AndesCore™ AX45MP Multicore

64-bit Multiprocessors with Level-2 Cache-Coherence

AndesCore™ AX45MP Overview

  • 64-bit in-order dual-issue 8-stage pipeline CPU architecture
  • Symmetric multiprocessing up to 8 cores
  • Level-2 cache and cache coherence support
  • AndeStar™ V5 Instruction Set Architecture (ISA)
    • Compliant to RISC-V RV64 GCP little endian
    • RV-GC: Integer, single/double precision floating point and 16-bit extensions
    • RV-P DSP/SIMD extensions (draft)
    • Andes V5 performance/code size extensions
  • Separately licensable Andes Custom Extension™ (ACE) for custom acceleration
  • 64-bit architecture for memory space over 4GB
  • 16/32-bit mixable instruction format for compacting code density
  • Branch predication to speed up control code
  • Return Address Stack (RAS) to speed up procedure returns
  • Memory Management Unit (MMU), Physical Memory Protection (PMP) and programmable Physical Memory Attribute (PMA)
  • Level-1 and level-2 cache controllers with 64-byte cache line size
  • Flexibly configurable Platform-Level Interrupt Controller (PLIC) for supporting wide range of system event scenarios
  • Enhancement of vectored interrupt handling for real-time performance
  • Advanced CoDense™ technology to reduce program code size

AndesCore™ AX45MP 64-bit multicore CPU IP is an 8-stage superscalar processor based on AndeStar™ V5 architecture. It supports RISC-V standard “G (IMA-FD)” extensions, “C” 16-bit compression instructions, DSP/SIMD ‘P’ extension (draft),  and Andes performance/ functionality enhancements for faster memory accesses and branch handling, plus Andes Custom Extension™ (ACE) to add user-defined instructions. It features MMU for Linux based applications, branch prediction for efficient branch execution, level-1 instruction/data caches and local memories for low-latency accesses.

The AX45MP symmetric multiprocessor supports up to 8 cores and a level-2 cache controller with instruction and data prefetch. Coherence manger implements MESI protocol to manage level-1 cache coherence, including I/O coherence for cacheless bus masters. Other AX45MP features include ECC for level-1/2 memory soft error protection, Platform-Level Interrupt Controller (PLIC) with enhancements for vectored dispatch and priority-based preemption, CoDense™ and StackSafe™ for software quality improvement, and suspend to standby/memory, CPU HotPlug, PowerBrake, and WFI for power managemen

AndesCore™ AX25MP Multicore

Applications

  • High performance application processor
  • Machine/Deep learning acceleration
  • Video and image processing
  • Large-scale network controllers
  • ADAS/V2X/IVI

Block Diagram

Development Tools

  • AndeSight™ IDE (Eclipse-based)
    • Compiler, Debugger, Profiler, Register Bit-field Display/Update, RTOS Awareness, and more
    • Tested platforms: Windows and Linux, single- or multi-user installation
  • Andes BSP
    • Optimized DSP library
    • Demo examples and sample projects
  • RTOSes
    • Open-source: FreeRTOS, Zephyr, RT-Thread
    • Commercial: ThreadX, μC/OS-II
  • SMP Linux kernel and platform drivers
  • COPILOT: Custom-OPtimized Instruction deveLOpment Tool for ACE
  • FPGA Development Boards
  • Debugging Hardware
    • AICE-MICRO and AICE-MINI+

Key Features and Performance

AndeStar™ V5 Architecture

Key FeaturesBenefits
RISC-V RV64GCP instructions
  • State-of-the art ISA from latest developments in computer architecture
  • Industry standard and open architecture
Andes Extended InstructionsAndes exclusive performance and functionality enhancements
Andes Custom Extension™ (ACE) option to create custom instructions for software acceleration
  • Add custom extensions to facilitate Domain-Specific Architecture/Acceleration (DSA)
  • Boost application performance significantly, at the same time maintain the programmability
  • Powerful constructs are available to define high level instruction
  • ACE design is based on Verilog and C languages which are familiar to the designers
  • The COPILOT tool automatically generates the extended CPU and software toolchain
  • Do not require expertise in processor pipeline to design ACE instructions
16/32-bit mixable instruction formatFor compact code density
32 general-purpose registersFor better code size and performance
Machine (M), User (U) and Supervisor (S) Privilege levelsFor Linux and advanced operating systems with protection between kernel and user programs

CPU Core

Key FeaturesBenefits
5.63 Coremark/MHz, 3.27 DMIPS/MHz*Superior performance-per-MHz
8-stage in-order superscalar pipelineSuperior performance-efficiency, while allowing for high speeds

