening to the inter-core bus; if a cache sees a variable on the bus which it has a copy of, it will read the updated value. This ensures that all caches have the most up-to-date value of the variable.

Another method which utilizes the inter-core bus is called invalidation. This method sends an invalidation message onto the inter-core bus when a variable is changed. The other caches will read this invalidation signal and if its core attempts to access that variable, it will result in a cache miss and the variable will be read from main memory.

At first glance it probably seems like the update method is the preferred method because we all know that cache misses can cause significant performance costs. However, the update method causes a significant amount of traffic on the inter-core bus because the update signal has to be sent onto the bus every time the variable is updated. The invalidation method only requires that an invalidation signal be sent the first time a variable is altered; this is why the invalidation method is the preferred method.

This example of the invalidation method is very basic. Much work has been done over the years to improve cache coherency performance. This has resulted in a number of cache coherency protocols.

MSI

MSI is a basic but well known cache coherency protocol. MSI stands for Modified, Shared, and Invalid. These are the three states that a line of cache can be in. The Modified state means that a variable in the cache has been modified and therefore has a different value than that found in main memory; the cache is responsible for writing the variable back to main memory. The Shared state means that the variable exists in at least one cache and is not modified; the cache can evict the variable without writing it back to the main memory. The Invalid state means that the value of the variable has been modified by another cache and this value is invalid; the cache must read a new value from main memory (or another cache).

MESI

Another well known cache coherency protocol is the MESI protocol. MESI stands for Modified, Exclusive, Shared, and Invalid. The Modified and Invalid states are the same for this protocol as they are for the MSI protocol. This protocol introduces a new state; the Exclusive state. The Exclusive state means that the variable is in only this cache and the value of it matches the value within the main memory. This now means that the Shared state indicates that the variable is contained in more than one cache.

MOSI

The MOSI protocol is identical to the MSI protocol except that it adds an Owned state. The Owned state means that the processor “Owns” the variable and will provide the current value to other caches when requested (or at least it will decide if it will provide it when asked). This is useful because another cache will not have to read the value from main memory and will receive it from the Owning cache much, much, faster.

MOESI

The MOESI protocol is a combination of the MESI and MOSI protocols.
Directory-Based Cache Coherency

The protocols described above work very well and are commonly seen in both multi-core and multi processor systems. In the example of the multi-core processor I showed above, these protocols would work well. In the future, when processors contain dozens of cores, I am sure that these types of protocols will produce an abundance of traffic on the inter-core bus. One method of dealing with this will be to move towards a directory based approach. In a directory based approach each cache can communicate the state of its variables with a single directory instead of broadcasting the state to all cores.

In figures 1 and 2 the cores of the processor are connected via a bus, this may not be the case when the number of cores starts to increase significantly. I hope that I have shown you that the processor itself is really a network of cores which is quite similar to a network of computers with different interconnected topologies. With this in mind, many processors will be designed with their cores connected in a ring topology. Of course, if we are thinking of the processor as a network of cores there are a number of issues, other than topology, that need to be looked at. For example; application protocols, sessions, and basically any of the topics I discussed in my OSI Reference Model series, will apply to these networks of cores discussed.

Another factor that complicates the cache coherency issue is that some systems contain multiple processors, each with multiple cores. In this case each processor must keep its caches coherent with each other as well as coherent with the caches of the other processors. This requires a cache coherency strategy within each processor and a higher level strategy for keeping the caches coherent on the system level. Keep in mind that the strategies within each processor need not (and should not) be the same strategy used within an adjacent processor. These strategies should be tailored to the capabilities and needs of the individual processor. For example, even today many computers have a separate graphics processor which may have a different number of cores which are used in different ways. This would likely lead to a cache coherency strategy which is different than the strategy optimized for the CPU.

source:
http://www.windowsnetworking.com/articles_tutorials/Cache-Coherency.html

It’s important to know a few basic facts about processor cooling. All desktop processors available today are pretty easy to handle, as the actual processor die is protected by a large metal plate covering the entire upper part of a processor. This metal plate is called the heat spreader, as it provides a larger footprint for dissipating heat from the processor to the heat sink. It also protects the processor die from physical damage, such as might occur if you install a heat sink without taking sufficient care.

Processor heat dissipation depends on the processor architecture – Core 2 is much more energy efficient than the older Pentium 4 or Pentium D. The number of processing cores is also a factor, since a dual core will always require more energy than a single core, as is the clock speed at which the processor runs. The higher the clock speed, the more voltage has to be applied, which also causes increased power consumption. Power requirements increase exponentially as you increase either parameter.

Classic air CPU coolers differ in terms of size, layout, material and the fan type and size. Heat sink size, layout and material correlate greatly, as the issue is all about conducting as much heat as possible away from the hot spot onto a metal surface that is as large as possible. This is why efficient heat sinks will have lots of thin fins and use copper where possible: the metal heats up, and an air flow created by a fan is used to blow the hot air away from the cooler. Copper is much heavier than aluminum, which makes it somewhat tricky to build a full-copper heat sink.

Processors are often sold with a boxed cooler and fan. Today’s coolers are efficient and quiet, and they will be sufficient if you don’t intend to overclock your processor. Go for an aftermarket cooling solution if you want more cooling performance, though.

There are several possible points of failure for processor coolers, the first obviously being installation. Always make sure that you either use the included thermal pad or thermal compound. Use as little compound as possible, as it only has to ensure airtight contact between the CPU heat spreader and the heat sink. If the compound comes leaking out at the sides so that you have to remove it, you used too much.

As you install the heat sink, pay close attention to the installation. The surface has to sit on the heat spreader evenly. If you’re using a sophisticated aftermarket CPU cooler, it will probably have an efficient layout and provide basic heat dissipation for most desktop systems. However, the fan is required as soon as the CPU is actually used and placed under a higher load, which brings us to the point of failure we’ve looked into: every fan is a mechanical device with a limited life span. As the fan fails, the cooler will not be able to remove all the heat dissipated by the processor.

source: http://www.tomshardware.com/reviews/cpu-cooler-fails,1695-4.html

Introduction

Being the brain of the computer, the CPU plays a very important role in determining the performance of the system. Unfortunately, when it comes to choosing the best CPU (especially for a gaming computer), you will probably feel like a lost sheep. With different brands, models, speeds and specifications to choose from, it can really be a difficult task to decide which CPU is the right one for you.

In this guide, we give you a complete overview of what a CPU is, what are the important factors that will affect its performance and how you should go about choosing the the CPU that is best suited to your needs.

What is a CPU?

The CPU (Central Processing Unit), or sometimes known as processor, is one of the most important component in a computer system. Being the brain of the computer system, its task is to take care of all the data calculation and make sure they are processed in the fastest time possible.

