home › BY › Is it possible to disassemble the ssd disk. SSD for fast processors. What to do with an old hard drive

Is it possible to disassemble the ssd disk. SSD for fast processors. What to do with an old hard drive

On budget computers, even with a good processor and other components, manufacturers sacrifice hard drive speed for the sake of economy, relying on volume.

Replacing the HDD in a laptop with an SSD will speed up the computer and, if desired, you can get additional storage if you purchase a special adapter.

What you need to know

Before replacing, you can take care of transferring the system. If you plan to install a new OS, you can skip this step. Just transfer important files to a cloud service or USB flash drive.
If you want to change the drive while keeping the system, then the new memory storage must be large enough to hold all the necessary information.
If you have new laptop with a valid warranty, then after opening the laptop yourself you will lose it.

How to keep a copy of Windows

When replacing an old HDD with a new SSD in a laptop, many users think about how to transfer the system to a new drive. For this, special programs from laptop manufacturers have been developed.

Some of them:

Acer provides the "Acer eRecovery Management" utility;
in Sony - "VAIO Recovery Center";
Samsung has - Samsung Recovery Solution 5";
Toshiba satellite - "Recovery Disc Creator";
HP Recovery Manager;
Lenovo solution center
Asus has a "Backtracker" program;
MSI Recovery Manager

Over time, the list may grow. You can find and download new versions of programs from official sites.

You can also use universal ones: Macrium Reflect Free, Macrium Reflect. They are supported on all Windows operating systems.

For each program there is detailed instructions on the developers website, but basically all the functionality is the same: run the program, choose what and where to copy, wait until the process is completed. After replacing the disk, you will see the desktop as it was.

Starting to replace the hard drive

Below we will look at an example of replacing a hard drive with an SSD in Asus laptop. If your laptop is from another manufacturer, it's okay, the principle is always the same for most models.

Before disassembling the laptop, be sure to turn it off and remove the battery. And when working, try not to touch the components on the motherboard with a screwdriver or hands, even the slightest scratch can damage it.

Getting Started:

If you decide to install a new system after replacement, then use Windows 7 and higher, Windows xp and Vista are not designed to work on an SSD drive and you may experience a drop in write speed. Also, versions 10 and 8 of the system are most optimized for working on a solid state drive.
Otherwise, after you have installed the SSD, installing the OS will be no different from the usual one.

What to do with an old hard drive

1) The HDD can be installed as an additional storage instead of the DVD drive. They have long lost popularity and are practically not used.

To do this, you need a special adapter that is inserted into the drive. When choosing, pay attention to its height and width, since the dimensions disk drive depends on the thickness of the laptop itself. Also, the width of the adapter can also be different. Size mismatch won't hurt. work hard disk, but if you are a perfectionist, then this shortcoming will get on your nerves.

Connecting a hard drive instead of a drive is not difficult, usually along with an adapter comes instructions and the necessary tools. This method of use will be optimal for replacing a hard drive without reinstalling the system.

2) Or, you can buy an external case with a USB adapter and use HDD as portable storage.

First, let's look at what an SSD is. An SSD is a solid state drive (SSD, Solid State Drive or Solid State Disk), a non-volatile, rewritable storage device with no moving mechanical parts using flash memory. SSD fully emulates the work of a hard drive.

Let's see what the SSD has inside and compare with its close relative USB Flash.

As you can see, there are not many differences. Essentially, an SSD is a large flash drive. Unlike flash drives, SSDs use a DDR DRAM cache memory chip, due to the specifics of their work and the increased speed of data exchange between the controller and the SATA interface.

SSD controller.

The main task of the controller is to provide read / write operations, and manage the data layout structure. Based on the block placement matrix, which cells have already been written to and which have not yet been written to, the controller must optimize the write speed and provide maximum long term SSD disk services. Due to the nature of the construction of NAND-memory, it is impossible to work with each cell separately. The cells are combined into pages of 4 KB each, and information can be written only when the page is completely occupied. You can erase data in blocks that are equal to 512 KB. All these restrictions impose certain responsibilities on the correct intelligent algorithm of the controller. Therefore, properly configured and optimized controller algorithms can significantly improve the performance and durability of an SSD.

The controller includes the following main elements:

Processor - usually a 16 or 32 bit microcontroller. Executes firmware instructions, is responsible for mixing and aligning data on Flash, SMART diagnostics, caching, security.

Error Correction (ECC) - ECC error control and correction unit.

Flash Controller - includes addressing, data bus and control of Flash memory chips management.

DRAM Controller - addressing, data bus and DDR/DDR2/SDRAM cache memory management.

I / O interface - responsible for the data transfer interface to external SATA, USB or SAS interfaces.

Controller Memory - consists of ROM memory and a buffer. Memory is used by the processor to execute firmware and as a buffer for temporary data storage. In the absence of an external RAM chip, the memory acts as the sole data buffer of the SSD.

On this moment The following controller models are used in SSD:

Indilinx "Barefoot ECO" IDX110MO1

Indilinx "Barefoot" IDX110M00

Intel PC29AS21BA0

Marvel 88SS9174-BJP2

Samsung S3C29RBB01-YK40

SandForce SF-1200

SandForce SF-1500

Toshiba T6UG1XBG

Flash memory.

In SSD, as in USB Flash, three types of NAND memory are used: SLC (Single Level Cell), MLC (Multi Level Cell) and TLC (Three Level Cell). The only difference is that SLC allows you to store only one bit of information in each cell, MLC - two, and TLC - three cells (using different levels electric charge on the floating gate of the transistor), which makes MLC and TLC memory cheaper in terms of capacity.

