Can hardware defects be supplemented by software? You will know after reading this.

Can the hardware digging pits be filled in by software? In fact, many hardware products have some unsatisfactory places. This may be a short-board of a technology, or it may be a problem caused by improper design. . In order to deal with these hardware deficiencies, the most commonly used filling method in the industry may be software! Many manufacturers will try to fix or alleviate some hardware problems through some specific software programs. Is this really effective? Today, let’s take a look at some famous software solutions for hardware filling!

SSD life is short? Balance algorithm to continue life

SSD is now widely used. If you don't use SSD now, you will even be ridiculed by some DIY players who don't understand the computer. However, it is not the norm for SSDs to achieve such high recognition. At least in the early years, many people are guilty about the longevity of SSDs.

SSD consists of flash memory, but to solve the life of flash memory, special software algorithms are needed.

SSDs are made up of flash memory, which has a lifetime limit. For example, MLC flash can only be erased thousands of times, TLC flash can only be erased hundreds of times, and so on. If you just use the flash memory to assemble an SSD, then the actual life performance may be very disappointing - when reading and writing data, it will focus on reading and writing some of the flash memory of the SSD, and this part of the flash memory will be particularly fast. Once this part of the flash is hung up, the entire SSD will hang.

The number of flash erases is limited, so the flash of the entire SSD is worn out.

This unbalanced wear situation may result in a 100G SSD, which is only due to the loss of a few M of flash memory, and the whole block is discarded. Let the flash memory life of the number M end, even if it is MLC, it may only need to erase dozens of G data. However, we all know that there is no SSD that will be so easily hanged, which is related to the special software algorithms of SSD.

To compensate for the shortcomings of SSD flash memory and maximize the life of SSDs, the industry introduced the Wear Leveling algorithm for SSDs to keep all flash wear as consistent as possible. The wear balance algorithm of SSD is roughly divided into dynamic and static. The dynamic algorithm is that when writing new data, it will automatically write to the newer block, and the old flash memory will be placed on the side; the static algorithm is more advanced, even if there is no data to be written, the SSD monitors Some flash blocks are older and will automatically allocate data, allowing older flash blocks to take over storage tasks that do not require data to be written, while allowing newer flash blocks to make room, and regular data reading and writing in newer blocks. In the middle - as a result, the life loss of each block is almost the same.

With this software algorithm blessing, even the SSD of TLC flash memory has a long life. For example, if the 256G TLC flash SSD has a lifetime of 500 erasures (P/E), then you need to write 125TB of data, and the flash memory will be dead – even if you write 10G data every day, it will take more than 30 years. Write flash memory, not to mention that few people write 10G data to SSD every day.

Some cottage SSDs use U disk solutions, while some U disk solutions do not provide wear leveling algorithms.

However, the wear leveling algorithm requires the master chip to be responsible for the calculations, and there are still some flash products that do not carry this algorithm, resulting in extremely short life - such as some low-end U disk. In the early years, the cottage U disk with unsatisfactory homes without the wear balance algorithm pretended to be SSD, resulting in a very short life of the so-called "SSD". SSD's unreliable bad word of mouth may have a lot to do with this.

LCD screen smear? Insert black algorithm to help

LCD screen LCD has become an absolute mainstream, but in the early years, LCD is a technology that is not very optimistic. Compared with CRT and plasma, the display effect of liquid crystal is obviously inferior. One of the more criticized ones is the smear.

Due to the hardware principle, the LCD needs to continuously deflect the liquid crystal molecules when displaying the dynamic picture. Liquid crystal molecular deflection is a continuous, steady state process that is not completed in an instant. One can observe that no matter how fast the liquid crystal molecules deflect, the LCD will have a more pronounced smear than the CRT and plasma. Smear is more noticeable when the screen displays high-speed moving objects, such as fast-moving trains, athletes in sports games, and the like.

Plugging in black is a typical solution to improve LCD smear

What to do? In order to solve the liquid crystal smear, the black insertion algorithm came into being. The so-called black insertion algorithm is actually inserting a black cymbal between a frame of the picture, so that the steady-state display of the LCD is converted into a pulse-like display like CRT and plasma, so that there is a time difference between each ,, which can Greatly reduce the appearance of afterimages. Of course, this also brings side effects such as flickering darkness, but the LCD improves the refresh rate and brightness by simply increasing the deflection speed of the liquid crystal molecules, so the black insertion algorithm is still practical.

