Powered by Blogger.

Featured

Language/Idioma

Computer hardware

Tuesday, March 12, 2013



When talking about computer hardware we refer to the physical components, all of their electronic circuitry and mechanical parts.  Generally speaking, no matter the type of computer, we may clasiffy the hardware components as follows: The central Processing  unit, Iput/Output peripherals and other peripherals.
 
CPU (Central Processing Unit)



The central processing unit is a conjunction of electronic devices conected together that are responsible of performing all of the data processing. These devices are generally found inside a metallic case, however newer computers, tablets and handhelds have all of their processing electronics integrated inside the same case as the screen.

The CPU, case or processing electronics can be viewed as the computer itself. This is because the CPU can continue processing information even with no peripherals attached. I always rather to call this block "processing electronic" because many people and even the hardware manufacturers themselves use the CPU acronym to name the microprocessor.


Input/Output (I/O) Devices
The I/O devices make possible the human-computer interaction. The essential input devices are the keyboard and the mouse. On the other hand, the essential output device is the computer monitor or screen. Below, every peripheral device that can be found is listed:

 
Input
  • Keyboard
  • Mouse
  • Joystick
  • Web cam
  • Microphone
  • Optical Pencil
  • Scanner
  • Digital camera
 
Output
  • Monitor
  • Printer
  • Speakers
  • Proyector
  • Headphones
 
Other peripherals
Among the peripherals, we may find
Storage Peripherals
    • External hard drive
    • CD/DVD/BluRay external drives
    • Flash USB pendrives and SD cards
    • Tape backup drives
 
Communication peripherals
    • Modem
    • Wireless adapter Wi-fi or bluetooth
    • Radio/TV Tuner
 In this web we will focus mainly on the hardware that is found in desktop computers, because in general desktops are the base desings that manufacturers use to develop  other types of computers such as laptops and tablets.

Below the most important components inside a computer are detailed.
  • Microprocessor
  • Motherboard
  • Video Card
  • RAM Memory
  • Storage devices
  • I/O devices
  • Peripherals



What is a computer?

Tuesday, February 26, 2013


A computer is an electronic device capable of processing data at blazing-fast speeds and storing massive amounts of information. A computer executes, at their lowest level, basic logic-arithmetic computations, but also performs move, copy and store operations. Such simple operations allow creating the complex programs and applications we use today.

Although computers execute rather simple operations, they process data by using a language that can be unintelligible for the average human. While people communicate through language symbols and decimal numbers, computers use the binary numeric system in order to execute all of the processing and storing operations. The binary system has only two symbols : 0 and 1. By using just those two symbols we can represent any numeric quantity or any symbol of any alphabet. It is clear that human beings are not able to directly understand binary machine language, for this reason it is necessary that computers convert their binary information into something humans can understand. The binary information is then converted by a computer into text, images, and sounds.



Now, let us put things into perspective. A latest generation mainstream computer is capable of performing operations at blazing-high speeds when compared to humans. For example, Let us consider a basic operation such as the addition of two 8-bit numbers -8-bits represents a number between 0 and 255 -. While an average human can spend between 2 to 5 seconds to obtain the result of such operation, a latest generation computer with a 3GHz CPU can obtain the same result in less than 0.33 nanosecods. This means that such computer is 15 million times faster than a human brain when performing logic-arithmetic operations.

But, how is it possible to achieve such overwhelming calculation speed? How do computers work? I will start by saying that such amazing processing speed can be achieved because computers are built using very fast electronic components that work using high frequency electrical signals. Such components have been in constant evolution since Jack Kilby created the first integrated circuit in 1959. Today, the most advanced components are capable of working at GHz frequencies.

When talking about computer, we may distinguish between two essential components: Hardware and Software. The hardware refers to the physical components of the machine, all the electro-mecanical parts that are in charge of the processing and data storing. For example: The Microprocessor, the main memory, storage devices, the power supply, input peripherals, output, peripherals, etc.