However, MLC/TLC memory has a shorter resource (100,000 erase cycles for SLC, an average of 10,000 for MLC, and up to 5,000 for TLC) and worse performance. With each additional level, the task of recognizing the signal level becomes more complicated, the search time for the cell address increases, and the probability of errors increases. Since SLC chips are much more expensive and their volume is lower, MLC / TLC chips are mainly used for mass solutions. At the moment, MLC/TLC memory is actively developing and is approaching SLC in terms of speed characteristics. Also, low speed MLC / TLC manufacturers of SSD drives compensate by algorithms for interleaving data blocks between memory chips (simultaneous writing / reading to two flash memory chips, one byte each) by analogy with RAID 0, and low resource - by mixing and monitoring the uniform use of cells. Plus, a part of the amount of memory is reserved in the SSD (up to 20%). This is unavailable memory for standard write/read operations. It is needed as a reserve in case of cell wear, by analogy with HDD magnetic drives, which have a reserve for replacing bad blocks. The extra cell reserve is used dynamically, and as the primary cells physically wear out, a replacement cell is provided.

How an SSD drive works.

To read a data block in a hard drive, you first need to figure out where it is located, then move the block of magnetic heads to the desired track, wait until the desired sector is under the head and read. Moreover, chaotic requests to different areas of the hard disk affect the access time even more. With such requests, the HDD is forced to constantly “drive” the heads over the entire surface of the “pancakes”, and even reordering the command queue does not always save. And in SSD everything is simple - we calculate the address of the desired block and immediately get read / write access to it. No mechanical operations - it takes all the time to translate the address and transfer the block. The faster the flash memory, controller and external interface, the better faster access to the data.

But when changing / erasing data in an SSD drive, everything is not so simple. NAND flash memory chips are optimized for sector-based operations. Flash memory is written in blocks of 4 KB and erased in 512 KB blocks. When modifying several bytes within a certain block, the controller performs the following sequence of actions:

Reads the block containing the block being modified into the internal buffer/cache;

Modifies the necessary bytes;

Performs a block erase on the flash memory chip;

Calculates a new location of the block in accordance with the requirements of the shuffling algorithm;

Writes a block to a new location.

But once you have written information, it cannot be overwritten until it is cleared. The problem is that the minimum size of the recorded information cannot be less than 4 KB, and data can be erased at least in blocks of 512 KB. To do this, the controller groups and transfers data to release the entire block.

This is where the optimization of the OS for working with the HDD comes into play. When files are deleted, the operating system does not physically clean up the sectors on the disk, but only marks the files as deleted, and knows that the space occupied by them can be reused. This does not interfere with the operation of the drive itself, and interface developers did not care about this issue before. If this method of removal helps to improve performance when working with an HDD, then when using an SSD, it becomes a problem. In SSDs, like traditional hard drives, data is still stored on the drive after it has been deleted by the operating system. But the fact is that the solid state drive does not know which of the stored data is useful and which is no longer needed and is forced to process all occupied blocks according to a long algorithm.

Read, modify and write back into place, after clearing the memory cells affected by the operation, which, from the point of view of the OS, have already been deleted. Therefore, the more blocks on the SSD contain useful data, the more often you have to resort to the procedure read>modify>flush>write, instead of a direct write. This is where SSD users are faced with the fact that the speed of the disk noticeably decreases as they fill up with files. The drive simply does not have enough pre-erased blocks. Pure drives demonstrate maximum performance, but during their operation, the real speed gradually begins to decline.

Previously, the ATA interface simply did not have commands to physically clean up data blocks after deleting files at the OS level. For HDD, they simply were not required, but the advent of SSD forced us to reconsider our attitude to this issue. As a result, the ATA specification introduced a new DATA SET MANAGEMENT command, better known as Trim. It allows the OC at the driver level to collect information about deleted files and transfer them to the drive controller.

During periods of inactivity, the SSD independently cleans and defragments blocks marked as deleted in the OS. The controller moves the data so as to get more previously erased memory cells, freeing up space for subsequent recording. This makes it possible to reduce delays that occur during work.

But to implement Trim, this command must be supported by the drive firmware and the driver installed in the OS. At the moment, only the latest SSD models "understand" TRIM, and for older drives, you need to flash the controller to enable support for this command. The following operating systems support the Trim command: Windows 7, Windows Server 2008 R2, Linux 2.6.33, FreeBSD 9.0. For other operating systems, you need to install additional drivers and utilities.

For example, for an Intel SSD, there is special utility SSD Toolbox, which can perform the synchronization procedure with the OS on a schedule. In addition to optimization, the utility allows you to perform SSD diagnostics and view SMART data for all computer drives. Using SMART, you can evaluate the current degree of SSD wear - the E9 parameter reflects the remaining number of NAND cell cleaning cycles as a percentage of the standard value. When the value, decreasing from 100, reaches 1, we can expect the appearance of “broken” blocks soon.

About SSD reliability.

It would seem that there are no moving parts - everything must be very reliable. This is not entirely true. Any electronics can fail, and SSDs are no exception. With a low resource of MLC chips, you can still somehow fight with ECC error correction, redundancy, wear control and mixing of data blocks. But the biggest source of problems is the controller and its firmware. Due to the fact that the controller is physically located between the interface and the memory chips, it is very likely to be damaged due to a failure or power problems. At the same time, the data itself, in most cases, is saved. In addition to physical damage, in which access to user data is impossible, there are logical damage, in which access to the contents of the memory chips is also violated. Any, even a minor mistake, bugs in the firmware, can lead to a complete loss of data. Data structures are very complex. Information is "smeared" across multiple chips, plus interleaving, making data recovery quite a challenge.

