An instruction execution method for an artificial intelligence chip, the artificial intelligence chip comprising an instruction memory, a data memory, at least one general execution unit, and at least one dedicated execution unit, the instruction memory being communicatively connected with a central processing unit, the method comprising;
- receiving a code block from the instruction memory, the code block being acquired by the instruction memory from a kernel code sent by the central processing unit, the code block being associated with a dedicated execution unit;
- locking the dedicated execution unit associated with the code block;
- and
- sending an instruction in the code block to the locked dedicated execution unit, the dedicated execution unit being configured to execute the received instruction and store an execution result in the data memory.