On the other hand, the software refers to the programs. Those that are in charge of controlling the hardware and giving  essential instructions so we can do actual work in a computer. There are different types of software we can find, some examples are: The operative system, device drivers, office applications, etc.

In this website we will uncover what is under computer's hood and will delve into the details of hardware and software.

In don´t want to close this article without showing all the types of computers that exist. In the beginning of the computers era there were only massive electro-mechanical monsters that were built using vacuum tubes, and occupied entire rooms. It was with the invention of transistor and later the integrated circuit, that desktop and personal computers arised, followed by laptops and other mobile devices.

The types of computes we can find today are described as follows:


Supercomputers:
Those are the most massive and most powerful. Supercomputers have a processing power ranging the PetaFLOPS (a hundred trillion operations per second), they can fill entire rooms and consume huge amounts of power (in the order of KiloWatts). They are used mainly for research tasks such as simulation of complex physics and biological systems.


Servers:
Server computers hold an intermediate processing power. The main requirement for server applications is availability. For this reason they must use more reliable and durable components that the ones found in regular consumer PCs. Servers are used mainly to store data that needs to be accessed by multiple users through a network. Such a network may be a Local Area Network or the Internet itself.


Desktop Computers:
These are the regular computers that we use every day in our homes, our jobs, our university, etc. They are used for mainstream applications such as office work, games, graphic design, video editing, web browsing, video conferencing, chatting, etc.


Laptops y tablets:
The primary characteristic of a laptop is its ability to be carried wherever the user wants to go, this is defined as portability. In order to achieve a good degree of portability a computer must have reduced size and weight. In addition, they must have a good battery life and the ability to connect to wireless networks.

Tablets, on the other hand, are a variation from traditional laptops; the main difference is that tablets use touchscreens instead of physical keyboards. In general, tablets are considered devices with better portability features, because they are smaller, thinner, more lightweight, and have a much better battery life.  However, the increased portability poses an important trade-off: A  better portability implies, in general, reduced performance. For a given price range, a laptop tends to have less performance than a desktop computer, at the same time a tablet has way less computing capability and less functionality than a laptop.


Handheld Computers:
The main characteristic of such devices is that they can be held on the palm of a hand. The handheld devices find their main application as personal assistants/organizers. Today it is very difficult to find standalone handheld computers, however they can be widely found integrated 

Supercomputing companies

Wednesday, February 20, 2013

When talking about supercomputers the first names that come to mind are Cray, IBM, SGI, etc. Those are some of the most influential companies in the history and evolution of supercomputing. A list of the most important companies, their achievements, and systems are described in table below.

