preface

I was approached by a friend who wanted to learn about Android devices and asked me a series of questions:

“I heard you’re an Android developer. What’s the difference between an LCD and an OLED screen?”

“What is LPDDR5? Can you explain?”

“What is UFS 3.1, and does it help?”

I:

“That’s a good question. Don’t ask it again.”

Boy, a string of questions. It’s silly to ask me.

I’m sure you’ll see the same thing when you’re learning about Android devices. Let’s talk about this knowledge

A, screen,

There are a lot of words to know, like LCD and OLED, high brush, and LTPO technology, which caught fire last year.

But before we get to the technical terms on the screen, let’s talk about the basics.

Resolution 1.

We all know that what is displayed on a screen is made up of pixels that are too small to be discernable by the naked eye.

Something like this:

For example, the resolution of my phone is 2376 * 1080, which means that the phone can store 2376 pixels on the side and 1080 pixels on the short side.

We buy mobile phones often talk about 2K screen, generally divided like this:

  • 720P: 1280 x 720
  • 1K screen: 1920 x 1080
  • 2K screen: 2560 x 1440

PPI is a word often used by manufacturers to promote mobile screens. What is PPI?

PPI stands for pixels per inch and is calculated as follows:

Again, take my phone, which has a resolution of 2376 x 1080 and a screen size of 6.57 feet.

The formula is the square root of (2376^2 + 1080^2), divided by 6.57, which is 397.

Many Android developers have a headache when looking at the resolution, because Android fragmentation is a serious problem, there are always too many resolutions to adapt.

2. LCD and OLED

OLED screens have been used in high-end models for years. Does that mean OLED screens are ready?

2.1 Principles

To understand the difference between the two screens, the first optical primary colors, three colors are red, green and blue, through the configuration of different columns, can form different colors.

Each pixel on the screen is made up of three sub-pixels, like this:

Although both use this method, the principle of LCD and OLED is quite different. To put it simply, LCD is the backlight, while OLED is the spontaneous light.

Pull out their blueprints:

The back of the LCD emits a pure white light that passes through a thin film of red, green and blue colors to get the three primary colors. If we don’t do any processing, red, green, and blue, we get white, so we have to add a switch in the middle layer, which is called the liquid crystal part, which can change the degree of opening and closing, change the intensity of light emission, so that the three colors make up different light.

OLED is basically a small red, green and blue light bulb that controls light intensity.

So, for LCD, even if the screen is all black, the back also needs to shine, which is often said light leakage.

Let’s talk about the differences in how they arrange their pixels:

LCD displays are arranged in RGB (left), while OLED displays are arranged in Pentile (right). As a result, OLED displays have low pixels for the same size.

No 2K (2560X1440), no A-screen (AMOLED screen)

2.2 Advantages and disadvantages

OLED vs. LCD:

  • Pros: Thinner, can be used as a flexible screen, brighter colors, can be lit alone, save power, responsive screen
  • Disadvantages: easy to burn screen, stroboscopic, pixel density

For a more detailed look at the differences, see:

What’s the difference between OLED and LCD?

3. Screen refresh frequency

It seems that from the year before last, major manufacturers have launched 120 HZ high-brush mobile phones.

120 HZ means that the screen can draw 120 frames per second. If the screen is described by seeing, it looks like this:

Remember colleagues just use the high brush mobile phone, said such a sentence: after using the high brush mobile phone, can not go back!

But higher refresh frame rates mean higher power consumption.

Starting in 2021, LTPO technology has been introduced by major manufacturers.

LTPO stands for Low Temperature Polycrystalline Oxide, which stands for Low Temperature Polycrystalline Oxide.

To put it simply, it dynamically changes the screen’s refresh frame rate.

For example, we in the case of reading a novel, don’t need 120 HZ high brush, so mobile phone frame rate will be set to 30 Hertz, and in the case of playing games, brush will be we need high, mobile phones and frame rate will be change to 120 HZ, of course, also get the game support high brush, such a gain a reduction, can reduce the energy consumption and improve life.