In such cases, the controller firmware helps to restore the drive with low-level formatting when the service data structures are re-created. Manufacturers are constantly trying to improve the firmware, fix bugs, and optimize the performance of the controller. Therefore, it is recommended to periodically update the firmware of the drive to eliminate possible failures.

SSD security.

In an SSD drive, as in an HDD, data is not deleted immediately after the file has been erased from the OS. Even if you overwrite the file with zeros at the top, physically the data still remains, and if you get the flash memory chips and read them on the programmer, you can find 4kb file fragments. It is worth waiting for a complete erasure of data when an equal amount of data is written to the disk. free space+ amount of reserve (approximately 4 GB for a 60 GB SSD). If the file ends up on a “worn out” cell, the controller will not soon overwrite it with new data.

Basic principles, features, differences in data recovery from SSD and USB Flash drives.

Restoring data from SSD drives is a rather time-consuming and lengthy process compared to portable flash drives. The process of finding the right order, merging the results and choosing the right assembler (an algorithm/program that fully emulates the operation of an SSD drive controller) to create a disk image is not an easy task.

This is primarily due to the increase in the number of chips in the SSD drive, which greatly increases the number options actions at each stage of data recovery, each of which requires verification and specialized knowledge. Also, due to the fact that SSDs are subject to much more stringent requirements for all characteristics (reliability, speed, etc.) than mobile flash drives, the technologies and methods for working with data used in them are quite complex, which requires an individual approach to each solution and the availability of specialized tools and knowledge.

SSD optimization.

1. In order for the disk to serve you for a long time, you need to transfer everything that changes frequently (temporary files, browser cache, indexing) to the HDD, disable updating the last access time to folders and directories (fsutil behavior set disablelastaccess 1). Disable file defragmentation in the OS.

2. Before installing Windows XP on the SSD, when formatting the disk, it is recommended to “align” the partitions to a multiple of the power of two (for example, diskpart utility), otherwise the SSD will have to do 2 reads instead of one. In addition, Windows XP has some problems with supporting sectors larger than 512kb (SSDs use 4kb by default) and the resulting performance problems. Windows Vista, Windows 7, latest versions Mac OS and Linux already align disks correctly.

3. Update controller firmware if old version does not know the TRIM command. Install latest drivers on SATA controllers. For example, if you have an Intel controller, you can increase performance by 10-20% by enabling ACHI mode and installing the Intel Matrix Storage Driver in the operating system.

4. You should not use the last 10-20% of the free space of the partition, because this may adversely affect performance. This is especially important when TRIM is running, as it needs space to regroup data: for example, defragmentation utilities seem to work, because they also need at least 10% of the disk space. Therefore, it is very important to monitor this factor, because due to the small volume of SSDs, they fill up very quickly.

Benefits of an SSD

High reading speed of any data block, regardless of physical location (more than 200 Mb / s);

Low power consumption when reading data from the drive (approximately 1 watt lower than HDD);

Reduced heat dissipation (internal testing at Intel showed that laptops with an SSD heat up 12.2° less than those with an HDD, it was also tested that laptops with an SSD and 1 GB of memory in common benchmarks are not inferior to models with an HDD and 4 GB of memory);

Noiselessness and high mechanical reliability.

Disadvantages of an SSD

High power consumption when writing data blocks, power consumption grows with the growth of the storage capacity and the intensity of data changes;

Low capacity and high cost per gigabyte compared to HDD;

Limited number of write cycles.

Conclusion.

Due to the high cost SSD drives and a small amount of memory, it is not practical to use them for data storage. But they are perfect as a system partition on which the OS is installed and on servers for caching static data.

1 - SATA interface

SSD drives communicate with the computer via the SATA interface. Therefore, for tuning, the SATA hard drive in a PC or laptop can be replaced with a faster SSD drive. At the same time, the interface version is important: most older models have a SATA 2 connector, which theoretically provides a maximum speed of up to 300 Mb / s. Modern SSDs typically offer a SATA 3 interface (also called SATA 6 Gb/s) with a maximum data rate of 600 Mb/s.

2 - Controller

The controller is the "brain" of the SSD, it controls the data exchange between the SATA interface and the memory modules. The more powerful the controller, the faster the SSD will run. For example, Marvell 88SS9174 can read or write up to 500 MB of data per second. To prevent premature wear of the SSD, the controller distributes write operations so that all memory cells are used as often as possible.

3 - Buffer memory

To increase speed, SSDs have an intermediate buffer that is several times faster than flash memory. In most models, buffer memory ranges from 256 to 512 MB and, like PC RAM, consists of DDR3 modules. Frequent write operations to the same areas of memory are taken over by the cache memory. This reduces the number of flash write operations and increases the life of the SSD.

4 - Flash memory

Each memory module in an SSD contains billions of flash memory cells. The tiny structures in a memory chip (such as data paths) are only 34 nm wide. For comparison: a human hair is on average two thousand times thicker. To ensure high read and write rates, data from many memory modules is requested at the same time. Thanks to this, the data transfer rates of the individual chips are summed up.

A lot has been written about SSD drives as a new generation of hard drives. And now, due to the floods in Thailand, I think the SSD position will be pumped all the way.

Since I have experience in repairing computers and components, I will consider the operation of this device from a practical point of view, that is, taking into account all the conveniences of using an SSD, plus problems and their solutions when the device fails.

SSD is an abbreviation of the English Solid State Drive, which means solid state drive. It has no mechanical parts, it cannot classify it as a drive or hard drive. It is customary to say that this device has three main advantages over a conventional hard drive.

The first advantage is speed. SSD is three times faster on average at boot operating system, when accessing programs such as Photoshop and when working in the programs themselves.

Second, it is completely silent.

And finally, the third: it is less power-intensive compared to a conventional hard drive.

