Hardware components: everything you need to know

Hardware components is the set of physical elements that make up the computer. From the box to the motherboard, through all external peripherals for special applications.

In this document we study each component providing considerations on its specifications and benefits, and how these influence the operation and performance of the computer system.

Index of contents

Hardware components

The motherboard; and more specifically CPU, auxiliary integrated circuit, ROM memory, connection buses and CMOS battery, make up the indispensable processing units for the correct operation of any computer.

CPU or central processing unit

The CPU, also known as the central processing unit, is the element in charge of interpreting the instructions of the software . The computing power of our computer depends on it.

From its inception, not all CPUs are created equal. The materials and processes used to manufacture these elements have a decisive effect on the performance of the microprocessors.

Low cost production usually involves the use of thermal pastes, plastic insulators and alloys for pins or poorer quality solders; a saving that is detrimental to the quality, durability and reliability of the CPU. In summary, the use of suboptimal materials reduces the life expectancy of the part. This can lead to problems like:

• Bottlenecks when interacting with other components Inability to operate at maximum capacity Increased chances of failure when subjected to thermal or computational overstress Early component failure

When studying which CPU best suits our needs, another vitally important feature is the clock frequency. This specification limits the number of operations per second that the computer may undertake.

Today's high-end CPUs have clock rates between 3.5 and 3.8 GHz. Through the practice known as overclocking, it can exceed 4.5 GHz, but not all CPUs allow this technique. Manufacturers specifications indicate which models accept overclocking and which do not.

In older processing units the clock frequency was closely tied to computing power, two other characteristics of the CPU currently influence the true capacity of the system.

We are talking about the number of cores and the processing threads . The cores act like subprocessors: they cooperate to divide the tasks in which the computer works. Threads optimize waiting times between operations of the same task. In a multitasking oriented computer, multicore processors take on greater relevance, whereas in raw computing applications, multithread is the preferred option.

The user-level CPUs available on the market have 4 to 16 cores (new models we will see soon), with single-core and multithread models.

Another important aspect of the central processing unit is internal memory. Although the CPU takes instructions directly from RAM, it also has cache memory. Cache memory time and energy spent reading and writing information that is needed repeatedly. The larger the available cache memory, the better the performance of the drive.

Modern CPUs typically have their cache memory tiered. The basic level or L1 is associated with a particular nucleus; L2 and higher levels can cater to all or some of the threads. The actual operation depends on the topology of the memories. The upper (or external) level always interacts with all cores, while the lower levels are linked to individual cores or groups of cores.

L3 is the current standard in retail equipment, but the L4 CPU cache is also a reality. In addition there are special caches more or less appropriate depending on the application: WCC, UC, smart cache, etc.

Another relevant aspect of CPUs is word size. Word size measures the maximum length of instructions that the CPU can receive from RAM. The older the better.

Finally, it is interesting to know what is the power demanded by the central processing unit. In special applications, consumption can be one of the decisive factors when choosing one or the other CPU: in computing centers, small differences in consumption can have very different economic performances.

Considering the electrical aspect of the unit, it is worth knowing also the efficiency with which the received energy is used. Low efficiencies point to large heat losses, which force the use of better cooling systems on the equipment. Recall that the optimal performance of the CPU occurs in the thermal range of 30 to 50 degrees Celsius, although most computers tolerate up to 80ºC without marked changes in performance.

Auxiliary Integrated Circuit

The auxiliary integrated circuit is made up of a series of specialized chips for audio, video and control applications. Formerly it was made up of more than a dozen small chips, but today its architecture has been profoundly simplified, with three well-differentiated blocks: the north bridge, the south bridge and the connection between bridges.

The chip that makes up the north bridge is also known as the northbridge , Memory Controller Hub (MCH), or memory controller hub. It has the tasks of controlling the memory, the PCI Express and the AGP bus, as well as serving as the data transmission interface with the chip of the south bridge.

Modern Intel CPUs include memory control and PCI Express functions, the north bridge is unnecessary. At AMD there is northbridge , but it is only in charge of controlling the AGP or PCI Express; memory controllers are integrated into the processor. Older chipsets have an even more inefficient architecture in which various buses are used to control RAM and the graphics card.