CPU is not something you can see from the outside of the computer. In fact, you won’t be able to see the CPU on a fully-assembled PC. To see it, you have to remove the computer casing, unplug the wire and remove the heatsink (and fan), only then can you see the surface of the CPU. The shape of the CPU is a small square chip with lot of connector pin underneath.

How CPU works

How a CPU works is actually very simple and can be illustrated with the following 3 steps:

1. When you click to execute an application, the raw instruction is first fetched from the hard disk (sometimes from the memory) and sent to the CPU for processing.
2. When the CPU receives the instruction, it will execute the logic and compute the result.
3. Once the CPU finishes processing, it will send the result to the respective device to output to the user.

While it may seems easy, all these 3 steps must be completed in a split second. Delays in any of these steps will result in a lag in the computer, which we do not want it to happen on our gaming computer at all.

Therefore, to improve the performance of the computer, it is not enough to have a fast CPU, you still have to ensure that the transfer of information to the CPU are done in the shortest time.

Factors that affect a CPU performance

It is easy to think that the speed of the CPU is directly link to the performance of the CPU. In actual facts, this is only true to a certain extent. A CPU with fast speed will not be efficient if it has only a limited data to process each time. To achieve maximum efficiency, the hardware (especially the hard drive and memory) that linked to the CPU must supply data as fast as the CPU speed. Failure to do this will result in a lagging computer, regardless how fast the CPU is.

Below, let’s look at the other factors that affect the CPU performance.

CPU Clock Speed

The operating frequency of the CPU (also known as the clock speed) determines how fast it can process instruction.

The speed is measured in terms of Hertz, and it is usually lies in the megaHertz (MHz) or gigaHertz (GHz) range. A megaHertz means that the CPU can process one million instruction per second whereas a gigahertz CPU has the capability to process one billion instructions per second. In today technology, all CPUs run in the gigahertz range and you seldom see CPU with speed in the MHz range anymore.

Theoretically, a 500 MHz CPU is six times slower than a 3 GHz CPU. Equally, a 3.6 GHz CPU is faster than a 3 GHz or a 3.4 GHz CPU. In general, the higher the frequency of a CPU, the faster the speed of the computer.

Cache

Remember we mentioned above that for the CPU to work at its maximum efficiency, the data transfer from the other hardware must be as fast as its speed? The purpose of a cache is to ensure this smooth and fast transition of data transfer from the hardware to the CPU.

To understand how the importance of a cache, it is necessary to understand how the whole process works.The main bulk of information comes from the hard drive. When an application is requested, the motherboard will fetch the required information from the hard drive and deliver it to the CPU for processing. Since the hard drive processing speed is much slower than the CPU, data transfer often takes a long time. To fasten thing up, the RAM is used to store temporary information from the hard drive. Instead of heading straight to the hard drive, the motherboard now checks and retrieves the data from the RAM. Only when the required information is not found in the RAM then will the motherboard go to the hard drive.

As CPU speed increased to the point where the RAM is no longer able to catch up, the transferring of information again become a serious problem. To solve this issue, a cache, which was effectively a small and extremely fast memory, was added to the processor to store immediate instruction from the RAM. Since the cache runs at the same speed of the CPU, it can rapidly provide information to the CPU at the shortest time without any lag.

There are different levels of cache. Level 1 (L1) cache is the most basic form of cache and is found on every processor. Level 2 (L2) cache has a bigger memory size and is used to store more immediate instructions. In general, the L1 cache caches the L2 cache which in turn caches the RAM which in turn caches the hard disk data.

L2 cache plays the greatest part in improving the performance of the processors. The larger the cache size, the faster the data transfer and the better the CPU performance. However, cache is very costly. That is why you don’t find 1GB of cache in your system.

Multi-Core

In the past, if you want to get a faster computer, you have to get a faster CPU. Today, this is only partially true. The reason being, CPU speed can’t increase forever. There is limitation as to how fast the transistors can run and when it reaches a plateau, you won’t be able to increase the speed anymore.

To tackle this problem, CPU manufacturers (in this case, Intel and AMD) adopted a multi-core technology, which literally means putting multiple cores in a CPU chip. While increasing the CPU speed resulted in faster data calculation, putting more cores in a chip resulted in more work done at the same time.

Intel vs. AMD, Which Is Better?

You may have seen report saying that Intel is better, and on the next day, another report saying AMD is better.

You are confused…which one is better? AMD or Intel?

Both AMD and Intel CPUs are built on different circuitry and for that, it is impossible to compare apple to apple. If you were to ask me which one is better, I can only say that both are equally good and whether you choose an Intel or AMD CPU depend entirely on your preferences.

Below we will discuss the unique features of each CPU brand.

Intel: HyperThreading

Hyper threading is an Intel technology that enables the operating system to treat a single CPU as two separate CPU. In this case, the OS can split its workload into multiple threads and sent them to the two CPU concurrently. With the same amount of time spent, twice the amount of work can be done.

However, things don’t always happen as good as it sounds; HyperThreading does not necessarily lead to a performance increase. Let see why:

If you are running two pieces of software, each under its own thread, then HyperThreading can be effectively utilized to process the data simultaneously. In this case, you will see a noticeable boost in your system performance. However, in the event that there is only an application with a large chunk of data that cannot be easily split into smaller parts, the OS can only load one CPU with the calculation and leave the other idle. While this won’t cause your system to slow down, it is really a waste of resource. Such incidents are particularly true for games where all the logics are dependent to each other, and it is just not possible to split the tasks and processed with different CPU.

For HyperThreading to really increase the system performance, the software using it has to be specifically programmed for this optimization.

Hyper transport is an AMD technology designed to increase the communication speed between various components in computers. It is a completely different technology from Intel HyperThreading, but can achieve the same effect of raising the system performance. While HyperThreading serves to increase the amount of work done per CPU, HyperTransport serves to improve the data transfer process from other hardware to the CPU. What it does is to reduce the number of connection (buses) in a system, such that data can be transported from a component to another component in a shorter amount of time. This reduces the system bottlenecks and enables the CPU to use system memory more efficiently.

Core Frequency

As mentioned earlier, both Intel and AMD CPU have different circuitry and you can’t compare them apple to apple. This applies the same for their clock speed as well. If you have noticed, Intel’s speed always seems to be higher than AMD. Be careful, this does not imply that the Intel CPU is better.

The higher clocker speed simply means that there are more work cycles per second, not the amount of work done per second. Intel CPU has the tendency to divide its task into many small parts for easy processing. As such, the amount of work done per cycle is relatively small. On the contrast, AMD has lesser work cycle, but it processes more data per cycle. Thus, when it adds up, the amount of work done can be quite significant.