Let's take a closer look at these benefits. According to the first, I can say that the speed is mainly felt when loading the operating system. Indeed, on an SSD, the system boots about three times faster.

When accessing programs, it is also fast, but not so much, somewhere twice, and this is felt when loading heavy programs such as Photoshop, AutoCAD and others.

When loading other programs, the force of habit probably plays a role: we are so used to distracting ourselves with something while the program is loading that the difference is almost not felt.

And the turn does not reach the speed of work in the program itself, because the SSD is subject to rapid wear, and no one wants to use the drive in programs once again.

Moreover, the wear of a conventional hard drive is not so terrible compared to the wear of an SSD. In case of wear or failure of the HDD, there are many utilities that allow you to programmatically restore a damaged disk or its individual sectors.

There are many ways, starting from the usual defragmentation - an option built into the operating system itself Windows system, to the extreme case of mechanical damage, when the only remaining option is to mechanically transfer the disks to another casing.

Thus, in 90% and even more cases, corrupted and even lost information from the HDD can be restored, which is almost impossible on an SSD.

Only the operating system and the Program Files folder are suitable for using an SSD. All other information, file and data base, as well as intensive work with programs, let it better remain on a regular mechanical hard disk HDD.

The advantage in terms of energy consumption is an important thing - this, of course, is the lower power consumption of an SSD, but given that the possibility of irretrievable loss of information is very high in the event of a power failure, this advantage also becomes, to put it mildly, very controversial.

And, finally, the financial side, the price of the issue, so to speak: an SSD is expensive, a normal 120 GB drive costs about $ 240 in Moscow. There are no such prices in the regions. In addition, if the price of hard drives is inversely proportional to upgrades, upgrades and capacity increases, then in the case of SSDs, exactly the opposite is true.

For example, in an SSD, there are two types of controllers. This is a programmable power supply and distribution chip for work and information in an SSD. The Sand-Force and JMicron controller program did a very poor job of handling these features. They recorded information very unevenly (for HDD, this issue is solved by regular defragmentation).

When one cell of a drive fails, the entire drive fails. By the way, a damaged HDD cell is the simplest defect that has a bunch of solutions from software "bypass" of the cell (transfer to quarantine) to software magnetization of the disk.

So, to solve this problem, the Trim command was invented for SSD, which should ensure uniform wear of the drive. Oddly enough, along with this innovation, the SSD has risen in price, when, according to all the canons of business and logic, it should have been the other way around.

Due to flooding in Thailand, 80% of the production of hard drives has been suspended. Until spring, it is unlikely that at least minimal work on restoring production will be able to begin. Stores that sell computers stop selling HDDs separately from computers. Not to mention that HDD prices have doubled.

So what is an SSD

Translated from English, solid-state drive means "a disk without moving parts." A solid state drive is a storage device, the principle of which is based on the use of rewritable chips and a controller. Often users confuse the terminology and call SSD a hard drive. This is wrong because technical features hard drives. Distinctive feature This type of media from the HDD is that when reading data from an SSD, there is no need to perform mechanical operations, all the time it takes only to transfer the address and the block itself. Accordingly, the faster the memory of the device and the controller itself, the faster the general access to the data.

However, the process of changing or erasing data from SSD drives is not so simple. This is due to the fact that the memory is written in blocks of 4 KB, and erased in 512 KB.

When modifying blocks, the following sequence of actions occurs:

1. The block that contains the changes is read into the internal buffer.

2. The necessary modification of the bytes is made.

3. The block is erased from the flash memory.

4. The new location of the given block is calculated.

5. The block is written to a new location.

During the deletion of files, they are not physically deleted, but only marked by the system as deleted, however, the SSD does not know which data is user data and which is deleted, and in fact all blocks have to be processed according to the above scheme. This system leads to the fact that with a large amount of data on the disk, the total operating time increases significantly, which slows down all work.

SSD security and reliability

If we talk about the possibility of recovering data from an SSD, we can note the following points:

The data is not deleted immediately, as in the HDD, even if the file is overwritten with other data on top.

The process of data recovery is quite laborious, due to the fact that it is necessary to choose the right order, combine the results, and also select the necessary algorithm that emulates the operation of the media controller.

The reliability of an SSD directly depends on the reliability of the controller and its firmware, since it is the controller that is located between the interface and the memory chips and the likelihood of damage to it in the event of a power failure is very high.

Rules for working with solid-state media to extend their life cycle and increase overall speed:

All data that changes frequently (various temporary data, swap files, etc.) should be transferred to a regular HDD.

Disable disk defragmentation.

Periodically update the controller firmware.

Leaving about 20% of the disk partition permanently free will improve overall performance.

Advantages of SSD over hard drives:

Very high read speed of data blocks, which is actually limited only by throughput controller interface.

Low power consumption.

Noiselessness.

The absence of mechanical parts, which leads to fewer possible breakdowns.

Small overall dimensions.

High temperature resistance.

Disadvantages of SSD:

Limited number of memory cells rewriting cycles (from 10,000 to 100,000 times). Upon reaching the limit, your drive will simply stop working.

High price. Compared to the price of an HDD for 1 GB (about 1.6 rubles / GB for a 1 TB HDD versus 48 rubles / GB for a 128 GB SSD).

Low disk space compared to HDD.

Compatibility problem with some versions of operating systems (some operating systems simply do not take into account the specifics of solid-state media, which leads to very rapid wear of the media).

SSD Companies and Manufacturers You Can Trust:

Intel, Kingston, OCZ, Corsar, Crucial, Transcend, ADATA.

