Take the PC side of our manufacturer's computer as an example. We plug the USB flash drive with USB3.0 port into our USB3.0 computer port. How fast is their transmission speed?
The answer is: the transmission speed is 625MB/s in theory, but 500MB/s in fact.
Some people may ask, "500MB/s, this is a bit exaggerated. It is also a USB flash drive with USB3.0, and my file transfer speed has never exceeded 70MB/s ".
According to the theoretical guidance of the specification, the maximum transmission bandwidth of USB2.0 is 480Mbps, while that of USB3.0 is as high as 5.0Gbps.
USB3.0 uses differential signaling. What is a differential signal? Simply put, it is two transmission signals. Take USB2.0 as an example, the maximum transmission bandwidth of USB 2.0 is 480Mbps. To achieve the transmission rate of 480Mbps, the working frequency of the differential signal transmitted by USB 3.0 must be 240MHz, and the differential signal frequency of 2.5GHz is needed to achieve the transmission rate of 5Gbps. But in fact, high working frequency is easy to produce signal interference. USB3.0 is spread spectrum, which is no longer a simple 2.5Hz (some USB 3.0 cable specifications require 7.5GHz), so it is not accurate to calculate according to the fixed 2.5Hz.
In other words, let's consider it from the perspective of unit conversion. How many Mb/s is 1Gbps? The first conclusion is: 1 gigabit/sec =125m/sec. Here are the calculation steps.
# Make clear the concepts of Gbps and MB/s first. # Gbps means gigabit per second, and MB/s means megabyte per second. 1 byte = 8 bits 1 bit = (1/8) bytes 1 bit = 0. 125 bytes 1 megabyte = 1000000 bytes/kl.
So 5Gbps = 625mb/s. But remember the converted speed, all conditions are limited to the theoretical value under the USB3.0 specification. When PC and hardware manufacturers implement USB3.0 specification, it is often lower than the theoretical value of the technical specification due to transmission mode and firmware loss.
First of all, talk about the mode of communication. USB3.0 adopts 8b/ 10b coding in the transmission coding stage. What is 8b/ 10b coding?
8b/ 10b coding was put forward by Al Widmer and Peter Franaszek of IBM in 1983, applied to ESCON, and was later patented. At present, it is widely used by serial bus. The USB3.0 specification released in 2008 also explicitly adopted the 8b/ 10b coding.
8/ 10b encoding encodes 8-bit data into 10-bit data for transmission, which has two advantages.
In order to avoid transmitting a high-frequency clock signal alone when transmitting high-speed data, it is easy to cause EMI (electromagnetic interference) problems. 8/ 10b technology also encodes the clock signal, thus avoiding this problem. The number of "0" and "1" sent is basically the same, and the number of consecutive "1" or "0" does not exceed 5 digits, every 5 consecutive "60". However, when the link times out, DC offset will not occur. By 8b/ 10b coding, the transmitted data string can be correctly recovered at the receiving end. In addition, some special codes (K code in PCI-Express bus) can help the receiver to recover, and can find the transmission error of data bits at an early stage to prevent the error from continuing. Simply put, from the initial encoding stage to the final decoding stage, the actual data only accounts for 80%. Strictly speaking, from the actual transmission of useful data, the actual transmission speed should be 80% of 625MB/s, that is, 500mb/s. ..
The second is the loss of firmware. Losses such as hardware interface differences will also greatly reduce this transmission speed. For example, computer port USB2.0, USB disk USB3.0, so the file transmission speed will also be reduced because of the port; Whether the computer disk is solid or not and whether the mobile hard disk is solid or not will affect it.