It is important to know the structure of the north bridge, the number of PCIe point-to-point lanes (x1, x4, x8, x16 and x32 are the usual ones) and the transfer speed of the connection before acquiring the chipset .

The PCI-SIG standard associates each denomination with a unique bandwidth makes it easy to know the component specifications. The first generation of PCI Express, the PCIe 1.0 released in 2003, features data transfer rates of 2.5 GT / s; the PCIe 5.0 released this year reaches 32 GT / s.

To choose a PCIe connector it is necessary to know what use it will be given. The following list gives a general idea of ​​the lanes required by different hardware components:

• 1 lane: network drivers, audio, USB connectors up to 3.1 Gen. 1.2 lanes: USB 3.1 Gen. 2 and higher, SSD drives. 4 lanes: firmware based RAID controllers, Thunderbolt applications, M.2 expansion cards (old NGFF).8 or 16 lanes: specialized PCIe cards, graphics cards.

The number of total lanes of the auxiliary integrated circuit or CPU is relevant when the number of connected components is expected to be high. Today's high-end models have up to 128 lanes.

Returning to the general outline of the chipset , another of the basic blocks that make it up is the south bridge. This is also known as the Southbridge , I / O Controller Hub (ICH), Platform Controller Hub (PCH), I / O Controller Hub, or Platform Controller Hub.

The south bridge controls input and output devices, as well as integrated audio, network, and imaging equipment. Below is the complete list of these elements:

• Storage ports (SATA and parallel) USB ports Integrated audio Integrated local area network PCI bus PCI Express lanes Real-time clock RTC CMOS memory or ROM: BIOS and Unified Extensible Firmware Interface (UEFI) Chip Super I / O (for DMA control, PS / 2 and other outdated technologies)

Finally, the north bridge and the south bridge are linked through a PCI connection known as an inter-bridge. If this element presents poor transfer speed, it will form a bottleneck in the auxiliary integrated circuit.

Each processor company presents its own solution. At Intel there is a dedicated connection known as Direct Media Interface or DMI, similar to a full-duplex PCIe. It achieves a bandwidth of 1 GB / s per direction, or 10 Gbps between the four peer-to-peer lanes that configure the DMI. AMD uses an information path known as A-Link with three versions: Basic, II, and III. These are PCIe 1.1 and 2.0 lines (for A-Link III) with four lanes.

ROM memory

ROM or read-only memory is an internal piece of hardware that is usually built into the motherboard.

It cannot be modified (or at least not easily) so it usually contains the firmware that allows the equipment to function. Its storage capacity is limited. Modern computers have 4, 8 or 16 Mb, enough to host the SMBIOS code, responsible for initializing basic processes in the computer such as activating POST, detecting hardware , establishing the basic execution environment or loading priority RAM paths.

The ROM has changed over time, from being unalterable memory (MROM) to operating as flash memory. The different types of ROM available today are:

• Programmable read-only memory (PROM) or one-time programmable (OTP). Reconfigurable with specialized equipment. It offers the highest security as it is resistant to rootkit attacks. Programmable and erasable read-only memory (EPROM). Allows up to 1000 erase and rewrite cycles. They are usually equipped with a label that protects them from ultraviolet light (UV erases information). Electrically erasable programmable read-only memory (EEPROM). The most common in current commercial applications. They are slower than traditional ROM memories. Flash memory is a particular type of EEPROM that is faster and stronger (supports up to a million erase and rewrite cycles). It is also worth mentioning the EAROM subtype, slow but more secure.

The main specifications of RAM memory units are: read speed, write speed, resistance and robustness of storage against high temperatures and radiation emissions.

Storage units in hardware components

Although ROM is seldom handled outside of the chipset environment, its inclusion within this segment could be argued. We have preferred not to do so to protect the prominence of RAM memory cards and physical storage units, blocks that we investigate in the following sections.

RAM

RAM or random access memory is a storage device that allows you to accelerate the speed of access and reading of information in use. They minimize the time used to obtain the required data.

RAM differs from physical storage units in that it is volatile: Stored memory is lost when the power goes out.

This hardware has undergone multiple evolutions since its conception in 1959 (MOS transistor, also known as MOSFET). Currently, RAM comes in two main branches: SRAM or static RAM and DRAM or dynamic RAM.