Unless we do benchmarking to determine the performance of each AMD and Intel CPU, it is definitely not a good idea to say that Intel is a better chip because it has a higher clock speed.

Front Side Bus

The Front Side Bus (FSB) is the communication channel that transfers data between the CPU and the other components in the system. Generally, the bandwidth of the FSB determines how much data can be transferred per second. The higher the bandwidth, the better is the system performance.

Since all the expansion cards (especially graphics card) connect to the CPU via the FSB, it is important to have a fast FSB speed to avoid any lag in the system performance.

In AMD, the HyperTransport technology has replaced the FSB with an integrated memory controller to control the data transfer to and from the components and the CPU. Due to lesser buses and more controllability, an AMD system is now able to send and receive information from various components simultaneously, and this resulted in a better performance.

Intel CPUs still use a more traditional approach, with the CPU communicating with the memory controller via the front side bus. So with Intel systems, a faster FSB often means somewhat improved performance.

Socket Type

The main reason why you can’t use an AMD and Intel CPU on the same motherboard is because they don’t have the same pin configuration. Because of the different in circuitry, the number of connection pins for both brands of CPU is also different. Even within the same brand, a specific model might use different pin configuration from another model. Thus, when choosing the CPU, it is important to bear in mind the socket type used by your motherboard.

How to Choose the Best CPU That Suits Your Needs?

For a gamer, you don’t want to have a CPU that is only good enough for word processing. What you really want is one that has a great deal of power to run the highest end games out there. While you may not have the budget to get the top end CPU in the market, you shouldn’t scrimp and get the cheapest CPU too. The more important thing is how you can strike a balance between performance and price. Here are three steps to choose your CPU.

1) Determine your budget

The last thing that you want to do is to max out your credit limit to get the most expensive CPU out there. Before you even start shopping, first determine how much money are you willing to spend on the CPU. While there is always a CPU for almost every price range, you will have to set aside about $200 for a decent gaming CPU

2) Select the brand

Choosing either a Intel or AMD CPU is really based on one preferences. Many benchmarking reports have shown that Intel score better than AMD in term of performance and heat generation, but it is more expensive. If you are low on budget, you may want to choose AMD CPU since it costs at only 3/4 of the price of a equivalent Intel CPU and still give you the performance you want.

3) Select the model

When choosing the model, focus on the no of cores, speed, and price. Check out forums/review sites to see how that particular model performs. If you are upgrading the CPU for your existing system, make sure the CPU uses the same socket as the one in your motherboard.

How to Save Money When Buying A CPU?

If you don’t have $1000 to throw around on a processor, that’s fine. There are many processors of different price range and you don’t necessary have to get the most expensive one. Here are some ways to save a few bucks.

Buy the next best processor

You don’t necessary need to get the latest processors in the market. More often than that, they are expensive and the support from other hardware is also not mature yet. You can easily save quite a bit of money simply by getting the next best processor.

Get an AMD CPU

For the same specification, AMD CPUs are generally cheaper than Intel CPUs. While many benchmarking reports have shown that Intel CPUs are better, the truth is that the differences is too small for you to notice. Using an AMD CPU for your game will not affect its performance to a great extent.

source: http://www.build-gaming-computer-guide.com/the-complete-guide-to-choose-a-cpu.html

The term overclocking is thrown around a lot, for better or worse. If you’re one of the many who has never overclocked, this guide will explain what it is and how to do it to the computers’ processor, motherboard and memory.

The prospect of overclocking a computer system can be intimidating for a computer newcomer, to say the least. The idea is simple enough; make the computer’s processor run faster than its stock speed to gain more performance without paying for it. The execution of this idea though, can be anything but simple.

Successful overclocking is as often a matter of ‘what you know’ as ‘what you have’. Understanding the maze of hardware dependencies and tweaks that can make the difference between a successful overclock and total failure is a demanding practice.

In this Beginners Guide, we will explore the process of overclocking processors, motherboards and memory to achieve a faster yet still stable computer. The article will guide readers step-by-step through understanding overclocking concepts, how to discover their hardware’s overclocking options and the actual process of overclocking. If you consider yourself an expert already, read on – there are a few tips and tricks packed into this guide that you may not know… or have a look at our recent experiment with underclocking. For insight into videocard overclocking, please see our companion guide on that subject right here.

What Does Overclocking Do?

Overclocking a computer’s processor or memory causes it to go faster than its factory rated speed. A processor rated at 2.4GHz might be overclocked to 2.5GHz or 2.6GHz, while memory rated at 200MHz might be pushed to 220MHz or higher. The extra speed results in more work being done by the processor and/or memory in a given time period, increasing the overall computing performance of the PC.

Can Overclocking Damage Computer Hardware?

Yes, but it’s typically unlikely. Generally speaking, when computer hardware is pushed beyond its limits, it will lock up, crash or show other obvious errors long before it gets to the point where the processor or memory might be permanently damaged. The exception to this is if extreme voltages are used when attempting to overclock, but since most motherboards do not support extremely high voltages, and neither does this guide, it’s not likely to be an issue.

For older processors, heat is also a factor worth keeping a close eye on. Modern processors have thermal sensors which will slow down or shut off the PC, but older CPUs do not necessarily feature these safety devices. The best know example of this is the AMD AthlonXP (socket A/462), which was famous for burning itself up in less than 5 seconds if the heatsink was not installed properly (or at all).

The Purpose of Overclocking

The most obvious reason to overclock a computer system is to squeeze some additional performance out of it at little or no cost. Overclocking the processor and system memory can significantly boost game performance, benchmark scores and even simple desktop tasks. Since almost every modern processor and memory module is overclockable to at least a slight degree, there are few reasons not to attempt it.

Important Overclocking Concepts

The following terms will be used throughout this guide, so it’s important to get a good grasp on them now.

FSB (FrontSide Bus): The data bus that carries information from the processor to the main memory and the rest of the system. A processor’s internal multiplier multiplied the FSB speed of the system = that processor’s speed in MHz or GHz.

Increasing the clock speed of the FSB (and thus the speed of the memory and the processor as well) is the most common and effective way of overclocking a modern computer.

AMD Athlon 64-based systems do not use a conventional FSB since the memory controller is built right onto the processor’s core instead of being located in the motherboard’s core logic chipset. Instead, a value called motherboard clock speed is used to determine the speed of data transfer between the processor and the memory. For the purposes of this article, FSB and motherboard clock speed are interchangeable terms.