More details can be found at:

LTPO Technology Popularization: The Next Industry Benchmark for Mobile Screens

Second, running memory RAM

RAM refers to the operating memory. In recent years, Android phones have started with 8GB of RAM and even reached 16GB.

LPDDR is the most commonly heard form of RAM, and LPDDR5 has been advertised with a lot of effort:

Take a look at the LPDDR family tree:

DRAM is a way of storing data in memory, which is converted into binary by charging the capacitor and calculating the capacitance difference. However, the capacitor is leaky, so it needs to be charged periodically.

CPU frequency is generally measured in GHZ, for example, the latest Snapdragon 8 Gen1 is the largest core frequency at 3.0 GHZ.

The operating frequency of the running memory is still in MHZ, so the CPU and the running memory have reached an agreement that the CPU must be able to get the data at the specified period without waiting, so the S in SDRAM stands for synchronous.

DDR SDRAM = DDR SDRAM = DDR SDRAM

In the past, data access was done once in a clock cycle. Now, data access can be done twice in the rising edge and the falling edge.

Finally, LPDDR is introduced in this paper. Its LP means Low Power and Low Power consumption, so its full name is Low Power Double Data Rate SDRAM, which is the standard for Low Power memory developed by JEDEC Solid State Technology Association.

LPDDR has low power consumption and small size, which makes it naturally suitable for mobile devices. Finally, let’s take a look at LPDDR parameters:

As of November 2021, some companies have developed the latest LPDDR5X DRAM, which is 30% faster and 20% less power.

Three, storage memory ROM

ROM is called read-only memory, and unlike RAM, data does not disappear because the power is turned off. And it’s the same number that comes with 32GB \ 64GB \ 128GB, which determines how many videos and games we can store on our phones.

Remember when the network ridicule, Lu Xun said, buy mobile phones do not buy 16G storage space mobile phone!

1. Classification of ROM

Although ROM is read-only memory, can only read out can not write information, just like the Android device in the storage system of that area we can never modify delete the same truth, but in order to facilitate user production and use, again appeared:

  • Programmable read-only device (PROM)
  • Erasable programmable read-only device (EPROM)
  • Electric erasable programmable read-only memory (EEPROM) and other different types

The ROM used in mobile devices is flash memory, which is defined as:

Flash memory is a form of electronically erasable programmed read-only memory that allows multiple erasure or writing during operation.

Flash memory is a special type of EEPROM written with macro block erasers. It is widely used by mobile devices because of its low cost, data loss and high reliability.

2. It and eMMC

What is UFS 3.1?

Common classifications of flash memory are SSD, eMMC, and UFS.

SSD needless to say, is used in PC, remember when I was in college, I replaced the mechanical hard disk in the laptop with SSD, the computer was like a freshman, playing masturbated ah can win more.

EMMC and UFS are targeted at mobile devices, which are characterized by small size, low power consumption and moderate performance. In recent years, many mobile phones have launched 1TB large-capacity mobile phones, with the intention of catching up with PC.

In recent years, mobile phones have adopted UFS instead of eMMC, because UFS adopts interface serialization and improves transmission frequency. Let’s see the parameters of UFS:

EMMC’s bandwidth is only 400 MB/S, so abandoning eMMC is inevitable.

More information is available at:

How does UFS 3.1 make apps start and load faster? Discussion on SSD, eMMC, UFS

3. File system

When we learn IO call link in Android:

Often referred to as a file system, VFS provides a unified file interface to applications, leaving the implementation to the file system.

The main file systems are ext4 and F2FS, for example, on my Android phone:

My phone uses F2FS for internal storage and EXT4 for system storage.