The first group concluded its evolution in 1995 with a 256 Mb device developed by SK Hynix, at that time Hyundai Electronic Industrial. DRAM reached up to 4 Gb in 2011 at the hands of Samsung, and then it derived in new technologies such as synchronous dynamic RAM or SDRAM that in its DDR2, DDR3, LPDDR2, LPDDR3, LPDDR4 and LPDDR5 types are widely used today; or synchronous graphics RAM and high-bandwidth memory (HBM and HBM2) which are also in force.

Different typologies have very different specifications that make them incompatible with each other.

The latest developments in RAM are the GDDR5X and GDDR6 types, the technology used in Nvidia's Ray tracing applications.

Another possible classification refers to SIMM (Single In-line Memory Module) memories and their evolution: DIMMs (Dual In-line Memory Module). Modern RAM memory cards are included in this last family. Laptops are often equipped with smaller memory sizes called SO-DIMMs (only the form factor changes, not the technology).

The most important RAM specifications are: capacity, capacity limit tolerated by the installed operating system, frequency and latency.

RAM limits the number of running processes on the computer. The operating system contains an address known as swap or swap space, which can come in the form of a file or partition. This item helps manage data from RAM when the random access memory in use is close to being fully occupied. This excess available RAM is known as virtual RAM; the name should not be misleading since this memory is located on the SSD or HDD and does not have the defining characteristics of RAM.

When the available RAM is exceeded, this file increases its weight. When the defined weight limit is exceeded, errors appear. In general, operating with RAM memory to the limit slows down computer processes and is not recommended, both from a performance and hardware preservation point of view.

It should also be known that memory that has gone through a period of inactivity in RAM may be compressed. This state is sometimes known as ZRAM (Linux) or ZSWAP (Android). This prevents disk paging (with much lower read and write speeds) and increases RAM performance. Optimized use of this technology allows you to get the most out of installed RAM without the need for hardware expansion.

Physical storage drives

Currently within this category, only the HDD or SSD on which the OS is installed can be considered as the main hardware . There are also hybrid applications known as hybrid hard drives or SSHD, but their use is not widespread.

HDDs or hard drives are storage elements that use an electromagnetic data accumulation system. The information is recorded on a rotating disk known as a platter thanks to the action of the read and write head.

The capacity of HDDs is greater than that of other storage devices. Currently there are already 20 terabyte models, although the 4, 6 and 8 TB corresponding to the previous generation are more common.

Apart from the capacity, there are other characteristics of the HDD that should be known:

• Error rates and correction firmware . The more resistant the system is to the introduction of errors in the accumulated bits, the greater reliability the component will have. Today many hard drives use code to alleviate typing errors. Thus, a hardware protected partition is assigned for Error Correction Codes (ECCs), Low Density Parity Checks (LDPC), or private manufacturers' software . Rotational speed. It measures the number of revolutions per minute of the disc. Modern models use engines of up to 7200 rpm. At higher rotational speed; faster reading and writing speed, electrical consumption, noise produced and physical wear. Search time, rotational latency and data transmission speed. They affect the speed of reading and writing. The first two are physical impediments to the structure of the hard drive; they depend on the position of the plates to be read and the location of the read and write head. The data transmission rate acts as a bottleneck when the connectors are inadequate. Form factor. This is a ratio of the size of the HDD envelope. We have to choose a form factor that can be attached without problems on our tower or laptop. Connection interfaces and buses. The buses used by contemporary computers are ATA, Serial ATA (SATA), SCI, Serial Attached SCI (more commonly known as SAS), and Fiber Channel or FC. Auxiliary equipment. They are components that are an inseparable part of HDD: temperature sensors, filters, adaptations for demanding atmospheres…

HDDs have been used in desktop computers, laptops and consumer electronics not only to accumulate information, but also to install the operating system and software that is used on a daily basis. However, in recent years a new technology based on flash memory has begun to displace this element in its most basic function, that of hosting OS.

We are talking about SSDs or solid state drives. This is persistent storage that improves several properties of traditional HDDs: they are silent, they do not have moving parts that can degrade with use, their speed of reading and writing is higher, and their latency is lower. Its only drawback is the price, and it continues to drop.