Hard disk device

The very design of the hard disk consists not only of direct storage of information, but also of a mechanism that reads all this data. Here is the main difference between hard drives and floppy disks and optical drives. Moreover, unlike random access memory(RAM), which needs constant power, the hard drive is a non-volatile device. The data on it is saved regardless of whether the computer is powered on or not - this is especially important when you need to restore information.

A little about the hard disk device. The hard drive consists of a sealed block of disks filled with ordinary dust-free air under atmospheric pressure, and a board with electronic circuit management. The block contains the mechanical parts of the drive. One or more magnetic disks are rigidly fixed on the spindle of the disk rotation drive motor.

Also there is a preamplifier-switch of magnetic heads. The magnetic head itself reads or writes information from the surface of one of the sides of the magnetic disk, the speed of which reaches 15 thousand revolutions per minute.

Internal HDD

When the power is turned on, the hard disk processor tests the electronics, after which the spindle motor is turned on. When a certain critical speed of rotation is reached, the density of the air layer that flows between the surface of the disk and the head becomes sufficient to overcome the force of pressing the head against the surface.

As a result, the read/write head "hangs" above the plate at a distance of 5-10 nm. The operation of the read / write head is similar to the principle of the needle in a gramophone, with only one difference - our head does not make physical contact with the plate.

When the computer's power is turned off and the disks stop, the head descends to the non-working area of the platter's surface, the so-called parking zone. Early hard drives had a special software, which initiated the head parking operation.

In modern HDDs, the head is brought to the parking zone automatically when the rotation speed drops below the nominal value or when the power is turned off. The heads are brought back into the working area only when the rated speed of the engine is reached.

Naturally, the question may arise - how tight is the disk block itself and what is the likelihood that dust or other small particles will leak there? After all, they can lead to a failure in the work of the hard drive or even to its breakdown and loss of important information.

The disc block with the motor and the heads are located in a special hermetically sealed housing - a HDA (chamber). However, its contents are not completely isolated from the environment, it is necessary to move air from the chamber to the outside and vice versa.

This is necessary to equalize the pressure inside the block with the external one in order to prevent deformation of the body. This balance is achieved by a device called a barometric filter. It is located inside the HDA.

The filter is capable of capturing particles larger than the distance between the read/write head and the ferromagnetic surface of the disc. In addition to the above mentioned filter, there is another one - the recirculation filter. It captures particles that are present in the airflow within the block itself. They can appear there from the shedding of magnetic pollination of disks. In addition, this filter captures those particles that its barometric "colleague" "missed".

HDD connection interfaces

Today, to connect a hard drive to a computer, you can use one of three interfaces: IDE, SCSI, and SATA.

Initially, in 1986, the IDE interface was developed only for connecting an HDD. Then it was modified into an extended ATA interface, to which you can connect not only hard drives, but also CD / DVD drives.

The SATA interface is faster and more productive than ATA.

In turn, SCSI is a high-performance interface that can connect various kinds of devices. This includes not only storage media, but also various peripherals. For example, faster SCSI scanners. However, when the USB bus appeared, the need to connect peripherals via SCSI disappeared.

SCSI interface

Now a little about connecting to the IDE interface. The system can have two controllers (primary and secondary), each of which can connect two devices. Accordingly, a maximum of 4 devices: primary master, primary slave and secondary master, secondary slave.

After connecting the device to the controller, you should select the mode of its operation. It is selected by setting a jumper (jumper) to a certain place in the connector on the device (next to the connector for connecting the IDE cable).

It should be remembered that the faster device is connected to the controller first and is called master. The second is called slave (slave). The last manipulation will be to connect the power, for this we need to select one of the power supply cables.

DE interface

Connecting a SATA drive is much easier. The cable for it has the same connectors at both ends. The SATA drive has no jumpers, so you won't need to select the device operation mode. Power is connected to the SATA drive using a special cable (3.3 V). However, it is possible to connect through an adapter to a conventional power cable.

SATA interface

Let's give one helpful advice: if friends often come to you with their hard drives, and you are already tired of spinning all the time system unit, we recommend that you purchase a dedicated hard drive pocket (called the Mobile Rack). They are available with both IDE and SATA interfaces. To connect another hard drive to your computer, simply slip it into your pocket and you're done.

SSD drives - a new stage in development

Now begins the next stage in the development of information storage devices. Hard disk drives are being replaced by a new type of device - SSD. Next, we will tell about it in more detail.

So, SSD (Solid State Disk) is a solid state drive that works on the principle of USB flash memory. One of its main distinguishing features from hard drives and optical drives is that its device does not include any moving parts and mechanical components.

Drives of this type were originally developed for military purposes, as well as for high-speed servers, since the good old hard drives for such needs were no longer fast and reliable enough.

We list the most important advantages of an SSD over a hard drive:

Firstly, writing information to an SSD and reading from it is much faster (tens of times) than from an HDD. The work of the hard drive is slowed down by the movement of the read / write head.

Secondly, due to the simultaneous use of all the memory modules installed in the SSD drive, the data transfer rate is much higher than that of the hard drive.

Thirdly, they are not so susceptible to blows. While hard drives can lose part of the data upon impact or even fail.

Fourthly, they consume less energy, which makes them convenient to use in devices powered by batteries.

Fifthly, this type of drive practically does not produce any noise during operation, while during operation of hard drives we hear the rotation of the disks and the movement of the head.

Perhaps there are two lack of SSD- 1) you will pay much more for its certain capacity than for a hard drive of the same amount of memory; 2) SSD drives have a relatively small limited number of read/write cycles.

An ordinary solid state drive is a printed circuit board with a chipset installed on it. This set consists of a NAND controller chip and, in fact, NAND memory chips.

Square printed circuit board solid state drive is used in full. Most of it is occupied by NAND-memory chips.