OLED will burn screen? Offset algorithm to alleviate

As a display technology for the future, OLED has many advantages over liquid crystals, such as thin and light, power saving, high contrast, high color gamut, etc., but it also brings a problem that cannot be ignored - burning screen. Even the iPhone X, which is now in full swing, has brought the problem of burn-in after using the OLED screen, which is officially recognized by Apple.

iPhoneX uses an OLED screen, and the burn-in phenomenon is also mentioned in the description page.

Burning the screen is indeed one of the most troublesome problems of OLED, and it is closely related to the display principle of OLED. Unlike traditional LCD screens, OLEDs do not illuminate liquid crystal pixels through the backlight module, and each pixel of the OLED can self-illuminate. This brings many advantages, such as avoiding light leakage, stacking high contrast, etc., but there is a problem that different pixel points have different illumination times, and some pixels that often emit light/infrequently emit light will attenuate faster/ Slower, the brightness is significantly different from other pixels. We observed that these brightness decays faster/slower pixels, and the direct perception is that somewhere is dark/bright, which is “image sticking” or “burning screen”.

This is a typical burn-in screen, in addition to OLED, plasma screen is also very easy to appear

How to deal with burn-in? Either increase the lifetime of OLED pixels, so that users do not experience brightness decay during use - but this is difficult to achieve, and the cost is too high. Thus, the algorithm for preventing the burn-in offset was born.

One idea of ​​using software to solve OLED burn-in is to reduce the display of fixed images. Samsung is experienced in using OLED screens, and it has its own set of software algorithms to prevent OLED burn-in. In many Samsung mobile phones with OLED screens, the position of the image that is often fixedly displayed, such as a virtual button, is periodically displaced to avoid long-term illumination/non-luminescence of the same pixel, so that burn-in can be avoided to some extent.

Some Samsung mobile phones will move the virtual button position through the offset algorithm to prevent burn-in

A similar mechanism exists on the iPhone X. Someone had previously unpacked the iOS11 firmware and found that Apple also set up a program to prevent burn-in screens for OLEDs. In addition, there is no virtual button like Android on iOS, and iPhone X uses gestures, which undoubtedly greatly reduces the probability of burning.

However, the software algorithm to prevent OLED burn-in does not completely guarantee that OLED will not burn the screen, Apple does not dare to say so. Limited by hardware, OLED screen burn-in will still be a problem that is difficult to avoid completely. However, through software optimization and correct usage habits, the situation of burn-in can still be greatly reduced or not, and it is hoped that more manufacturers using OLED screens will join the anti-burn screen algorithm.

The CPU design has a bug? Patch BIOS to solve

In many people's minds, if used normally, the CPU should be the least likely part of the computer. However, if the CPU itself is not well designed, it is quite a headache. In fact, there have been cases where CPU design has bugs, but the CPU is still entering the consumer market. For example, AMD has done this.

AMD's phenom processor has TLB Bug

AMD's first-generation Phenom processor was given the responsibility of meeting the Intel Core processor. For the first time, the design of the third-level cache was used, which once gave A meal high hopes. However, unfortunately, Phenom, the first B2 stepping device, has a TLB bug. TLB is a bridge for connecting memory and CPU cache. In a Phenom processor with a bug, TLB will cause the CPU to read the page table and cause a crash. If you have a bug, you have to fix it. The CPU is difficult to return to the factory. What should I do? AMD then uses software to solve the problem.

AMD avoids this hardware bug through system patching and BIOS, but it will cause performance degradation.

AMD released a new BIOS and also provided a patch for the Win system, no matter which method, its role is to block a certain page table and even CPU cache. This of course can avoid the emergence of bugs, but it will also cause performance degradation. It can be said that this software repair solution is only a matter of expediency, but it is not perfect. Since then, AMD has introduced the B3 stepping Phenom processor, which has been completely modified by hardware modification. The CPU after the step change has been renamed from 9X00 to 9X50, which shows the importance of this repair.

to sum up

It can be seen that the software can make up for a lot of hardware defects, but it is not perfect for this time. For example, AMD's Phenom must solve the problem by modifying the hardware. The OLED offset display algorithm can't cure it. Screen problems. While new technologies will have unique advantages, some new hardware will bring new problems. I hope that manufacturers can truly experience the user experience and bring better products.

Endlaunch Adapter

Zysen offer 180degree Endlaunch Waveguide To Coaxial Adapter, from WR19 to WR430, N,SMA 2.92mm, 1.85mm connector are available. most are in stock!

End Launch Adapter,Coax Connector Adapter,Wr28 Waveguide Adapter,End Launch Coaxial Adapter

Chengdu Zysen Technology Co., Ltd. , https://www.zysenmw.com

This entry was posted in on