32-bit, high-performance CPU with SIMD DSP extensions
The eSi-3260 32-bit CPU is the top-of-the-range member in the eSi-RISC family of processor cores from EnSilica with SIMD (Single Instruction Multiple data) DSP extensions.
It is targeted specifically for applications needing DSP functionality with minimal silicon area and very low-power including smart sensors, IoT sensing nodes and always-on applications such as voice command recognition.
32-bit RISC IP: eSi-3260 Technical Overview
- 32-bit RISC architecture
- 16-bit SIMD (single instruction multiple data)
- single cycle MAC coupled with the option of 64-bit precision
- 32 general purpose registers
- 104 basic instructions and 10 addressing modes
- Optional IEEE 754 floating point unit (FPU)
- Supports up to 90 user-defined instructions
- 5-stage pipeline
- Optional memory management unit (MMU)
- Configurable instruction and data caches (1-64kB, direct mapped or 2 or 4 way associative)
- AMBA AXI or AHB interconnect and APB peripheral bus
- User and supervisor modes
- Up to 32 interrupts plus NMI and system call
- Fast interrupt response time of 6-9 cycles
- JTAG or serial debug
- Up to 3.69 CoreMark per MHz
- Intermixed 16 and 32-bit instructions result in exceptional code density without compromising performance
- High quality IP:
- Verilog RTL
- DFT ready
- Silicon proven
- C and C++ software development using license-free toolchain, under industry standard Eclipse IDE
- Easy migration path to cacheless version
- Always-on sensing nodes for audio
- Smart Sensors
- IoT sensing nodes
- Industrial control
The eSi-3260 32-bit CPU is the top-of-the-range member in the eSi-RISC family of processor cores from EnSilica. It is targeted specifically for applications needing DSP functionality with minimal silicon area and very low-power.
The processor features a single cycle MAC coupled with the option of 64-bit precision also makes the eSi-3260 ideal for audio, high-accuracy sensor hub; motion control and touch screen applications. Support for 16-bit SIMD (single instruction multiple data) operations with integrated multiplication and addition performed in a single cycle also targets filtering and complex arithmetic operations. The inclusion of saturation and rounding arithmetic provides excellent FFT acceleration with a 1024 point calculation being performed in less than 10µs when implemented in a 40nm LP process.
The processor features optional instruction and data caches that can be configured in size (from 1- 64kB) and associativity (direct mapped, 2 or 4-way associative) to increase performance when accessing off-chip memory. The optional paged memory management unit (MMU) enables the implementation of virtual memory and the ability to run operating systems such as Linux.
The 5-stage pipeline allows extremely high clock frequencies to be achieved. Static branch prediction is employed to minimize the cost of branch instructions.
The eSi-3260′s instruction set includes everything you would expect in a high-performance processor. There are also a number of optional application specific instructions and addressing modes. For example, a set of IEEE-754 compliant single-precision floating point instructions are available. For those applications that require extreme performance or ultra low power operation, user-defined instructions and registers can be implemented.
Instructions are encoded in either 16 or 32-bits, with all of the commonly used instructions being encoded in 16-bits, maximizing code density and improving cache performance.
Hardware debug facilities include hardware breakpoints, watchpoints, null pointer detection and single-stepping for fast debugging of ROM, FLASH and RAM based programs.
The toolchain for is based upon the industry standard GNU toolchain, which includes an optimising C and C++ compiler, assembler, linker, debugger, simulator and binary utilities. All these tools can be driven by the customizable Eclipse IDE. The debugger can connect to the target CPU either via JTAG, a serial interface or the Verilog PLI.
Complete C and C++ libraries are supplied. Ports of Micrium’s uC/OS-II RTOS, FreeRTOS and the lwIP TCP/IP stack are available. The toolchain is available for both Windows and Linux hosts and is available to use at no cost.
eSi-RISC is delivered as a Verilog RTL IP core. The design is target technology independent, although alternative implementations of some modules are available, such as the multiplier and JTAG interface, which are specifically optimised for FPGAs. The design is DFT ready, supporting full scan insertion for all flip flops and memory BIST. Example scripts are provided for popular EDA tools.
A selection of AMBA peripherals are supplied with the core, including: UART, SPI, I2C™, Timer, PWM, Watchdog, GPIO, PS/2, Ethernet MAC as well as a static memory interface and DMA engine. By using an industry standard bus, a wide range of 3rd party IP cores are also compatible with eSi-RISC.