SSDs consist of controllers, the memory unit, a cache or buffer, a battery or supercapacitor, and a connection interface with the equipment. The controller is one of the most relevant elements since the number of NAND chips that make it up establishes the reading and writing speed of the device.

The SSD supports around a million rewrites. Depending on the range that is accessed, it is equipped with non-volatile NAND flash memory or with triple, quad or multi-level cell flash memory (TLC, QLC and MLC) that are cheaper and have worse features. There are also items on the market with memory based on DRAM, 3D Xpoint (Intel and Micron technology), NVDIMM (Hyper DIMM) and ULLtraDIMM. The speed of the SSD depends on the type of memory used; the best option is DRAM.

The available data transfer interfaces are: SAS, SATA, mSATA, PCI Express, M.2, U.2, Fiber Channel, USB, UDMA (or Parallel ATA) and SCSI.

In general, SSDs are more robust, durable and faster, hence the current preferred option.

Hardware components of input peripherals

It is understood as peripheral input to the external equipment to the computer tower that allows the introduction of information to the system. Within the main hardware we must consider the keyboard and the mouse.

Keyboard

The keyboard features a collection of keys (matrix) that enables you to enter commands into the system and perform certain predefined operations. The keyboard has a microprocessor that transforms the signals that arrive from the matrix to electrical information interpretable by the equipment to which it is connected.

There are different types of keyboards on the market depending on the utility that will be given:

• Flexible keyboards roll up or fold down to take up little space. These specialized wraps are highly appreciated by travelers, who save space on their bags. They are also used in environments where the level of cleaning required is very high (laboratories and hospitals, to name a few cases). The projected keyboards work thanks to a projector, cameras and sensors. The matrix image is projected onto a flat surface and hand movement is captured on it. They are still insufficiently developed, but they are used in the same applications as the previous ones. Another case of specialized keyboards are those of the gaming segment. The most appreciated are those that come equipped with mechanical keys, although the ability to configure shortcuts , macro programming, simultaneous key registration and aesthetics are also appreciated. The transmission latency of these devices is very low to minimize the impact on the user's games. In keyboards for drafting, programming or databasing , the resistance of the keys is lower to avoid injuries associated with efforts by repetitive movements. They also allow a more comfortable position of the hands on the device to decrease the incidence of carpal tunnel syndrome. Ergonomics is one of the fundamental factors in the design of these models.

The use that will be given to the keyboards is not the only factor that allows a classification. According to the method of connection with the computer we differentiate wired and wireless keyboards. The latter use a wireless connection via Bluetooth, wifi, radio or infrared. The former use USB or PS / 2 cabling.

The mechanism behind the operation of the keys also allows for fundamental differentiation. There are mechanical keys, classic keys, membrane keys, and chiclet keys (rare).

The first ones deserve a separate paragraph. The mechanical keys have an individual push-button switch that improves the precision of the device. Multiple switches are available: Cherry Mx (most popular), Razer, Kailh, Romer-G, QS1, and Topre. When buying mechanical keys you have to consider its actuation point, travel, percussion sound and weight.

A little-known benefit of mechanical keyboards is the ability to replace broken keys individually without parting with the entire keyboard. This positively affects the longevity of the equipment, making mechanical keyboards an environmentally responsible option.

Finally, the keyboard layout should be considered. Term that refers to the available keys and their position in the matrix; topology that varies geographically as follows:

• AZERTY: specially designed for francophone countries, with combined French, Belgian and Arabic variants (present in North African countries such as Morocco, Algeria or Tunisia). QWERTY: the most common distribution, available in German, Spanish and Japanese versions. QWERTZ: used in German-speaking countries almost exclusively: Germany, Austria, Switzerland… Limited use distributions: Colemark, Dvorak, HCESAR… Special distributions: Braille and the like

Hardware components focused on the d

The mouse is a small pointing device designed to be guided on a flat surface with the palm of the hand. It is an ergonomic device with several buttons, a motion capture system, a controller and an information transmission system.

Depending on the characteristics of some of these constituent elements, mice can be classified in different ways.

According to your transmission system:

• Wireless mice. They use wifi, radio frequencies, IR or Bluetooth to exchange information with the computer. Wired mice. They use a USB or PS / 2 port to connect to the tower.

