10 most terrible computer disasters in history

Just because of a computer crash, the entire US power system had trouble for hours, and Russia's missile detection system almost blamed the Third World.

1. False nuclear disaster

On November 9, 1979, all American fighter jets and warheads were placed on alert and ready to face the end of the world. At military command centers across the United States, the screens all announce an enormous amount of Soviet nuclear missiles pointing straight at them. The emergency plane 'Doomsday' for the US president is also about to take off.

But then they checked the data from the satellites of the nuclear missile defense system and realized that a rehearsal tape had been accidentally put into the system and alarmed the whole nation.

2. The genuine Windows has been turned into a mess

Because of a bug in the anti-piracy tool (Windows Genuine Advantage) that Microsoft accused thousands of customers of being a thief.

Famously annoyed with confusing confirmation steps, WGA mistook thousands of "genuine" Windows versions for clones. According to Microsoft, this incident occurred because a member of the WGA development team uploaded the wrong software onto the company's server and it automatically deleted the installation code on Windows, then the WGA group organized to eat and forget. Do not retest this software.

As a result, the next day any computer connected to the WGA server is considered to be using pirated Windows. Windows XP customers are warned to use unauthorized software, while Vista users are blocked from certain features.

3. Loss of network due to software error

AT & T's network manager can only look in vain when its 72-inch screen displays the red lines, marking the collapse of its phone network. In a "good day" business , the operator could make up to 70% of long distance calls in the US, about 115 million. However, January 15, 1990 was not a "good day" for AT&T.

The problem comes in New York, when the AT&T 114 signal circuits (each can make about 700,000 calls per hour) suddenly turn off for 4 seconds and restart because of overload. These 114 circuits are connected together under cascading and parallel network systems to find the best line for each call. And the circuits all inform each other of their operating status.

When these circuits return to active status, they send a signal that they are online and ready to receive calls. Then about 0.01 second a confirmation signal needs to be transferred, but because the software on the circuit has a problem, this signal is sent before the online notification signal. This incident made AT&T unable to handle 50% of their calls within 9 hours, and the report later assessed that they had lost up to 60 million USD.

4. Protection of car accidents is disastrous

Volvo car attaches great importance to protection in car accidents and is a pioneer in the field of accident mitigation. However, they happen in the most unexpected circumstances. Both times Volvo announced to the world new protection features, both failed miserably.

While the company is demonstrating a collision-avoidance system on the S60, its engine suddenly ignites and burns the car. The S60 was designed to anticipate imminent collisions, but the problem between the control system and the battery made Volvo shy.

Not discouraged, the company continues to test the S60 with a performance system that makes it possible to avoid pedestrians. This system uses sensors and cameras to determine the pedestrian's position and make a sudden braking if a collision is anticipated. And in this test, Volvo's car caught the brakes before 9 out of 12 dummies, the rest were thrown into the sky.

How did the experts answer this case? A study at the University of Washington hacked into the control system of many different cars and found that these systems could automatically activate control almost the entire function of the car and disobey the Driver's steering, including brakes.

5. Aircraft "half lean half fat"

In the spring of 2005, the Airbus A380 was preparing to fly for the first time in Toulouse, when engineers received an error message from the flight software. That error caused the airline to lose up to $ 6 billion because it had to delay the first passenger flight until two years later.

The problem is that the French airline uses CATIA 5 software to design for the previous part, while the German partner produces the tail part using the CATIA version 4. Easy to understand why they are not compatible with each other . The difference between the two software will result in the cables being malfunctioning, and needing to be reconnected.

Even if people write code to help them be compatible, it still needs a separate signal line to connect both planes. But engineers say the Airbus A380 does not have enough space to connect a signal wire far enough away from the signal to avoid signal interference. Perhaps many people think that just connecting the ends of the cable is finished, but keep in mind that the cable on the A380 has a total length of up to 530km, more than 100,000 different wires and 40,000 connectors.

6. Explode gas pipe

In the early years of the 1980s, the Soviet Union began looking for new technologies for industrial control systems, the most economical and simple method of 'chôm' of other countries. However, this mistake caused them to pay the price because the United States soon discovered the conspiracy to steal software from the Soviet rival and planned to prank them.