As you can see, there are no mechanical parts and disks in the SSD drive - only microcircuits.

Types of memory in SSD.

Now that we have figured out the design of SSD drives, let's talk about them in more detail. As already mentioned, an ordinary SSD consists of two interconnected parts: memory and controller.

Let's start with memory.

To store information in an SSD, memory cells are used, which consist of a huge number of floating-gate MOSFETs. Cells are combined into pages of 4 kB (4096 bytes), then into blocks of 128 pages, and then into an array of 1024 blocks. One array has a capacity of 512 MB and is controlled by a separate controller. Such a multi-level model of the drive device imposes certain restrictions on its operation. So, for example, information can be erased only in blocks of 512 kB, and recording is possible only in blocks of 4 kB. All this leads to the fact that a special controller controls the recording and reading of information from memory chips.

Here it is worth noting that a lot depends on the type of controller: read and write speed, failure resistance, reliability. We will talk about which controllers are used in SSDs a little later.

There are 2 types of NAND memory used in SSD: SLC and MLC. SLC (Single-Level Cell) memory uses single-level transistors (they are also called cells). This means that one transistor can store 0 or 1. In one word, such a transistor can only store 1 bit of information. It won't be enough, right?

Then big-headed men "scratched their turnips" and figured out how to make a transistor-cell 4-level. Each level represents 2 bits of information. That is, one of four combinations of 0 and 1 can be written on one transistor, namely: 00, 01, 10, 11. That is, 4 combinations, against 2 for SLC. Twice as much as on SLC cells! And they called them multi-level cells - MLC (Multi-Level Cell). Thus, on the same number of transistors (cells), you can record 2 times more information than if SLC cells were used. This significantly reduces the cost of the final product - SSD.

But MLC cells have significant drawbacks. The life of such cells is less than that of SLC and averages 100,000 cycles. For SLC cells, this parameter is 1,000,000 cycles. It is also worth noting that the read and write times for MLC cells are longer, which reduces the performance of the solid state drive.

It also considers options for using three-level cells (Triple-Level Cell) in SSDs, which have 8 levels, and, therefore, each TLC cell can store 3 bits of information (000, 001, 011, 111, 110, 100, 101, 010).

Comparison table of flash memory types: SLC, MLC and TLC. Characteristics of NAND SLC MLC TLC

Bits per cell 1 2 3

Overwrite cycles 100,000 3000 1000

Reading time 25 µs. 50 µs. ˜75 µs.

Programming time 200 - 300 µs. 600 - 900 µs. ˜900 - 1350 µs.

Erase time 1.5 - 2 ms. 3 ms. ˜4.5 ms.

The table shows that the more levels used in a cell, the slower the memory based on it. TLC-memory clearly loses, both in speed and in "lifetime" - rewriting cycles.

By the way, USB flash drives have long used TLC memory, which, although it “wears out” faster, is also much cheaper. That is why the cost of USB-flash and memory cards is steadily declining.

Despite the fact that various companies produce SSDs under their own brand, many people buy NAND memory from a small number of its manufacturers.

Manufacturers of NAND memory:

Toshiba/SanDisk

Thus, we learned that SSD drives come with two different types Memory: SLC and MLC. SLC-based memory is faster and more durable, but expensive. Memory on MLC cells is noticeably cheaper, but has a lower resource and speed. In the general market, you can only find SSDs based on MLC type flash memory. Hard drives with SLC memory are almost non-existent.

SSD controllers.

At the time of this writing, the following controllers are most widely used:

Sandforce controllers.

One of the most common SandForce controllers is SF2281. This controller supports the SATA-3 interface and is found in SSD drives Silicon power, OCZ Vertex 3, OCZ Agility 3, Kingston, Kingmax, Intel (Intel 330, 520, 335 series).

Marvel controllers.

Marvell 88SS9174. Used in Crucial C300, M4/C400, and Plextor M5 SSDs. This controller has established itself as one of the most inexpensive, reliable and fast.

Marvell 88SS9187. This controller is used in the Plextor M5 Pro, M5M series, and the updated M5S series of solid state drives. New features include a DRAM controller with support for up to 1 Gb DDR3. Also implemented modern system ECC error correction and reduced power consumption.

LAMD controllers (Hynix).

LAMD (Link A Media Devices) is a division of Hynix. LAMD's LM87800 controllers are used in Corcair's Neutron and Neutron GTX series drives. The LM87800 controller itself is eight-channel and supports the SATA 6Gb/s interface.

Indilinx controllers.

Everest. Since Indilinx is a subsidiary of OCZ, it is not surprising that the Everest2 controller is at the heart of such SSDs as OCZ Vertex 4, OCZ Agility 4. The advantage of the Indilinx controller is high write performance. It is also worth noting a good balance - the read and write speeds are almost the same.

Barefoot 2. The basis of the controller is the ARM Cortex-M0 core. This SATA II controller supports eight MLC and SLC type memory access channels. LPDDR memory can be used as buffer memory, as well as DDR. Base Solid State Media Capacity this controller can reach 512 GB.

Barefoot 3. The latest chip, made according to the 65 nm process technology and independently developed by OCZ. The basis of the controller is an ARM core and an Aragon co-processor (32-bit, 400 MHz). Thanks to the support of special RISC commands for working with solid state drives, this controller is a leader in speed. The Barefoot 3 controller is 8-channel and supports SATA 6Gb/s interface. Based on this controller, OCZ produces a line of SSD drives under the OCZ Vector brand.

Samsung controllers.

Samsung uses the Samsung MDX controller in its SSDs. For Samsung 840 Pro and Samsung 840 drives, an eight-channel MDX controller based on a 3-core ARM Cortex-R4 chip (300 MHz) is used.