Extensive branch predication features

  • Branch Target Buffer (BTB)
  • Branch Histroy Table (BHT)
  • Return Address Stack (RAS)
  • Branch Target Buffer and Branch History Table to speed up control codes
  • Return Address Stack to speeds up procedure returns

MMU (Memory Management Unit)

  • Sv39, Sv48 virtual-memory systems
  • 4/8-entry fully associative ITLB/DTLB
  • 32-512-entry 4-way set-associative shared TLB
  • Hardware page table walker
  • Virtual memory support for full address space and easy code/data sharing
  • Support for full-featured OS such as Linux
  • Protection of supervisor and user privilege
  • Hardware for fast address translation
Physical Memory Protection (PMP), configurable up to 32 regionsBasic read/write/execute memory protection with minimum cost
Programmable Physical Memory Attribute (PMA), configurable up to 16 regions

Configurable memory attributes:

  • Memory, I/O, None
  • Cacheable/Non-cacheable
  • Write-back/Write-through
  • Read/write/read & write allocate, no allocate
  • Access fault for non-existent regions
Performance monitorsProgram code performance tuning
StackSafe™ hardware stack protection
  • Easy identification of stack size threshold during development
  • Hardware error detection of stack overflow and underflow at runtime

Multiplier options

  • Fast multiplier: pipelined, 2-cycle
  • Small multipliers: producing 1, 2, 4, or 8 bits per cycle
Option to choose between speed and area according to application's requirements
PowerBrake technologyPerformance throttling to digitally reduce power consumption

Cluster power management

  • Suspend to Standby
  • Suspend to Memory
  • CPU Hotplug
Different low power modes for different use cases

* BSP v5.1.0, DMIPS/MHZ follow Dhrystone’s no-inline ground rules, best performances 

Memory Subsystems

Key FeaturesBenefits

Level-1 I-Cache & D-Cache

  • Size: 8KB to 64KB
  • Cache line size: 64 bytes
  • Set associativity: 2-way or 4-way
  • Accelerating accesses to slow memories
  • Flexible cache configurations

Level-2 I/D Unified Cache

  • Configurable from 128KB to 8MB
  • 64-byte cache line size
  • 16-way, pseudo random replacement
  • 2 tag banks, 2 data banks with interleaving
  • Configurable memory cycles for SRAM timing
  • Capture large working-set instruction and data to improve performance
  • Flexible selections to meet performance and timing requirements

ILM & DLM

  • Size: 4KB to 16MB
  • SRAM interface support
  • Bus masters accessed by AXI slave port
  • For deterministic and efficient program execution
  • Flexible size selection to fit diversified needs
MemBoost Optimize cache reads and writes to achieve higher bandwidth and lower latency
Optional ECC error protection with SRAM interfaceCode and data integrity protection
Bus master port: AXI with 128/256-bit data, I/D joint or separate busHigh throughput with wide data path
BUS Slave Port: AXI with 128/256-bit data, for ILM/DLM accessesEfficient data transfer between CPU and SoC masters
Core/bus clock ratio of N:1Simplified SoC integration

Multicore Cache Coherence

Key FeaturesBenefits
  • Support up to 8 cores
  • MESI cache coherence protocol
  • 128/256-bit I/O coherence port for cacheless bus masters
  • Symmetric multicore and L2 cache controller with cache coherence between level-1 (L1) caches and I/O coherence for bus masters without caches
  • Convenient and efficient interface for SoCs with rich I/O transactions

Platform-Level Interrupt Controller (PLIC)

Key FeaturesBenefits

Implements RISC-V PLIC specification

  • Up to 1023 PLIC interrupt sources
  • Up to 255 PLIC interrupt priority levels
  • Up to 16 PLIC interrupt targets
Allow individual interrupts to be serviced and prioritized without sharing

Enhanced interrupt features

  • Vectored interrupt dispatch
  • Priority-based preemption
  • Selectable edge trigger or level trigger
  • Faster interrupt handling for real-time applications
  • Complete hardware preemption support for faster response
  • Flexible interrupt source interface for simpler SoC design

Debug Support

Key FeaturesBenefits
Implements RISC-V debug specifications ver 0.13Supported by industry debug tool suppliers
JTAG Debug PortIndustry-standard support
Embedded Debug Module with up to 8 triggersFlexible configurations to tradeoff between gate count and debugging capabilities
Exception redirection supportEntering debugger upon selected exceptions without using breakpoints

Trace Support

Key FeaturesBenefits
Implements RISC-V Trace 1.0 Instruction Trace InterfaceSupported by Andes tools

Product Package