XMOS Microcontroller MCU

An XS1 combines a number of XCore™ processors, each with its own memory, on a single chip. The programmable processors are general purpose in the sense that they can execute languages such as C; they also have direct support for concurrent processing (multi-threading), communication and input-output. A high-performance switch supports communication between the processors, and inter-chip XMOS Links are provided so that systems can easily be constructed from multiple chips. The XS1 products are intended to make it practical to use software to perform many functions which would normally be done by hardware; an important example is interfacing and input-output controllers.

Each XCore processor provides the following resources:
32-bit processor providing up to 500MIPS
Eight hardware threads and 32 channel ends
Ten timers and six clock blocks
Four XMOS Links
64KBytes SRAM and 8KBytes OTP memory

The XCore is a multithreaded processing component with instruction set support for communication, input-output and timing. Thread execution is deterministic and the time taken to execute a sequence of instructions can be accurately predicted. This makes it possible for software executing on an XCore to perform many functions normally performed by hardware, especially DSP and I/O.,Each XCore thread has a dedicated set of registers and scheduling of threads is performed by hardware. Communication between threads is performed using hardware channels. Communication instructions transfer data directly between processor registers and channels, automatically scheduling and descheduling the communicating threads to control the flow of data.,The XCore has an efficient set of instructions to support conventional sequential programming languages. Its multithreading, communication and input-output instructions are designed to support modern concurrent programming languages. The instruction set is easily extensible to provide applications-specific instructions; these currently include support for long-arithmetic, CRC, DSP and cryptography.,Each XCore has a single unified memory system which is shared for program and data by all of the threads executed by the core. Multiple threads can share the same program in memory and can pass ownership of data between them. No caches are used.,The XCore has a tightly integrated set of I/O ports controlled directly by instructions. Data is transferred directly between processor registers and ports, avoiding the use of memory and minimizing latency. The ports can serialize and deserialize data enabling the processor to handle high speed data streams. They can also timestamp data arrival, and accurately control the time at which data is transferred to or from the pins.