According to its motion capture system:

• Mechanic They have a rigid rubber ball at the bottom that moves by activating two internal wheels that work as a sensor when the user moves the mouse over the surface on which it rests. It has poor durability characteristics due to the presence of moving elements, being especially susceptible to jamming due to dirt accumulated in the mechanisms. Opticians. It achieves an accuracy of 800 dots per inch (dpi or dpi). They are more durable, but require a mouse pad to function properly. Laser. Evolution of the previous one that provides higher dpi values: up to 2000 dpi. They are preferred by professional video game players and graphic designers. Trackballs . Similar to the mechanical mouse. The buttons have priority over the movement of the device. The rubber ball migrates to the top of the mouse and its control is assigned to the plex. Multitouch. It is a hybrid between a mouse and a touchpad .

When choosing a mouse ergonomics is important. In this sense, gaming mice usually offer the greatest configuration possibilities: distribution of installed buttons, resistance opposed by the buttons, dimensions of the grip envelope, etc.

WE RECOMMEND YOU DRAM Calculator for Ryzen: What is it, what is it for and configure it

Touchpads

It is a touch panel that fulfills the functions of the mouse in computer equipment such as netbooks and laptops.

Given its analogous functions, the touchpad also has buttons that allow you to control the computer. Although, the most important part is the touch zone. This detects the position of the finger calculating the electrical capacity present in the different points of the region. Accuracies of 25 microns are achieved.

Some touchpads have multitouch technology that allows multiple fingers to be used simultaneously to operate the system with greater control. Others allow to quantify the pressure used.

Touch screen

Some netbooks integrate touch control functions on the screen. Usually this solution is more common in mobile phones, tablets and consumer electronics.

Touchscreens can be resistive, capacitive, and surface acoustic wave. The former are the cheapest and most accurate, but their brightness is 15% less and they are thicker. Capacitive functions like previously documented touchpads . The weaker acoustic waves use sound localization.

Output devices

They are all those elements that present useful information for the user. In this article the only one that we consider to be strictly necessary is the monitor.

Monitor

It is a screen that converts bits of information into visual elements that are easily interpretable by the user.

There are multiple technologies used in monitors: cathode ray tube (CRT), plasma (PDP), liquid crystal (LCD), organic light emitting diodes (OLED) and lasers.

The specifications that matter to us in these peripherals are:

• Screen resolution. Currently it is rare to find screens with a resolution less than 1280 × 768 pixels (high definition or HD). Some common resolutions available in the market are Full HD, Retina Display and 4K. Resolution defines the aspect ratio of the image and the screen dimensions that can be used without losing perceived definition. Refresh rate. Also known as the refresh rate or vertical sweep frequency, this specification refers to the number of frames that can be displayed on the screen every second. The higher the number, the better the perceived fluency. Common refresh rate values ​​are 60, 120, 144, and 240 Hz. Size. It is measured in inches on the largest diagonal of the rectangle that forms the screen. Also the geometry has relevance, there are new generation screens with a concave design from the user's perspective that improve immersion by giving a more panoramic feeling; It is an optimal solution for media playback applications. Response times and latency. It measures the time from when the computer has certain information until it is presented. It is relevant in the competitive video game scene, among others. Technology panel. Configuration of connections, color correction, selectors for parameters, etc.

Power supply and other elements

For the equipment to function properly, an electrical power source is needed that is capable of supplying the required energy. The power supply is integrated into the tower and must be dimensioned considering the voltage demand of the computer components. These sources can be modular and semi-modular, and their nominal voltage is usually between 150 and 2000 watts.

The computer case and racks for special applications are support structures for the processing and storage components. It is questionable whether they are part of the main hardware , but we include them here as well.

Finally, taking into account the same details as in the previous paragraph, the inclusion of refrigeration in this section can be justified. The cooling system is the set of elements that keep the computer's temperature at acceptable values.

Cooling can be accomplished using fans, radiation plates, coolant lines, or a combination of the above. Effective heat dissipation is the most important parameter of these systems, but it is also important to know the useful life, the noise generated and the complexity of the installation.