Internal Multiplier: The ratio of a given processor’s speed (in MHz or GHz) as compared to the FSB (Frontside Bus) speed of the computer system it is installed in. A processor with an internal multiplier of 16x installed in a system with a FSB of 200MHz would run at 3.2GHz internally, since 16 x 200MHz = 3.2GHz. Most modern processors are ‘multiplier locked’ to some degree, meaning that their internal multiplier cannot be changed (or at least increased). This in turn means that increasing the FSB speed of a system is the only way to overclock the processor.

Memory Divider: Most modern Intel Pentium 4 and AMD Athlon motherboards allow a memory divider to be set. This divider allows the system memory to run slower than the actual FSB speed. By default, FSB speed and memory are usually set to a 1:1 ratio, meaning that increasing FSB speed (by overclocking) increases memory speed by the same amount. Most ‘generic’ system memory is not built for overclocking and thus may not be able to take the level of overclocking that the processor or motherboard can achieve.

The memory divider allows users to mitigate this problem by reducing the speed increase of the memory relative to that of the FSB and the processor. Setting a 5:4 memory divider would mean that memory speed increases at 4/5th the rate of the FSB, for example.

Reducing the relative speed of the memory does result in a slight decrease in performance as compared to the default 1:1 ratio between FSB and memory speed, but it may help users with generic memory achieve a higher overclock.

Stock Speed: The default or factory speed settings of computer hardware like the processor, memory and motherboard. With the processor, stock speed refers to the clock speed in MHz or GHz of the processor. With the memory, stock speed refers to the highest standard memory speed that the memory module is rated for (PC3200 DDR memory has a stock speed of 200MHz, for example). In the case of the motherboard, stock speed refers to the default speed at which the processor and memory work together, the FSB speed.

To tie this all together, say a motherboard has an Athlon XP 3000+ processor installed (stock speed 2.1GHz) which uses a FSB speed of 166MHz. A PC3200 DDR memory module (stock speed 200MHz) is installed. Since the processor requires a 166MHz FSB, the motherboard will set the memory speed to 166MHz which becomes its stock speed with the current configuration.

Core/Memory/Chipset Voltage: These three voltage values represent the amount of electrical power being fed to the respective components. When a processor, memory or motherboard is made to run faster due to overclocking, more voltage may be required in order for that component to run stably. With this in mind, voltage adjustment is one of the most important principles of overclocking.

If an overclocked computer becomes unstable, increasing one or more of these voltage settings by a very small amount (0.05V to 0.1V) can often mean the difference between an unbootable system and a stable overclocked one. That being said, it is important to make some distinctions with respect to voltage adjustments; more voltage does not necessarily mean faster speeds, rather minor increases can help improve stability. Computer circuits are designed to operate within very specific electrical ranges, and drastically increasing the electricity being supplied to a chipset will raise temperatures, and potentially damage it.

The role of the CPU, motherboard and memory in overclocking

When overclocking a computer, the processor, system memory and motherboard all have a different and important part to play in the process. The abilities and overclockability of each component has a significant effect on how successful the whole experiment will be. Let’s take a closer look at each component:

The Processor (CPU): As readers might know, two important variables govern how fast a modern processor goes. Its internal multiplier and the FSB (Front Side Bus) setting of the motherboard and memory. The FSB is the effective speed of data transfer between the processor and the main memory (it’s also the base speed that the system’s memory runs at), while the multiplier is an internal indicator of the speed of the processor.

A processor’s speed equals its multiplier (x) the FSB in MHz. Therefore, an Intel processor with a multiplier of 16 working with a FSB speed of 200MHz would run at 3.2GHz. There are two ways a processor can be made to run faster; increasing the multiplier, or increasing FSB speed.

Many modern processors have ‘multiplier locks’ which prevent users from changing the internal multiplier settings partially or completely, so increasing FSB speed tends to be the most common and effective method of overclocking.

The Memory: A system’s main memory speed determines the speed of data transfer between the processor, memory and the rest of the system. As you can imagine, this is the most important variable for computing performance in some systems. In all modern Intel and AMD systems, the FSB speed is directly linked to the speed of the memory by default, so the faster the memory is clocked, the faster the processor goes, since processor speed = (internal multiplier (x) FSB speed). This can be changed, but the 1:1 ratio between FSB and memory speed is the most desirable for overclocking.

AMD Athlon 64 systems do things a bit differently, since the memory controller is part of the CPU itself, so there is no conventional FSB carrying data from the processor to the rest of the system. Overclocking the memory still works essentially the same way, though the technology/terminology has changed. More on this in our AMD overclocking section below.

The Motherboard: Just as the motherboard is the heart of every computer system, it is also central to your overclocking efforts. The motherboard’s circuitry connects the processor and memory together and its BIOS options determine in what ways and by how much they can be overclocked. Even the highest quality memory and most overclockable processor can accomplish nothing if placed in a motherboard with no or limited overclocking options in its BIOS, or a board equipped with a new, poorly implemented or unstable core logic chipset.

Again, each component above depends on the other two when it comes to overclocking.

Hardware considerations for overclocking: Heat and cooling

The faster a computer goes, the more heat it produces. This is especially true when the voltage being fed to certain components is increased, a standard overclocking method. Excess heat in the processor, motherboard chipset or memory can cause crashes and system instability, and may be one of the limiting factors in determining the maximum overclock for a system.

The stock heatsinks included with most processors are perfectly adequate for cooling them at their stock speeds, but may not handle the additional heat generated by overclocking very well, especially if the computer chassis is not suitably ventilated. Readers may be better off investing in one of the many custom cooling solutions on the market, or at least buying some case fans to ensure an adequate flow of fresh air through their case. Take a look here for some cooling ideas. The same goes for the chipset and to a limited degree, the memory.

Hardware Considerations for Overclocking

It’s also important to keep an eye on the amount of heat the processor is putting out because of the ‘thermal throttling’ safety systems built into both AMD Athlon 64 and Intel Pentium 4 processors. If either of these CPUs gets too hot, they will slow themselves down drastically in order to keep from burning out. Users will notice this in terms of massively reduced system and benchmark speed, which should clue them into the fact that additional cooling is needed if they wish to continue overclocking. Thermal throttling should never occur in the regular use of a processor, but overclocking is NOT regular use.

To monitor the processor’s temperature, look for the ‘system health settings’ or similar entry in the BIOS screen.

Though some CPUs run hotter than others (Pentium processors tend to displace slightly more heat than their AMD siblings), all modern processors are happiest in the area between 35°C -65°C. If the processor is showing temperatures over 70°C in the BIOS, chances are that heat is going to be a limiting factor in the computer’s stability and overclocking potential. Time to consider a new heatsink and/or better case ventilation.

Power Supply Requirements

Overclocking a computer system also increases the amount of power it draws, and this may lead to system instability if its old 300Watt power supply is not up to the task. If overclocking a modern Pentium 4 or Athlon 64 system, plan on upgrading the power supply to at least 400Watts.

