CPU & GPU Processors
Their Purpose and Function
Intro
Now that we have covered what exactly HCI's are and their importance it is time to take a step down a more technological road and delve into the weird mechanical world of Processors and computer generated calculations. It of course would be impossible to talk of the current games platforms without talking about CPU's and their graphical counterparts. Some of which are relatively new inventions by technological standards.
So within the following section of this blog I will be discussing and detailing the inner workings and purpose of the mystical CPU and GPU.
CPU's
Featured within every home computer, console, mobile phone, smart TV and various other electronic devices is a CPU, better known as a Central Processing Unit. Without this 35mm sized thin layer of metal your computer or electronic device would be simply unable to function. You see, a CPU is the brain of a computer, the Shepard which guides and tells everything where to go, what to do and how to do it. It access millions of bytes of data at a time and performs billions of computational equations and cycles every second or so. It is put simply a mechanical wonder capable of allowing your device or platform to do so much. CPU's are constantly evolved and are dynamically changed from iteration to iteration ever changing and ever flowing from model to model each with a new set of pro's and cons. But, enough of their power and importance, as it is time to move on to: How they work, What they do and Why they do it.
Now, before we get into the major details about processors it's important to note their side components. Without these use of the processor may not be possible or at the very least hard. For example: processors have tendency to get extremely hot due to the reasons explained later, to counteract this heat they will have at least one fan connected to the processor itself built for the sole purpose of cooling down the CPU and aiding the heatsink, in modern computers due to the extreme amount of calculations and the power of processors it is now commonplace to see a heatsink attached to most processing units. A heatsink is in principle much like that of a car radiator, it is designed to draw and attract heat away from the CPU through the use of thermal fins which attract the heat and disperse it evenly throughout the computer therefore removing the heat from both the CPU and the heatsink the fan also blows excess away in various directions aiding with the cooling process. Without these safety's in place the CPU would simply cease to function from extreme temperatures and in many cases it would simply melt!
A CPU is (on average) a 35mm x 35mm wide chip which hosts layers upon layers of tiny transistors. Transistors are extremely tiny microscopic pieces of metal containing elements of gold, copper and palladium which are capable of blocking electrical charges when they are too weak and allowing electrical charges to pass through when sufficiently strong enough. They are conductive but begin as non-conductors, as sufficient electrical charges pass through them they transition from being non-conductive to conductive. This is due to the materials with which they are made of loses resistance as the electricity passes through it. This gives them the unique ability of being semi-conductors, something which is necessary for the processors operation and function as the brains of the machine
CPU's in modern days contain billions of transistors with all of them being incredibly vital to the CPU's workflow and power. Every transistor will receive sets of inputs and will in turn produce a new output. This works thanks to the semi-conductive state explained earlier whereby when one or more of the inputs receives an electrical pulse it changes the transistors state internally and produces a different result on the other side. It is this simple alternating pulse affect which makes it possible for the CPU to calculate logic operations and calculations.
A CPU's power can often be rated by it's clock speed and is one of the easiest ways of describing it to customers and users. The logic behind it is that the higher the clock speed the better the chip, (although there are a variety of factors along with this to take into consideration.) A CPU's clock speed, put simply is how many cycles per second an electric charge passes through the transistors of the chip. In modern processors it is usually easily in excess of billions of cycles per second. When it comes to computing terms a billion is represented as a Giga and seconds are represented as hertz so when talking about modern clock speeds of 3.2 billion cycles a second it would be 3.2 Gigaherz. Now the faster a clock speed goes means the faster a CPU is able to handle memory allocation, read/writing and various other tasks but clock speed is not simply the be all and end all of a processor's power as measuring how many actual instructions are being processed each second (represented as MIPS, BIPS and TIPS, which stand millions, billions and trillions of cycles per second.) is a far more accurate representation and has a greater impact on the processing overhead of video games.
Another way in which a CPU can be rated and described is by that of its bit depth. A processor's bit depth in an extremely simplistic term simply means how many bits it will process or transport across the internal processor bus in single cycle. The number of of bits in which a CPU will usually register or instruct across the bus is expressed within multiple's of 8. In current hardware, processors are capable of 64 bit processing with 128 bit looming within the horizon.
Each
transistor is a receives a set of inputs and produces output. When one
or more of the inputs receive electricity, the combined charge changes
the state of the transistor internally and you get a result out the
other side. This simple effect of the transistor is what makes it
possible for the computer to count and perform logical operations, all
of which we call processing. - See more at:
http://www.inetdaemon.com/tutorials/computers/hardware/cpu/#sthash.3CqqwmNH.dpuf
Each
transistor is a receives a set of inputs and produces output. When one
or more of the inputs receive electricity, the combined charge changes
the state of the transistor internally and you get a result out the
other side. This simple effect of the transistor is what makes it
possible for the computer to count and perform logical operations, all
of which we call processing. - See more at:
http://www.inetdaemon.com/tutorials/computers/hardware/cpu/#sthash.3CqqwmNH.dpuf
Each
transistor is a receives a set of inputs and produces output. When one
or more of the inputs receive electricity, the combined charge changes
the state of the transistor internally and you get a result out the
other side. This simple effect of the transistor is what makes it
possible for the computer to count and perform logical operations, all
of which we call processing. - See more at:
http://www.inetdaemon.com/tutorials/computers/hardware/cpu/#sthash.3CqqwmNH.dpuf
GPU's
The reason for you being able to play videogames, the reason this blog is able to be written and the reason for games getting better both visually and mechanically are all by-products of the power outputted and used by both the CPU and the next section of this blog the GPU.