Cray is one of the most renowed supercompuer companies in the world, also known for becoming pioneer in such area with the release of their Cray-1 supercomputer in 1976[1, 2]. If we take a look at the latest supercomputer rankings we can find up to 3 Cray systems in the top 10: Jaguar, Cielo and Hopper. All of them holding a computing power above 1 PetaFLOP.
The third most powerful supercomputer has been produced by Cray, and is located in the National Oak Ridge Lavoratory  in USA: The Cray XT5 Jaguar. The Jaguar is powered by 224.256 AMD Opteron cores, wich means a computing capability of 1.75 PetaFLOPs. Such amount of power is distributed among 18.688 nodes, each node featuring two Hex-core AMD Opteron-2435 processors and 16GB of RAM.[3, 4, 5]
IBM is also one of the most reowned supercomputing companies in computer history. They became pioneers during the 60's when they delivered the IBM 7030 in 7961. This computer had a computing capability of 4MIPS. During the 90's IBM built a chess supercomputer named Deep Blue. This machine was able to beat Kasparov, the chess world champion by then. Deep Blue was conformed by 30 RS/6000 nodes, each one holding 30 P2SC microprocessors running at 120MHz. The whole system had a total computing power of 11.38 GFLOPS and was able to execute 200 million chess movements per second. Later, IBM produced a improved version: The Deeper Blue, that was able to beat Kasparov in a 6 game encounter. In 1999 IBM founded the Blue Gene
project. Such project was oriented to the desing of PetaFLOP capability supercomputers, but consuming limited amounts of power. The most important machines produced during the lifetime of this project were the Blue Gene/L, the Blue Gene/P an the Blue Gene/Q. All of these computers dominated the supercomputing rankings for many years. Te fastest IBM supercomputer nowadays is the RoadRunner. It holds the world's 10th place with a peak performance of 1.7 PetaFLOPS. The RoadRunner is a Hybrid design consisting of 12960 PowerXcell 8i -similar to PS3's cell CPU- and 6480 dual-core opteron processors. The total amount of RAM is 103.6TB [4, 5, 6, 7, 8].
NEC has been a consistent supercomputer designer and manufacturer since they developed the SX-1 and SX-2 systems in 1983. Between 2002 and 2004, NEC held the record of having the fastest supercomputer on the planet: The Earth simulator. This computer consisted of 2400 nodes with intel Xeon CPUs and NVIDIA Tesla processors. According to the latest ranking, the earth simulator is the fifth fastest supercomputer in the world with a computing power of 1.192 PetaFLOPS.[4,9]
One of the most remarkable supercomputers built by Fujitsu was the Wind Tunnel Simulator (WTS). It was developed in conjunction with the Japan Aerospace Laboratory and held the record of being the most powerful supercomputer  between 1993 and 1996 with a sustained performance near 100GFLOPS[11]. The WTS had 166 vector processors, each one with a peak performance of 1.7GFLOPS. Fujitsu is the developer and builder of the fastest supercomputer in the world: The K computer, which is located in RIKEN Advanced Institute for Computational Science in Japan. It has a total of 705124 cores distributed among 80.000 nodes. Each node has an 8-core SPARC64 VIIIfx processor and 16GB of RAM.[10]
Silicon Graphics Inc. is another well-known manufacturers of high-performance workstations and supercomputers, especially  in the entrertainment/movie industry. In 1999 the company ventured into the supercomputing industry by  acquiring Cray for a brief period of time [12]. In 2004 SGI built the record breaking Columbia supercomputer, located in the NASA AMES research center. This machine consisted of a cluster of 20 Altix supercomputers, each one with 512 Intel Itanium 2 processors. The computing power of the whole system was 42.7 TeraFLOPS. In 2011 Fujitsu delivered Pleiades, a super computer developed in conjunction with NASA AMES research center. This machine reached a peak performance of 970 TeraFLOPS becoming the seventh fastest computer in the world. Pleiades was built using Altix ICE nodes and the whole system holds a total of 111104 Intel Xeon processors. It is expected that by performing hardware updates, this computer will be able to reach the 10 PetaFLOPS mark.[13]
REFERENCES

[1] http://www.cray.com

[2]http://es.wikipedia.org/wiki/Cray-1

[3]http://es.wikipedia.org/wiki/Cray_XT-5

[4]http://www.top500.org/

[5]http://es.wikipedia.org/wiki/Cray_Inc.

[6]http://www-03.ibm.com/systems/technicalcomputing/

[7]http://en.wikipedia.org/wiki/IBM#Research_and_inventions

[8]http://en.wikipedia.org/wiki/Category:IBM_supercomputers

[9]http://en.wikipedia.org/wiki/NEC_SX_architecture

[10]http://www.fujitsu.com/global/about/tech/k/

[11]http://en.wikipedia.org/wiki/Numerical_Wind_Tunnel

[12]http://en.wikipedia.org/wiki/Silicon_Graphics

[13]http://www.nas.nasa.gov/hecc/resources/pleiades.html

Introduction to Supercomputers

Thursday, February 14, 2013

IBM Blue Gene Supercomputer


During the course of our lives we use to hear a lot about supercomputers. But do we really know what all is this  about? What is a supercomputer? What are their key characteristics? How are they different from our ordinary personal computers?