Spontaneous reboots while under load are usually a sign of insufficient power. When that happens, it’s definitely time to pick up a new power supply. On that note, while flashy powersupplies full of lights may be good to look at, they do not always indicate the best power quality, or power efficiency. In our experience, powersupplies which support PFC or Active PFC are the best choices.

Principals of Overclocking

Before we (finally) get to the practical part of the guide, here’s a brief word to prepare readers for the almost inevitable periods of frustration to follow. Overclocking is a very imprecise science; the processor depends on the stability of the motherboard and memory in order to achieve overclocking, and vice versa. If one of these components cannot stand the stress of overclocking, it will limit the other two also.

Heat, voltage and power supply stability are also relevant to overclocking success. Excess heat, not enough or too much voltage and unstable power can all cause the premature failure of an overclocking adventure, and it’s next to impossible to pinpoint what is causing the problem.

To avoid frustration as much as possible, be patient. Follow the directions below and take overclocking one small step at a time, so that when trouble occurs you will have a smaller set of potential issues to troubleshoot.

Preparing for Overclocking

In order to get the best out of current hardware, the most recent drivers and BIOS version for the motherboard need to be acquired. System benchmarks should be run pre-overclocking to establish a performance ‘baseline’.

Readers should visit their motherboard manufacturer’s website to obtain the most recent set of drivers for their motherboard, as well as the most recent BIOS version. For instructions on finding the current BIOS version and overwriting it with a newer edition. Newer BIOS versions may add overclocking options and stability, so this is always a good first step.

Establish a Performance Baseline

In order to get a good idea of how overclocking increases the performance of a computer, it’s important to take benchmarks and establish a performance baseline for the system.

Download, install and run the following benchmarks:

* 3Dmark2001SE
* X2: The Threat (download the demo and use the ‘run as a benchmark’ checkbox when loading it.)
* PCMark04
* Sandra 2005 (CPU, multimedia and memory benchmarks)

Record the results of each test. This will be the performance baseline, a level to measure the soon-to-be overclocked computer system against.

Readers should also consider downloading the Prime95 burn-in program, since it is extremely useful for stress testing an overclocked PC to ensure stability.

Examining BIOS Options

Since the motherboard controls what options you have for overclocking, it’s essential to take a look at the BIOS (Basic I/O System) pages of the board to examine the options available.

Reboot the computer and go to the BIOS screen by pressing the DEL key repeatedly during startup.

Note that some motherboards respond to different key commands to bring up the BIOS. Dell computers for example often rely on F2 or F8, while other manufacturers may opt for F10 to bring up the BIOS.

When the BIOS comes up on the screen you will be greeted with a blue menu with about a dozen headings. Navigate through the list by using the arrow keys, and press ‘enter’ to select a menu. When you’re done press the ESC key to exit, and either save or discard any changes you’ve made to the BIOS settings.

The first features to look for are CPU and FSB speed adjustment controls. Generally, these will be in a section of the BIOS called ‘frequency/voltage control’.

As seen in the screenshots below, this page will contain the FSB adjustment controls and voltage adjustment controls.

Increasing the FSB or ‘CPU host frequency’ or (Motherboard Clock or FSB or a host of other terms for the same thing) will increase the FSB speed of the motherboard, overclocking the processor and memory at the same time.

Increasing the voltage to the CPU core, memory or chipset will feed more power to those components to aid in stability while increasing heat.

This page may also contain memory divider options depending on the motherboard.

Everything needed to overclock the system should be on this one BIOS page.

Different motherboard’s BIOS screens will look different and use different names for the various menus and options, but the options themselves should be grouped together in one menu as seen above. If the memory timings options are not visible, try hitting CTRL+ALT+F1 when entering the BIOS.

The second BIOS page that should be identified now is the ‘PC health status’ page, or similar.

This page contains the readouts from the motherboard’s temperature monitors, allowing users to check how hot the processor is running.

In Case of Disaster

Though it’s less common with modern hardware, it’s quite possible that during this overclocking adventure the computer system may be (over)driven to a point where it simply refuses to even POST, never mind boot into Windows. If this happens, don’t panic.

To restore a system in this condition, power down the computer, unplug it and restore the BIOS to its default settings. This can be accomplished in one of two ways:

Set the BIOS (CMOS) reset jumper on the motherboard into the reset configuration and leave for 20 seconds (consult your manual for its location), then reset the jumper to the default position and power the system on again. The BIOS should now be reset to its default settings.

Or – Remove the CMOS battery (the flat silver disk) from the motherboard using a pencil or similar implement. Leave it out for a minute or two, then replace it and power on the system. The BIOS settings should have been reset, allowing the computer to boot.

Or – While turning on the PC, hold down the Insert key until the POST screen is displayed, enter into the BIOS and select default settings, save and reboot.

Memory Performance (latency vs. speed)

Memory latency is another important consideration when overclocking a computer system. The latency settings of the memory determine how long it waits for certain states to clear before performing new read or write actions. The lower the latency, the faster the memory will perform. Lower latency settings put more stress on the memory and increase the chance of error though, so many lower-end memory modules cannot handle fast latency settings, especially when overclocked. Raising the memory’s latency settings may enable a higher overclock to be achieved at the cost of some performance.

Memory latency settings can generally be found in the ‘advanced chipset features’ section of the BIOS.

source: http://www.pcstats.com/articleview.cfm?articleid=1804&page=1

When it comes to PC performance, few people demand more power than gamers. Keeping up with developers who are constantly pushing hardware to its limits requires the best system possible, specifically tailored to peak every bit of performance it possibly can.

Unfortunately for buyers, this means a gaming PC can not really have a weakness. While the most important component is definitely the graphics card, you can not neglect other areas. If you are lacking in CPU or RAM power then you will find your PC bottlenecking at those parts and you will not be able to make the most of your card.

If you are on a tight budget however, there are certain shortcuts you can take to reduce the cost, including a cheaper monitor, mouse and a wise selection of other components. This guide will cover these shortcuts and also options for the hardcore enthusiast who wants to squeeze every frame per second out of their games.

Packages

A gaming PC

The gaming PC market is growing rapidly. Gamers are one of the few groups of people who are willing to spend more than $3000 on a PC to ensure they get the maximum possible performance. Originally this market was only catered for by small niche companies who produced high end PCs targeted specifically at the gaming sector; however in recent times sales of these PCs have increased dramatically and larger companies are beginning to mount an assault on the market.