GPU (Graphical Processing Unit) is a processing card placed within either a PCI/PCI-E port of a PC or in many other cases built into the processing chip itself. The dimensions, weight, power requirement and power output along with temperature and heat vary massively from card to card as there are constantly technological updates along with constant card revision almost every 6 months to a year.
A GPU is a single chip processor featured within a card like housing. It's responsible for computational mathematics of items such as the placement, shape, geometry and angle of 3D objects and their separate polygons along with the maths behind lighting, lighting effects, shadows and other extremely processor intensive tasks. Shaders have to be written to cater for a wide variety of GPU's as one is often never the same as the other as companies will often seek to create new methods of calculations and new technologies to help speed up, lighten the load of the task or increase performance of many of the tasks relied upon for the GPU. They will often in many cases also apply visual effects processing which is dependent on certain cards for example certain NVidia cards are capable of producing special Anti-Aliasing effects within certain chip set release's such as the GTX 9XX series.
Just like CPU's GPU's also have clock speeds although they sometimes smaller than that of their PC counterparts as they only have to calculate graphic specific calculations rather than the operations and calculations of an entire computer system. For example the clock speed of the PS4 graphic's card is 800mhz whilst it's CPU clock speed if that of 1.56 GHZ. Also they have built in memory caches to help with larger texture support and pixel/vertex shading methods as it is able to temporarily store and quickly load much more texture and shader information.
Now, as well as independent GPU's there are 2 other types of GPU's:
1. IGP
Integrated Graphics Processing (IGP) chips are again single chip graphical processing units with a difference. As touched upon above these are graphics cards which are directly integrated within the CPU themselves. They are commonplace among low end desktop, Laptop and Notebook Personal Computers. Often sharing the processing power and overhead of the CPU they can cause an extreme strain upon the CPU and cause unwanted errors such as slowdowns, lagging and often computer crashes of the system. These types of GPU were widely used in the early days of 3D as the graphical capabilities of games were not as intensive as to those of today's. This is the reason why independent GPU's have taken a surge within the market in the last 2 decades.
2. CDGP
Custom Designed Graphical Processing (CDGP) Are a type of graphical card custom designed for a specific hardware system the most commonplace being of course console. The Xbox One, Wii U and PS4 respectively use custom designed GPU's for their hardware. These are often developed a series of years by either the company themselves or a 3rd party manufacturer such as ATI or Nvidia. In many cases they hold significant advantages over those of commercial grade independent GPU's as they are developed specifically for the system ensuring greater compatibility and more optimisation of tasks. But, they also fail significantly compared to the areas of those of independent GPU's as they are not able to be upgraded every year and they are not susceptible to driver updates from the manufacturer which mean that after 2 years they become slightly obsolete.
The main 2 Manufacturers of GPU's are Nvidia and ATI with many other smaller companies developing variations of their GPU cards.
GPU (Graphical Processing Unit) is a processing card placed within either a PCI/PCI-E port of a PC or in many other cases built into the processing chip itself. The dimensions, weight, power requirement and power output along with temperature and heat vary massively from card to card as there are constantly technological updates along with constant card revision almost every 6 months to a year.
A GPU is a single chip processor featured within a card like housing. It's responsible for computational mathematics of items such as the placement, shape, geometry and angle of 3D objects and their separate polygons along with the maths behind lighting, lighting effects, shadows and other extremely processor intensive tasks. Shaders have to be written to cater for a wide variety of GPU's as one is often never the same as the other as companies will often seek to create new methods of calculations and new technologies to help speed up, lighten the load of the task or increase performance of many of the tasks relied upon for the GPU. They will often in many cases also apply visual effects processing which is dependent on certain cards for example certain NVidia cards are capable of producing special Anti-Aliasing effects within certain chip set release's such as the GTX 9XX series.
Just like CPU's GPU's also have clock speeds although they sometimes smaller than that of their PC counterparts as they only have to calculate graphic specific calculations rather than the operations and calculations of an entire computer system. For example the clock speed of the PS4 graphic's card is 800mhz whilst it's CPU clock speed if that of 1.56 GHZ. Also they have built in memory caches to help with larger texture support and pixel/vertex shading methods as it is able to temporarily store and quickly load much more texture and shader information.
Now, as well as independent GPU's there are 2 other types of GPU's:
1. IGP
Integrated Graphics Processing (IGP) chips are again single chip graphical processing units with a difference. As touched upon above these are graphics cards which are directly integrated within the CPU themselves. They are commonplace among low end desktop, Laptop and Notebook Personal Computers. Often sharing the processing power and overhead of the CPU they can cause an extreme strain upon the CPU and cause unwanted errors such as slowdowns, lagging and often computer crashes of the system. These types of GPU were widely used in the early days of 3D as the graphical capabilities of games were not as intensive as to those of today's. This is the reason why independent GPU's have taken a surge within the market in the last 2 decades.
2. CDGP
Custom Designed Graphical Processing (CDGP) Are a type of graphical card custom designed for a specific hardware system the most commonplace being of course console. The Xbox One, Wii U and PS4 respectively use custom designed GPU's for their hardware. These are often developed a series of years by either the company themselves or a 3rd party manufacturer such as ATI or Nvidia. In many cases they hold significant advantages over those of commercial grade independent GPU's as they are developed specifically for the system ensuring greater compatibility and more optimisation of tasks. But, they also fail significantly compared to the areas of those of independent GPU's as they are not able to be upgraded every year and they are not susceptible to driver updates from the manufacturer which mean that after 2 years they become slightly obsolete.
The main 2 Manufacturers of GPU's are Nvidia and ATI with many other smaller companies developing variations of their GPU cards.
No comments:
Post a Comment