About installing Windows on an SSD.

Installing Windows XP on an SSD is not recommended, as this operating system is not designed to work with an SSD. In Windows 7 and 8, SSD support is fully present. True, for a more durable and "correct" operation of the SSD with this system, it is recommended to configure some parameters of this OS.

The PC processor is the main component of the computer, its "brain", so to speak. It performs all the logical and arithmetic operations that the program specifies. In addition, it manages all computer devices.

Computer processor device - what is a modern processor.

Today processors are made in the form of microprocessors. Visually, the microprocessor is a thin plate of crystalline silicon in the shape of a rectangle. The area of the plate is several square millimeters, it contains circuits that provide the functionality of the PC processor. As a rule, the plate is protected by a ceramic or plastic flat case, to which it is connected by means of gold wires with metal tips. This design allows you to connect the processor to system board computer.

What is a PC processor made of?

address bus and data bus;

arithmetic-logical unit;

registers;

cache (small fast memory 8-512 KB);

command counters;

math coprocessor.

What is a PC processor architecture?

Processor architecture is the ability of a processor to execute a set of machine codes. This is from a programmer's point of view. But the developers of computer components adhere to a different interpretation of the concept of "processor architecture". In their opinion, the processor architecture is a reflection of the basic principles of the internal organization of certain types of processors. Let's say the architecture Intel Pentium denoted P5, Pentium II and Pentium III - P6, and not so long ago the popular Pentium 4 - NetBurst. When Intel closed P5 to competing manufacturers, AMD developed its K7 architecture for Athlon and Athlon XP, and K8 for Athlon 64.

What is a processor core?

Even processors with the same architecture can differ significantly from each other. These differences are due to the variety of processor cores that have a certain set of characteristics. The most common difference is the different system bus frequencies, as well as the size of the second-level cache and the technological characteristics by which the processors are manufactured. Very often, changing the core in processors from the same family also requires replacing the processor socket. And this entails problems with the compatibility of motherboards. But manufacturers are constantly improving kernels and making permanent but not significant changes to the kernel. Such innovations are called core revisions and are usually denoted by alphanumeric combinations.

What is a system bus?

The system bus or processor bus (FSB - Front Side Bus) is a set of signal lines that are combined according to their purpose (addresses, data, etc.). Each line has a specific information transfer protocol and electrical characteristics. That is, the system bus is a link that connects the processor itself and all other PC devices (hard drive, video card, memory, and much more). Only the CPU is connected to the system bus itself, all other devices are connected through controllers that are located in the north bridge of the system logic set (chipset) motherboard. Although in some processors the memory controller is connected directly to the processor, which provides a more efficient memory interface to the CPU.

What is a processor cache?

Cache or fast memory is a mandatory component of all modern processors. The cache is a buffer between the processor and the controller of a rather slow system memory. The buffer stores blocks of data currently being processed, and the processor does not need to constantly access slow system memory. Naturally, this significantly increases the overall performance of the processor itself.

In processors used today, the cache is divided into several levels. The fastest is the first level L1, which performs work with the processor core. It is usually divided into two parts - the data cache and the instruction cache. L2 interacts with L1 - the second level cache. It is much larger and is not divided into an instruction cache and a data cache. Some processors have L3 - the third level, it is even larger than the second level, but an order of magnitude slower, since the bus between the second and third levels is narrower than between the first and second. However, the speed of the third level is still much higher than the speed of the system memory.

There are two types of cache - exclusive and non-exclusive.

An exclusive type of cache is one in which information at all levels is strictly delimited to the original.

A non-exclusive cache is a cache in which information is repeated at all levels of the cache. It is difficult to say which type of cache is better, and the first and second have their advantages and disadvantages. The exclusive cache type is used in AMD processors, not exclusive - Intel.

What is a processor socket?

The processor socket can be slotted or socketed. In any case, its purpose is to install CPU. The use of a connector makes it easier to replace the processor during upgrades and remove it for a PC repair. Connectors can be used to install a CPU card and the processor itself. Connectors are distinguished by their purpose for certain types of processors or CPU cards.

The advantage of SSD drives over traditional hard drives is obvious at first glance. These are high mechanical reliability, no moving parts, high read / write speed, low weight, lower power consumption. But is everything as good as it seems?

We disassemble the ssd.

First, let's look at what an SSD is. SSD is solid state drive. SSD, Solid State Drive or Solid State Disk), a non-volatile, writable storage device with no moving mechanical parts using flash memory. SSD fully emulates the work of a hard drive.

Let's see what the SSD has inside and compare with its close relative USB Flash.

ssd controller.

The main task of the controller is to provide read / write operations, and manage the data layout structure. Based on the block placement matrix, which cells have already been written to and which have not yet, the controller must optimize the write speed and ensure the longest life of the SSD drive. Due to the nature of the construction of NAND-memory, it is impossible to work with each cell separately. The cells are combined into pages of 4 KB each, and information can be written only when the page is completely occupied. You can erase data in blocks that are equal to 512 KB. All these restrictions impose certain responsibilities on the correct intelligent algorithm of the controller. Therefore, properly configured and optimized controller algorithms can significantly improve the performance and durability of an SSD.

The controller includes the following main elements: Processor– usually 16 or 32 bit microcontroller. Executes firmware instructions, is responsible for mixing and aligning data on Flash, SMART diagnostics, caching, security. Error Correction (ECC)- ECC error control and correction unit. Flash Controller– includes addressing, data bus and control of Flash memory chips. DRAM Controller- addressing, data bus and DDR/DDR2/SDRAM cache memory management. I/O interface- responsible for the data transfer interface to external SATA, USB or SAS interfaces. Controller Memory- consists of ROM memory and buffer. Memory is used by the processor to execute firmware and as a buffer for temporary data storage. In the absence of an external RAM chip, the memory acts as the sole data buffer of the SSD.