There are two distinct ways you can purchase a gaming PC. Many companies offer traditional PC packages aimed towards gamers, which come with a predefined collection of components and hardware. Most retailers in Australia that offer these have several different options, ranging in price and performance that cater to different markets. Alternatively, many companies offer a range of customisable parts, sometimes grouped within different price brackets which give you more freedom but also allow room for error. Be sure to carefully read up on the parts you are purchasing if you select the latter option, to make sure everything is compatible and suits your needs.

The hardware

Whatever option you choose, you will need to make up your mind about which individual hardware configuration would best suit you. It is likely that any pre-built gaming PC you purchase will be able to adequately play most games, but knowing how different components relate to different needs will help you can maximise your PC’s effectiveness.

When buying a gaming PC, there are a large number of factors to take into consideration, but the most important are graphics card, memory and CPU in that order. There are also important peripheral considerations such as mouse, headphones and sound cards.

Graphics card

An ATI x800 graphics card

The graphics card is probably the most important element in creating a successful gaming PC. Very few other functions test your graphics card as much as a modern game. As a result, this is the area in which you should find yourself spending the most money.

There are two main competitors in this area, ATI and NVIDIA, who licence their GPU technology out to other companies. Neither is considered greatly ahead of the other in terms of game performance, although NVIDIA has recently released a new generation of cards, the 7800 range, which ATI has yet to respond to. At the moment many games are CPU limited, which means the power of your CPU will often be the factor that limits how well a game performs. This does not allow some high end cards to reach their full potential, but if you buy a high end system it should be capable of handling the most powerful card on the market. This may change in the coming months with the release of ATI’S R520 core, but no concrete specifications have been released.

There are a few things to look out for in a good graphics card, particularly DirectX 9 support and a reasonable quantity of memory. DirectX 9 is a system used by developers to render graphics and is used in the majority of modern, graphically intensive games, so this support is vital. These games also require a large quantity of memory bandwidth to operate within so you should aim for a card with a minimum of 128 megabytes of RAM. Most modern cards will have this, as well as support DirectX 9, but it is worthwhile to check. Graphics card memory comes in the same amounts as regular memory. Some of the common configurations include:

* 32 megabytes (older cards, two or three generations back)
* 64 megabytes (two generations back)
* 128 megabytes (last generation)
* 256 megabytes (last generation and this generation)
* 512 megabytes (next generation)

The absolute minimum you should spend on a gaming card is about $250-$350. This will get you either an NVIDIA 6600GT or an ATI x800, both of which will run modern games comfortably on medium to low settings. These cards are excellent value for money, and should definitely be part of a budget gaming system.

There are some great bargains at the higher end of the spectrum too. With the release of the 7800 range, prices are being driven downwards. The NVIDIA 6800GT and the Radeon x800xl are both hovering around the $500-$600 mark at the moment, and will hopefully drop in the near future. They can both run any game on the market with moderate to high settings and will please all but the most hardcore gamer.

Memory and GPU clock speeds are the primary factors in gaming performance and generally the higher these speeds are, the faster the card is. The pixel shader technology the card incorporates is important if you’re playing very modern games as this technology comes into play a lot. The number of pipelines is also important as these are the channels that convert the image from the card to your screen, so the more the better. Modern cards run at 16 or 24 pipelines, but an 8 pipeline card would probably suffice if you are on a budget.

One other option to consider is an SLI setup, which involves running two cards together to share the load, which increases performance by 30% or more. It only works with select cards such as the 6600GT or the 6800 Ultra and requires a specially designed motherboard with multiple PCI-E slots. Computer enthusiasts will relish the extra grunt, but for most situations a single high end card is more than enough. ATI are coming out with competing technology called Crossfire in the coming months, which purports even higher performance increases.

Most pre-built gaming systems will come with a quality graphics card because companies recognise this is vital to a great gaming experience. Pay attention to the brand however, as different brands have different strengths and weaknesses. Some clock their card at slightly higher rates, offer extras such as games and software, have DVI or TV-out connections, or offer overclocking warranties for that extra bit of power. You can squeeze a little extra out of your system with cards like this, which is great if you are on a budget, but you may also be paying extra for things you don’t want or need, such as useless software and outdated games. It is worth putting a bit of research into such a vital part of your gaming system.

Memory

The importance of memory when gaming cannot be underestimated. When it comes to basic PC use, web browsing, word processing, emailing etc, memory does not really factor in a great deal. When running complicated applications or games however, you will quickly find your system slowing down and will need to create virtual memory on the hard disk if you are not adequately prepared.

You can see our memory guide for a more detailed explanation of how memory works and what the different terminology means, but like anything in the computer world, higher numbers are better. Any gaming system you purchase will use DDR RAM now, with some utilising the faster DDR2 variation.

The minimum you can get away with in modern games is 512MB, but with windows XP using a large chunk of that, 1GB is definitely preferred. You can pick up 1GB of RAM for under $180 now, with prices really bottoming out, so it costs almost nothing to give your system a real shot in the arm. Some recent games such as Battlefield 2 will struggle even with that quantity and will perform better with 2 GB, which will set you back about $350-$400. RAM is the least expensive way to give your system a real boost and if you stick with 1GB, you can always purchase another 1GB stick later to increase performance.

PC3200 is the standard high end RAM, running at 400Mhz it gives plenty of grunt and is definitely one of the best options. There are RAMs that are clocked higher (such as PC3500, 3700 and 4000), but they are only really useful if you are an overclocker or a gaming enthusiast.

Similarly, DDR2, while boasting bigger and better numbers, is not currently much better than regular DDR, as systems and programs are not made to take full advantage of it. It currently boasts negligible gains over its predecessor. Like 64 bit technology, DDR2 will shine once programs catch up. It could be a worthwhile investment if you are building specifically for the future, but everybody apart from overclockers should be fine with PC3200.

CPU

The processor is the central part of any PC, controlling the majority of operations going on at any one time. Different games draw more on different parts of the PC with many being CPU intensive, so you a need a mid to high end CPU to run modern games at a comfortable level. We would recommend a CPU with a rating of at least 3.0GHz on the Pentium scale (or a 3000+ on the AMD scale) as being the absolute minimum required.

However, with the release of dual core technology and developments in high end chips (such as the Athlon FX57 and Intel’s Extreme Edition releases), mid range chips around the 3.5ghz mark are rapidly becoming more affordable, and should be strongly considered even if you are on a budget. They will give a much needed performance boost for comparatively little money.

For the games enthusiast, chips can be acquired that are clocked at over 4.8GHz, but they come with hefty price tags, often in the $1400-$1500 range. If you are looking at spending that sort of money, it may be worthwhile to consider a dual core processor, which is discussed below. A single core powerhouse might be better on current games, but some future releases may benefit from the new technology.