Hardware components

Within this group we will talk about the GPUs, the NIC and the expansion cards, elements that allow to expand the capacities and computing power in certain uses, but dispensable for basic applications.

GPU or graphics processing unit

The GPU is a coprocessor specially developed to work with graphics and floating point operations. It works in parallel with the CPU dividing in the work according to the implied information.

The most important parameters of a GPU (rarely referred to as a VPU) are the triangles or vertices drawn per second (it limits the complexity of the graphics it works with) and the pixel fill speed (which tells us how fast they apply. the textures on the drawn geometry). The GPU's clock frequency, the size of its memory bus, and other CPU and chipset parameters define how many frames per second the GPU can generate. This value is the third determining specification when speaking of graphic processing units.

Depending on the specific GPU model, it is also interesting to know the technology with which it can work and if it is possible to install several units in parallel (SLI).

NIC or network card

This hardware component receives many different names: network interface card (TIR), network interface controller (NIC), network adapter, network card, physical network interface, LAN adapter or, simply, network card, its name most common in Spanish.

It is an adapter that connects a computer equipment to a public or private computer network, so that the different connected systems can share information and resources with each other.

NICs can use different technologies to transfer information packets: polling , controlled IRQ-I / O, programmed I / O, DMA, third-party DMA, bus mastering …

When choosing a network card that meets the needs of the Internet user, you must pay attention to its transfer speed (limited by the equipped buses -PCI, PCI-X or PCIe-), the technology used, the types of network it supports and the connectors installed as standard (SC, FC, LC, RJ45…).

Expansion cards

These are devices with chips and drivers that increase the performance of the computer when connected. Both the network card and the GPU can be considered, in the most general sense of the term, expansion cards. Also included in this group are the following hardware :

• Sound or audio cards Graphics cards Internal modems Radio tuner cards

Storage units

When storing information, two aspects are important: having as much memory as necessary and ensuring that the information is not lost over time. In this sense, external storage units allow us to increase our memory capacity, while optical readers give us access to discontinued saving formats.

Optical reading units

This is hardware capable of reading out-of-date or abandoned storage devices: floppy disks, CDs, DVDs, etc. They are composed of mechanical elements such as motors and reading heads in a very similar way to those already defined in the case of hard disk drives.

External storage drives

In this case, we are talking about additional memory spaces, either in HDD, SSHD or SSD format that are attached to the computer via USB or similar connectors. They can be individual components or form large-capacity structures known as SAS, SAN, or NAS.

Output, input and I / O peripherals

Two of the most common items among companion peripherals are the headphones and the printer. There are many other important peripherals such as the fax, the webcam, the digitizing tablet… but covering all of them in detail could fill a book. In the following paragraphs we stick to the two devices already mentioned.

Headphones

The preferred option to enjoy audio files. With headphones we can set the maximum volume without disturbing those around us. Many headsets available in computer stores today are equipped with a microphone that favors telematic conversations.

To choose a good earpiece, the fidelity of the sound, the power developed by the integrated speakers, the transfer speed of the connections and wiring, and the ergonomics of the device are relevant aspects.

The only alternative to headphones is speakers, but they invade the space of other users.

Printers

This peripheral transforms virtual information into physical written or illustrated documents. Its use is declining as paper is abandoned, but it is still widespread.

Along with scanners, cameras, and webcams, one of the most important specifications for printers is the definition they work at. In the case of printers it is often referred to as dots per inch (dpi or dpi). The type of printing technology also matters:

• Inkjet printing. They are cheap but they consume ink quickly and the spare parts make the service provided extremely expensive. Laser printing (toner). They require a large initial investment, but they are worth it in the long term given their low consumption. Less common printing methods: solid ink, impact, dot matrix, sublimation ink, etc.

Final words and conclusions about hardware components

Since the printer is a hardware with moving parts, when buying one it is advisable to make sure that its construction is robust. Deciding on widely known manufacturers is always recommended.

We recommend the following guides:

• Best processors on the market Best motherboards on the market Best RAM memory on the market Best graphics cards on the market Best SSDs on the market Better chassis or PC cases Better power supplies Better heatsinks and liquid coolers

Do not miss it!

So we close this extensive article on hardware components . The main components necessary for the computer to work as well as the most common accessories have been thoroughly covered. We hope it has helped you.