What's the difference between a data stream computer and a traditional computer?

A data stream computer refers to a computer that uses data streams to drive instruction execution.

The working mode of traditional computer is that the content of program calculator determines the operation order.

In data flow mode, only when all operands required by an instruction or a group of instructions are ready can the corresponding instruction be executed. The execution of instructions is data-driven and does not need a program calculator, which is beneficial to the development of concurrency.

Data flow computer: In the structure of data flow computer, the execution of instructions begins in a "data-driven" way. In this way, as long as all the required operands of any instruction in the program are available and there are available computing resources, it can immediately start execution (called ignition). The operation result of the instruction can be used as the operand of the next instruction to drive the ignition execution of the instruction. This is the meaning of "data-driven". In the data flow computer model, there is no * * * shared data. After an instruction is executed, it will not be sent to memory for use by other instructions. Instead, it flows directly to the instruction that needs the result as the new operand of the next instruction. After the instruction is used once, each operand disappears.

Traditional computers:

Refers to the traditional computer commonly used by modern people. Its heart depends on silicon chips, but the area of a chip is always limited.

Academician Tao Ruibao of Fudan University said that if it is not fundamentally changed, the development speed of traditional computers will be slower and slower, and 10- 15 years later will be completely stagnant. This assertion has been recognized by academicians.