64 bit processors are becoming the standard these days and with good reason. 64 bit computing, while not yet readily used and available, offers a huge performance increase, and although you can spend less to get a high end 32 bit chip, you’ll be missing out by the end of 2005 when the technology really takes off.

With the current generation of chips, AMD seems to have the advantage in terms of gaming performance. The architecture of their socket 939 chips has shown greater performance in gaming benchmarks than comparable Intel chips (or socket 754 AMD chips) which makes it the best choice for the hardcore gamer who is not interested in other applications. Intel’s hyperthreading technology means it shines in other areas, such as multi-tasking. So if your machine is intended for other functions besides gaming, be sure to take this into account.

Both companies now produce dual core CPUs as well, which makes the decision process more complicated again. These chips essentially take two CPU cores and link them together on a single chip, offering increased processing power. Their primary strength is in multi tasking, which is not a big concern for the current generation of games; however much like 64 bit technology, future releases will incorporate and take advantage of these developments, and so they are worthwhile in the long term. They do cost quite a bit more than normal chips, coming in at over $1000 in most instances and so are only viable for the hardcore gamer.

Motherboard

While your motherboard is a vital component in any PC, the quality of the motherboard won’t offer much to your gaming experience. The motherboard is the component which all the other pieces of hardware plug into. It is the medium through which your various pieces talk to each other. Thus you need a reliable board, but it’s not something you can really use to squeeze extra performance out of your system.

With this in mind, the main selling point of a motherboard is going to be its extras. These include things like RAID support, firewire support and built in network support. Most of these won’t be particularly relevant to gaming. However there are exceptions, specifically PCI-E support. PCI-E stands for PCI Express, which is a new technology being phased in to replace AGP. It offers faster data transmission than its predecessor and motherboards are increasingly being released that support it. All graphics cards will either be PCI-E or AGP so we strongly recommend purchasing a system with PCI-E support. It’s the technology of the future and if you’re buying a new system there is no reason not to upgrade to it as there is virtually no price difference and a year from now it will be the standard format (Nvidia have already begun to ignore AGP with their latest card, the 7800). SLI is another technology that is worth considering, but that will be covered in more depth further on in the guide.

Another element that may be important is the overclocking options of the board. While the CPU itself is an obvious limitation on how far it can be overclocked, different motherboards offer different overclocking options, such as FSB, multiplier and voltage control. Some will overclock the same chip better than others. This is only important for the computer enthusiasts who want to get the most out of their system and for most purposes the majority of motherboards will perform the same.

Soundcard

Depending on who you ask, the soundcard is either the be all and end all of a system or, totally irrelevant. With regards to a gaming system, it depends on the sort of games you play. A competitive first person shooter for example, often requires stealth so the sound of footsteps and gunfire is vital to locating and dealing with opponents. In action and horror style games, sound helps create atmosphere and mood to immerse you in the game. If you play more racing and strategy games however, sound might not be so vital. It is background and somewhat token, as opposed to vital to the gaming experience. So depending on the purposes of the machine, your sound card needs will differ.

For most gaming purposes a standard 5.1 surround sound card will be more than ample. You could spend several hundred dollars on a high end card and hardly notice a difference in game. Expensive sound cards are largely targeted at an audiophile market, or sound and video editors who require crystal clarity in their audio applications. As long as you stay away from onboard sound (which no gaming PC should come with anyway) and purchase a good pair of headphones (discussed below) then your games should sound great.

Cooling

A PC case fan

Most PCs come with basic cooling. Increases in technology have meant that modern CPUs and graphics cards typically run at higher temperatures than their predecessors so most systems have a CPU fan and a heat sink or two to help ease the load. Gaming PCs are no exceptions to this; however some do come with added cooling features which are worth noting.

If you are an avid PC user, and are interested in trying to overclock your system, then extra cooling might be just the thing you need to push it that little bit further. A few more powerful fans in the right places can do wonders. Most vendors stock higher quality third party fans for cases, CPUs and graphics cards, so do a little investigation and choose the combination you think fits best. You can buy some extremely powerful heatsinks designed to allow you to push your system to its limits, but these can cost in excess of a few hundred dollars.

Alternatively, you could take it a step further with water cooling. Not as dangerous as it sounds, water cooling involves pumping water through small pipes placed against hotspots in your system. The water cools down vital system components and is considerably more effective than air cooling. It does come at a cost, but it is not out of reach with a basic setup being available for about $300. Some retailers might even help set it up if you decide to enquire about it, so be sure to ask.

There are other methods of cooling out there, including phase cooling (which uses refrigeration techniques) and several types that involve chemical combinations, but they are often unstable and only available from retailers that focus on extreme cooling.

Extra cooling is only necessary if you intend to push your PC past the standard limits. You can boost your performance if you know what you are doing but it can be dangerous. Hardware can be damaged or completely ruined if you push it too far, so overclock at your own risk.

Other pieces

There are other elements to a PC that are less vital to gaming, such as hard drives, DVD burners and floppy drives. You can tailor the amount of disk space, and the types of drives you want to your needs. Most gaming PCs should already come with a basic CD-Rom and probably DVD drive, as well as a reasonable quantity of hard disk space (50-100GB).

You may also need to consider the network card that comes with the system if online play is important to you. Most PCs will come with broadband enabled cards out of the box these days, but it’s worth checking to be safe. A standard 10GB card will be fine for gaming over a LAN, but you may wish to pursue a more powerful 100GB connection if speed is your thing.

Peripherals

In addition to what’s in your box, exterior peripherals are equally important for a proper gaming experience.

Case

Not necessarily a peripheral in the traditional sense, some companies that specialise in gaming PCs like to make their cases a little more aesthetically pleasing than normal. While this does not have an impact on the way you play games, some people make a hobby out of modifying their case with windows, LCD lights and other setups. This is not a big concern for most people, but if you take your PC to LANs (events where many people bring their PCs to network up and play games) then the look of your PC may be important.

Monitor

An LCD monitor

For many people, the monitor is one of the most important parts of their computer. It is the vessel through which you view the thousands of dollars you’ve spent on your system already, so it’s logical to buy a decent one. While it won’t actually improve your game play much, playing a modern game with all the settings cranked up on a 19 inch screen is a great experience.

The main choice is between LCD and CRT. Originally LCDs suffered from a “ghosting” effect, where fast movements in game would lead to a blurring that made it difficult to see. This was particularly evident in first person shooters and other fast paced games. Thus, CRT monitors have been the favourite amongst gamers.