Random I/O scenarios are more frequent for apps, so internal storage at the application level is more appropriate. However, under the same conditions, F2FS occupies a higher storage space, about 1.1 to 1.5 of ext4, and the performance of continuous reading and writing large files is not as strong as ext4, so ext4 is selected as the system storage.

Iv. Processor

The processor is the brain of the phone. The most common Android processors we buy are Qualcomm’s Snapdragon, Mediatek’s Breguet, Huawei’s Kirin, and Samsung’s Exynos.

Normally, on a PC, the processor is the CPU, like the Core I7-12700.

Processors in mobile devices are different. Considering the size and power consumption, they use highly integrated SoC (System on Chip), where a processor concentrates the critical parts of the System on a single Chip.

1. What is integrated into SoC

The sparrow is small, but it has all the five organs.

  • Central processing Unit (CPU)
  • Graphics Processing Unit (GPU)
  • Image Processing Unit (ISP)
  • Digital Model Processor (DSP)
  • Neural processing Unit (NPU)
  • Adjusting Modem
  • And other functional modules

Take a look at the latest SoC layout for a chip:

Many of the important components have been marked out.

Through SoC integration, the mainboard layout can be saved to the maximum extent, the device can be slim, the AC speed of IP unit can be improved, and the overall power consumption can be significantly reduced.

2. Process technology

SoC in mobile devices is cut from a single wafer. If its manufacturing process is more advanced, it means that it can unlock higher main frequency, lower heat, lower power consumption, not easy to reduce the frequency, and run stronger and more stable!

So, when manufacturers introduce chips, they will certainly introduce that they use XXNM process.

For example, the above chip uses Samsung’s 4nm process, while Mediatek’s Breguet 9000 uses TSMC’s 4nm process. Although the same 4nm process is used, TSMC’s 4nm process is stronger.

3. CPU

CPU is one of the most critical core units in SoC, which determines the performance of mobile phone.

Take the processor above, which is an octa-core processor:

Include:

  • 1 3.0GHz Cortex-X2 large core
  • Three new 2.5ghz Cortex-A710 midcore
  • Four 1.8ghz Cortex-A510 small cores

As it turns out, it’s very easy to get hot and slow down when the SoC is running at full speed, so this idea of small, medium and large cores fits the current trend.

In addition to CPU layout, we also need to understand a simple architecture.

When writing Android applications, we often need to be compatible with cpus of different architectures, such as the common Armeabi (ARM V5), ArmeabI-V7A (ARM V7), ARM64-V8A (ARM V8), and x86 architectures.

Arm starts with arm, which is the architecture of choice for mobile devices. The CPUS on the PC side are mostly x86 architecture, which has been around for a long time on mobile phones.

In a nutshell:

  • The ARM architecture processor is derived from RISC (Abbreviated Instruction Set), which optimizes commonly used commands to provide a more concise and efficient execution environment. RISC cedes complexity to the compiler, sacrificing program size and instruction bandwidth in exchange for a simple and low-power hardware implementation.
  • The x86 architecture’s counterpart, CISC (Complex instruction Set), offers higher performance at the cost of increasing the complexity of the processor itself.

Therefore, the low-power and high-efficiency ARM architecture is clearly more suitable for mobile devices. It is worth mentioning that the M1 chip for computers released by Apple last year is also based on ARM architecture.

4. GPU

GPU corresponds to the graphics card of mobile devices. The high resolution, refresh rate and the number of frames a mobile phone can run in playing games almost all depend on the GPU. It will provide many graphics rendering interfaces for games and applications, such as the common openGL.

Commonly used Gpus include Qualcomm’s Adreno series and Arm’s Mail series.

To learn more about processors, read:

Why does SoC performance depend on architecture and technology?

conclusion

Even application-level development engineers will inevitably encounter hardware knowledge when reading related documentation.

Therefore, learning simple hardware knowledge is somewhat necessary.

Because hardware knowledge is not much, this article will inevitably have mistakes, welcome to correct!

If you think this article is good, “sanlian” is the best support for the author!!