I will start by saying that a supercomputer is just a different type of computer; a computer that holds way higher computing power than the regular ones. However, this overwhelming computing power does not come without a price tag. A supercomputer consumes a great deal of power (in the order of  KWatts or MWatts), contains really expensive components (in the order of millions of dollars) and requires enormous amounts of space in order to accommodate, not only the computer components themselves, but also all the infrastructure that is needed for power and cooling.



After the concise definition of supercomputer we gave in the introductory paragraph, I guess we all wonder: How is it possible to obtain a computing power that is a hundred or a thousand times better than our desktop PCs using the existent microprocessor technology? The answer to this question lies in the huge computing power that can be achieved by taking advantage of a technique known as "massive parallel processing". Any supercomputer is made-up of hundreds or even thousands of smaller computers called nodes. Similarly, each node contains a set of multi-core processors that are capable of addressing large amounts of memory. Supercomputer nodes are generally designed in a slim form factor so they can be physically arranged inside vertical racks to form clusters. The next two pictures show the appearance of a node, and also the way how they are arranged inside the racks.




Supercomputer node

 Supercomputer nodes arranged inside racks

Now let’s put things into perspective. The computing power of a supercomputer is generally measured using the number instructions per second that it can process. The specific metric used for this purpose is the amount of Floating Point Operations per Second (FLOPS). Having this in mind, lets illustrate the huge difference in computing power between a regular PC and a supercomputer by using a simple example. While a really powerful personal computer equipped with a Hex-core Core i7 CPU and two Radeon HD5970 graphic cards has a potential computing power of 1.03 TeraFLOPS[1], the most powerful supercomputer, the K computer in Japan, holds a computing power of 10.51 PetaFLOPS[2]. If we analyze the previous data, we can conclude that the most powerful supercomputer has approximately 10204 times more computing power than a high end PC. Regarding the power consumption, while the consumption of high end desktop PC working at full load is on average 300 to 500W, a supercomputer consumes between 300 to 500MW on average [3].

A supercomputer requires a special cooling system consisting of a complex air conditioner array. The function of this cooling system is to keep the temperature of the nodes cold so they can work optimally. It is evident that the power consumption of the cooling system represents a significant contribution to the total power consumption of the whole supercomputer.



Now that we know the basics about supercomputers let us explore their applications. The supercomputers are used for scientific research such as in complex physics and biological simulations. Such applications take enormous advantage of massive parallel processing of supercomputers.

Let us enumerate some specific examples of supercomputing applications:



  • Climate modeling; Using supercomputers researchers can model past, present and predict future climate.
  • Space simulation: By using supercomputer space researcher are able to study the behavior of the sun, the stars and the space in general.

  • Supercomputers are used to simulate the effects of a tsunami over any given city or coast.

  • Supercomputers are used to simulate supernovas explosions

  • Supercomputers are used to test the aerodynamics of the most advanced military aircrafts

  • Supercomputers allow simulating protein folding, and how this folding can affect people suffering Alzheimer disease, and also many types of cancer.

  • Supercomputers are used to simulate nuclear explosions, eliminating the need of performing real nuclear tests.[3]

In this article we have learned about supercomputers, their characteristics, applications and physical appearance. It is evident that supercomputers have been really important tools for the advancement of science and technology. In spite of their high cost, even more powerful supercomputers will continue to be designed and developed because without them it would be nearly impossible to solve many of the critical scientific problems of the human kind.



REFERENCIAS

[1]http://en.wikipedia.org/wiki/FLOPS

[2] http://www.fujitsu.com/global/about/tech/k/

[3] http://es.wikipedia.org/wiki/Supercomputadora

Welcome




Welcome to this blog about computers and Information Technologies (IT). I will be writing really interesting articles about a wide range of topics related to computer hardware, computer networks and software. Enjoy!
 
Support : Creating Website | Johny Template | Mas Template
Copyright © 2011. ArchPC (EN) - All Rights Reserved
Template Created by Creating Website Inspired by Sportapolis Shape5.com
Proudly powered by Blogger