Recent LCD technology however, has drastically reduced the response time of the monitors. Now you can buy 12ms (millisecond) or 8ms LCD screens which have virtually no blurring. A few things to note about LCDs

Pros:

* They have a wider viewing area than CRTs (a 19 inch LCD is bigger than a 19 inch CRT.)
* They are easier on the eyes.
* They are lighter, which is a big factor if you go to a lot of LANs.
* With DVI input you can avoid any loss of quality when converting from analog to digital (as is popular with CRTs.)

Cons:

* There are still some minor ghosting issues (not enough to really be noticeable).
* They have a native resolution (which means you must play in that resolution to get the best image quality); this is typically 1280×1024, which is not small by anyone’s standards.

At this stage of development it comes down to personal preference. The LCD ghosting is all but gone from modern monitors, but if you go down that road be sure the monitor the computer comes with has an adequate response time for gaming (12ms minimum, 8ms preferred). LCD is definitely the technology of the future, but at the moment the differences are quite negligible. CRTs however can be bought cheaper than LCDs, so if you are on a tight budget that is probably the way to go.

Mouse and mouse pad

A greatly underestimated part of your gaming setup, many games rely on quick mouse reflexes and accurate responses, so a quality mouse is crucial. Regardless of whether you’re shooting down enemies or quickly managing large numbers of troops, a good quality mouse will improve your game play dramatically. Optical mice have basically taken over from ball mice, with technology having developed to the point where they are skip-free on all but the lightest, shiniest surfaces. They work by having a tiny camera that takes hundreds of pictures of the surface and uses them to work out how far the mouse has moved.

There are a number of choices in this market, ranging from laser mice to wireless mice. Many wireless mice suffer from response problems which are not noticeable during everyday office applications, but can become a pain when gaming. More recent developments seem to have this under control, but it is worth asking to test the mouse you wish to buy, just in case.

Pay attention to the DPI of the mouse you are getting, as that indicates the quality of the camera present and the number of shots it takes. The higher the number, the more accurate the mouse is. Some gaming mice have as high as 2000 DPI, while basic desktop opticals clock in at a mere 200 or 400. Typically a good quality mouse will set you back $80-$100, and will last several years if kept in good condition.

A Funcpad mousepad

To go with your mouse you need a good quality mousepad. While you may think the $2 pad that came with a game is a suitable surface to play on, many game enthusiasts would tell you differently. There are a number of well known gaming pads, ranging from cloth through to glass, metal and plastic. Some are specially created by professional gamers, others by individual companies that focus on gaming mice and accessories. Do a little research and find the surface that is right for you.

Keyboard

The quality of your keyboard is much less of a factor. As long as it is comfortable and functional it won’t really have an impact upon how you play. There are several gaming specific keyboards available, but they are generally acknowledged as being inferior to a regular setup, as their layout is unwieldy and in attempting to enhance gaming options become useless for anything else. There are also several wireless keyboards, which are a viable consideration if you are trying to minimise your cords. There are a few known problems with response times (as with wireless mice) so try to use one before you buy it if taking this path.

Headphones

While many home PC users enjoy the surround sound associated with a speaker setup, competitive gamers swear by headphones as their preferred audio source. As with sound cards, the importance of the sound depends on the type of games you play. With first person shooters for example, the ability to precisely locate the source of a sound is enhanced by the use of headphones and you can achieve the sort of immersion that comes from an expensive speaker setup for a fraction of the cost.

They are also very useful if you attend a lot of LANs, as speakers are prohibited at such events. Similarly, living with other people or in a small space, you can play your games as loud as you want without worrying about disturbing your room mates.

Most computers will not come with headphones as standard so it will be something you have to pursue as an extra or from a different company. You can get a good pair of headphones for around $80-$100, but the quality continues to increase the more you spend. If you enjoy listening to music while you game, spending a little more for a better listening experience can be well worth it.

Price

If you are looking at pre-packaged systems, one thing to do before buying is price the individual components to get a rough idea of what you are paying for. A quality gaming PC will never be cheap, with a full budget system setting you back around $2000. For a high end experience you will be looking at over $3000, with extremes available for over $5000-$6000.

That said some companies may place an exorbitant premium on their systems simply for the job of piecing them together. If you check the rough market price of the individual components that make up your package you will get a rough idea of what you should be paying and any great disparities will become obvious.

source: http://www.pcworld.idg.com.au/article/190248/gaming_pcs

Why does your CPU need cooling? OK you know it gets hot without cooling but do you know what happens if you don’t use adequate cooling. With different CPU’s different results can happen, some annoying some expensive. The basics are that heat equals slower CPU performance and possible damage.

Older CPU’s used to be made of simply transistors. Now CPU’s contain Transistors, Capacitors and Resistors. Resistors produce a lot of heat and this needs removing as quick as possible. If excess heat remains on the chip electromigration or oxide breakdown can occur which can lead to crashes and CPU failure. 10 degree’s extra heat on a CPU half’s its life span. A further 10 degree’s halves the life of the CPU again. You may think that you don’t keep your CPU’s for that long but life span is only one thing, the performance and stability is another.

Toms Hardware Guide have a great video to watch about what happens to CPU’s without cooling.

Overclocking, How does this effect the heat issue?

Overclocking your CPU does cause extra heat to be produced. The amount of extra heat depends on what type of overclocking you do. If you increase the CPU’s frequency by increasing the FSB then the extra heat will increase linearly. If however you have to increase the voltage the excess heat will be the square of the voltage increase. Simply put increasing the voltage will create far more heat than simply overclocking the Bus speed. However many overclockers found that increasing the voltage is the way to keep the system stable. There is a perfect article at Overclockers.com very in depth and informative.

Click here for the latest Prices on the top fans and other cooling equipment

I have a fan what else can I do?

Firstly you can take a look at improving your cooling solution. This could be simply a bigger and better fan. A heatsink with better spread technology or you could check out other types of cooling at our CPU cooling article. You may also want to consider Case fans and sorting out your wires in your case to allow better airflow.

The other option is a software option. Programs like Rain, Waterfall and CPU idle help keep your CPU cool by sending it to sleep. This is done by sending the HLT (Halt) command to the CPU, this command sends the CPU into suspend mode saving power and gives the heatsink/fan extra time to disperse the heat. The command is only sent during empty CPU cycles so performance is not compromised.

There is debate on whether these programs actually do your CPU any good, as the program has to re-issue the command constantly therefore putting undue stress on your CPU and reducing its life. My personal view on this is that I can see why there are these arguments but the advantages of heat being dispersed greatly out weigh the simply single command constantly being sent. Other items such as the system clock etc. Use many CPU cycles and don’t take up much CPU power. I do use Waterfall myself and see an average of 75% power saving with no bad experiences. There will always be debate, pro’s and con’s. This will have to be one you decide for yourself.

source: http://www.pantherproducts.co.uk/Articles/CPU/Whycool.shtml