Currently, the following controller models are used in the SSD: Indilinx "Barefoot ECO" IDX110MO1 Indilinx "Barefoot" IDX110M00 Intel PC29AS21BA0 JMicron JMF602 JMicron JMF612 Marvel 88SS9174-BJP2 Samsung S3C29RBB01-YK40 SandForce SF-1200 SandForce SF-15001XB T6

Flash memory.

In SSD, as in USB Flash, three types of NAND memory are used: SLC (Single Level Cell), MLC (Multi Level Cell) and TLC (Three Level Cell). The only difference is that SLC allows you to store only one bit of information in each cell, MLC - two, and TLC - three cells (using different levels of electric charge on the floating gate of the transistor), which makes MLC and TLC memory cheaper in terms of capacity.

However, MLC/TLC memory has a shorter resource (100,000 erase cycles for SLC, an average of 10,000 for MLC, and up to 5,000 for TLC) and worse performance. With each additional level, the task of recognizing the signal level becomes more complicated, the search time for the cell address increases, and the probability of errors increases. Since SLC chips are much more expensive and their volume is lower, MLC / TLC chips are mainly used for mass solutions. At the moment, MLC/TLC memory is actively developing and is approaching SLC in terms of speed characteristics. Also, low MLC/TLC speed is compensated by SSD drive manufacturers with algorithms for interleaving data blocks between memory chips (simultaneous writing/reading to two flash memory chips, one byte each) by analogy with RAID 0, and low resource - by mixing and tracking uniform use of cells. Plus, a part of the amount of memory is reserved in the SSD (up to 20%). This is unavailable memory for standard write/read operations. It is needed as a reserve in case of cell wear, by analogy with HDD magnetic drives, which have a reserve for replacing bad blocks. The extra cell reserve is used dynamically, and as the primary cells physically wear out, a replacement cell is provided.

I'll show you how to change the HDD hard drive to a high-speed SSD drive. I bought a Samsung 850 Evo 250 GB SSD drive. and installed it on my laptop. Then I installed Windows and all programs on a new SSD drive.

I bought my Samsung 850 SSD EVO 120 GB SATA III SSD on AliExpress . At first I wanted to order such a Samsung 750 SSD EVO 120 GB SATA III (it is 120 GB and cheaper), but in the end I ordered 250 GB, although I could have done with 120 GB. The Samsung 850 EVO SSD arrived after about 12 days (the fastest product that came from AliExpress).

The parcel is well packed and sealed with styrofoam. Inside the box is plastic, and in it is an SSD drive.

Here are the specifications for this SSD drive. My reading speed tests, write at the bottom of the page.

1. Copy all the information you need from your disk

If you, like me, have only one place for a hard drive in a laptop, then first copy all the information from your hard drive to yourself external drive or on another computer. Or buy. So that you can then connect your removed HDD via USB and download everything you need from it to your new SSD drive.

Here is a visual video of this adapter.

2. Remove the hard drive and install the SSD

Turn off the laptop, disconnect the laptop from all wires, turn it over and remove the laptop battery. Now on back cover laptop, find the inscription HDD - this is the place where your hard drive is installed. On my Samsung NP-R560 laptop, it is on the bottom left. The hard drive is covered with a cover with two screws.

We unscrew these two screws covering the laptop hard drive.

Remove the cover covering the hard drive. There should be arrows on it showing in which direction you need to pull to move the cover.

Here is my laptop hard drive. It has an aluminum cover to better dissipate heat and has a pull tab to make it easier to take it off. Simply grasp this tab and pull to the left to disconnect the hard drive from the connector.

Done, the hard drive is disconnected from the laptop and connectors. Pick it up and set it aside.

This is what a laptop looks like without a disk.

Now put the SSD in place. HDD drive.

Carefully insert it in place of the old HDD. I also put the aluminum plate from the old HDD onto the new SSD.

Close the hard drive cover.

Tighten the cover screws.

Ready. Now we turn the laptop over, insert all the wires into it, put the battery back and turn on the laptop.

3. Install Windows on a new SSD

There is nothing on the new SSD drive and there is no OS (Windows) either, so now you need to install Windows on it. You will get this error when trying to boot from a new SSD drive that does not yet have a Windows operating system.

Partition table invalid or corrupted. Press any key to continue…

You need to insert your bootable flash drive and boot from it.

If you don't have a bootable flash drive yet, it's time to make one.

Here is a video on how to set up the BIOS for Windows installation from a bootable flash drive.

Now when there is a bootable flash drive and the boot occurs from it, we install Windows on the new SSD. We select our SSD, it will be marked as "Unallocated space on disk 0" and click "Next" and install Windows.

Copy will start Windows files, then prepare for installation, install components, install updates, finish. The computer will restart several times. After the first reboot, you can remove the bootable USB flash drive.

If you have never installed Windows through the BIOS, then you will see a video on this topic.

After installing Windows on a new SSD drive, change the boot priority in the BIOS to first Windows bootloader searched on the SSD drive. Although if everything is loaded and working, then you can not change anything already. I will go into the BIOS, Boot - Boot Device priority.

And with the F5 or F6 key, I will move the SSD drive to the very top, so that the boot sector on the SSD drive is first searched, and then on the other drives, if it was not found on the SSD.

4. Comparing the speed of SSD with HDD and USB disks

Using CrystalDiskMark 3, I measured the write and read speed of my HDD before removing it and replacing it with an SSD. The read speed from it was about 100 MB / sec. for sequential reading and writing.

Popular in category: