P2P Distributed Parallel Computing

This year is the "dual-core year" of the processor ". For many years, Intel and AMD have been constantly striving to improve their CPU computing capabilities. After the frequency of a single chip reaches its limit, they finally pushed the CPU into the multi-core era.

In today's personal computers, the computing power is more than times that of large computers in the early days. This is thanks to the mutual promotion between the software industry and the hardware industry. The well-known example is the so-called "Microsoft-Intel" alliance, that is, operating systems and processors stimulate each other's market demands. Competition in the market leads to an oligarchy. In the PC market, two hard (Intel, amd) and one soft (Microsoft) products are quite fun.
 
Do we need such a fast computer? The eternal need for human beings to pursue the limits will allow us to continue on the way to pursue higher computing capabilities. However, "higher computing power" does not necessarily mean "faster CPU". The sum of all CPU computing power is always greater than a single CPU, distributed parallel computing, it provides the possibility for many CPUs to work collaboratively.
 
Multi-CPU (multiple computers) collaboration is basically a software-level issue, software-level issues, and ultimately an operating system platform issue. The internet gives Google a chance to put operations on the client for execution. This is a clever practice-it means less investment in central servers and better user experience. Google continues to launch new web-based services, making people feel that the Web OS era has arrived. This judgment is too early. It can be said that the computing capability of the Web Client platform is relatively low compared with that of a pure local program. Moreover, the browser itself still needs to run on the operating system. Third, this architecture still does not organically integrate customers' computers.
 
For many large companies such as IBM, the concept of grid computing has existed, and the actual results are not as great as that of a small BT/eMule. P2P technology is used in file sharing and has changed the entire entertainment industry landscape. If it is applied to computing power cooperation, the future will be limitless.
 
The Internet will become more popular. In the future, all personal devices will be online at any time, and devices (clients) will be intelligent agents of individual human beings. If you think of each device as a core of the CPU, it will be able to form a multi-core CPU with powerful computing power. Over the past few decades, the computer CPU and operating system have been moving towards a highly integrated direction, leading to a high degree of complexity in the software and hardware architecture, and the single-host computing capability is approaching the limit. P2P-based Distributed Parallel Computing solves the complexity brought about by High Integration of standalone servers.
 
P2P-Based WAN distributed parallel computing requires two solutions:
 
First, collaboration mechanism. How clients collaborate with each other, how to resolve problems, how to solve the latency caused by communication, and how to achieve "Hot swapping "......

Second, credit mechanism. WebService leasing and ASP (Application Service Provider) fail to become the mainstream market, largely because the credit mechanism is not sound enough. Credit has two levels: credit concept and credit technology. The former depends on the latter.

In addition to some scientific research projects, there are no widely-influential Wan distributed parallel computing applications. The Application of P2P technology remains at a lower level of file sharing. Operating system vendors are most likely to implement P2P distributed parallel computing. The English name of "matrix" is matrix, which describes a P2P parallel computing scenario. At the end of the film, a compromise between humans and computers may also represent the bright future of P2P parallel computing.

P2P distributed parallel computing is far from us. The day when it was widely used will also be the time for a major change in the hardware and software architecture. On that day, the CPU and our brain will be liberated.

(For programmers)