Knowing the KGB's goal is the gas pipeline control software of a Canadian company, US employees secretly install the revised version of the software into this company. The time bomb in the software passed the Soviet censorship and was included in the gas pipeline control system capable of supplying 40 billion cubic meters of gas a year to Europe.

This software has caused valves, pipes and turbines to automatically turn on and off at random and create tremendous pressure, causing these gas pipelines to explode. According to the report, no one was injured but economic losses were difficult to reach.

7. Pentium chip error

In the summer of 1994, Intel frantically advertised its new Pentium chip. Professor Thomas Nicely used this processor to run a program to calculate prime numbers. Most of these calculations are accurate, but most are not all.

Professor Nicely discovered unusual results in these calculations, and it took him five months to find the cause and it came from the processor he used. Unlike previous CPU versions, the 486DX and Pentium models contain a real number set (FPU) to support computing. The problem is that the FPU table on this processor is faulty and lacks a formula for dividing and in some cases will lead to errors. For example, if we split 4195835 for 3145727 we will get exactly 1.33374 but Pentium processor will give 1.33382, the error is about 0.006%. This error may not be a problem in normal computing tasks, but for science this is unacceptable.

Intel 'firefighters' with such false statements only appears at a rate of 1 out of 9 billion real numbers ' calculations ' or' it takes 27,000 years to use then a new user has encountered this error for the second time. only accepting chips for users who can prove that their work requires such precision. And so IBM computer company threatened by stopping all Pentium orders, Intel had to give in and accept to replace all the faulty chips.

5 million Pentium chips were shipped to the market, but luckily for Intel, most users didn't bother to replace the chip, otherwise the loss of the company didn't just stop at $ 500 million.

8. Missing spacecraft

The primary flaws in space tests can also be extremely expensive, namely $ 327 million. In 1999, the Mars Climate Orbiter probe approached Mars with the task of collecting weather information. But in the end the mission was completely failed.

The Mars Climate Orbiter is controlled by a faulty software and makes the controller's thrust error of 4.45. This error is equal to a pound force. Gradually this error changed the trajectory of the ship. The Mars Climate Orbiter should only have access to Mars's atmosphere at a height of 200 km on the surface, but it has actually been much lower. The gravitational pull of Mars pulled this ship down and friction with the atmosphere caught it.

A few weeks later the Mars Polar Lander also disappeared without trace but not the same cause as the Mars Climate Orbiter. Experts believe that the sensor on the ship mistook the ship's vibration when approaching the atmosphere with vibrations when the ship landed, and so the Mars Polar Lander's engine was allowed to shut down while it was still a few feet away from the ground. kilometer. As a result, everyone is sure.

9. The dark day of America

On August 14, 2003, the biggest electrical incident in US history came from the XA / 21 electricity management system. A small error occurred in the system's 4 million C lines causing a series of consecutive problems in just a millisecond. The data constantly overloaded the system, which had the alarm function disconnected and caused the main server to crash after 30 minutes. The data is transferred back to the standby server and continues to crash in the next few minutes.

Within an hour the line and fuse lines in turn overloaded the power outage across the Northeastern United States and part of Canada. Disconnected power lines lead to a decrease in the required power supply and cause the generators to stop working.

This incident occurred at rush hour and caused tremendous consequences. 256 generators stopped working, leading to information services such as mobile phone networks. According to the report, the best method of communication is to use a laptop with a dial-up modem using a landline phone line.

10. Imagine the sun is a rocket

During the Cold War when the great powers continued to race arms, an error in the missile detection system almost made up the Third World War.

The Soviet Union installed the Oko system to allow the detection of intercontinental missiles and send information about headquarters. This system detects missiles based on the satellite's ability to scan long distances on the surface of the Earth. However, on September 26, 1983, the sun, Soviet satellites and US missile launchers all sat in a straight line. A strong light of the sun from the US rocket launcher directly hits the Cosmos 1382 satellite sensor. And the satellite recognizes the light as the burning tail of the rocket.

Although the information recorded from the sensor software could not accurately identify this type of missile, the screen on the Soviet side still showed five nuclear warheads heading towards them. This was an extremely stressful time for Colonel Stanislav Petrov, the only thing that made him hesitate to press the alarm button just as a feeling. He argued that the United States never launched a war with only five missiles.