Shulou(Shulou.com)06/02 Report--

Single event reversal (SEU,Single-Event Upsets) is the existence of a large number of high-energy charged particles in the space environment. The electronic components in the computer are irradiated by the earth's magnetic field, cosmic rays, etc., causing the potential state to jump, "0" becomes "1", or "1" becomes "0".

This problem has already been noticed in the early days of space exploration.

On the earth's surface, with the protection of magnetic field and atmosphere, there are few high-energy particles that can reach the ground, which generally have little impact on computers. In computer design, this particular scenario does not need to be considered. The conventional hardware error correction mechanism combined with the operating system is sufficient to cope with occasional calculation errors. I really can't change it. I can only get a blue screen and shout "long live Sister Li" and it's over.

Long live Sister Li

The situation is very different in space, and the probability of single event reversal is much higher.

Two monitoring devices for single event measurement carried on the practice 4 satellite, which was launched on February 8, 1995, flipped a total of 65 times in the 19 days after entering orbit.

Satellite designers don't want to answer "what kind of experience is the blue screen of a satellite master computer?" no one can reach the satellite's Reset button. The core program is gone, but there may not be a chance to save it.

Take a look at the holes that China Aerospace has stepped on (we don't know how many of them are not public). )

Fengyun-1B was successfully launched on September 3, 1990. After only two months of operation, we encountered a flow of high-energy particles emitted by solar flares in early November, resulting in a single event, resulting in confusion in the attitude control computer program, unable to control the attitude of the satellite, and causing the satellite to flip in space. Fortunately, after this incident, the computer program was corrected in time and the satellite returned to normal operation. On February 14, 1991, the satellite's computer once again appeared a single event event, and the attitude of the satellite was abnormal again, and the fault could not be detected in time. When it is found that the attitude of the satellite is abnormal, the gas carried on the satellite has been sprayed, the attitude is completely out of control, and the cloud image cannot be photographed. The original design life of the satellite is to run for one year, but the satellite cannot work in less than half a year.

The analysis and summary after the event is that there is a loophole without anti-radiation reinforcement in the selection of components and software design, and the computer circuit chip is easy to produce single event flip problem under the bombardment of high-energy particles in space, which may cause the computer to work abnormally at any time. Attitude control system failure.

For equipment with hundreds of millions of magnesium, as long as the performance reaches the target, reliability is the most important.

The method to improve reliability is also very simple and rough, that is, add redundancy and check. Typical technologies are three-mode redundancy (TMR), error correction code (ECC), etc., at the cost of performance degradation, of course.

In addition, advanced processes that improve computing performance may be more affected by single particles. And as the bit unit gets closer and closer in the smaller process nodes, there may be more than one single event flip, which poses a greater challenge to the verification algorithm.

Therefore, the performance of the computer chip on board may not be as good as the one in your phone.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.