Multi-core capable RISC-V processor with vector extensions
The SiFive® Intelligence™ X280 Gen 2 is an 8-stage dual issue, in-order, superscalar design with wide vector processing (512 bit …
Overview
The SiFive® Intelligence™ X280 Gen 2 is an 8-stage dual issue, in-order, superscalar design with wide vector processing (512 bit VLEN/256-bit DLEN). It supports RISC-V Vectors v1.0 (RVV 1.0) and SiFive Intelligence Extensions to accelerate critical AI/ML operations.
The X280 Gen 2 is scalable up to a 4-core cache-coherent complex. With optional SiFive Scalar Coprocessor Interface (SSCI) and Vector Coprocessor Interface eXtensions (VCIX) for close coupling with customer-designed external AI accelerators or other coprocessors.
Compared to the highly popular first generation, the new product has substantial upgrades including a more configurable and efficient memory subsystem, new hardware accelerated AI instructions, support for new datatypes, and a configurable Vector Load Data Queue (VLDQ) to tolerate memory latencies.
Key features
- New in Second Generation
- New RVA23 support
- New RVV1.0 512-bit vector engine
- New SSCI interface added alongside VCIX
- New instructions and extensions
- New improved memory subsystem
- New memory latency tolerance
- New HW support of BF16 datatype
- New vector crypto support
- Plus many other new features, enhancements and upgrades
- 512-bit vector register length processor
- Variable length operations, up to 512-bits of data per cycle
- Ideal balance of control logic and data parallel compute
- Decoupled Vector pipeline
- INT8 to INT64 integer data type
- BF16/FP16/FP32/FP64 floating point data type
- SiFive Intelligence Extensions for ML workloads
- Custom instructions to greatly accelerate Neural Network computation
- Optimized IREE MLIR AI/ML software stack and TensorFlow Lite implementation
- Hundreds of Neural Network models ported
- Built on silicon-proven U7-Series core
- 64-bit RISC-V ISA
- 8-stage dual-issue in-order pipeline
- Coherent multi-core, Linux capable
- High performance vector memory subsystem
- Memory parallelism provides cache miss tolerance
- Virtual memory support with precise exceptions
- Up to 48-bit addressing
- Multi-core, multi-cluster processor configuration, up to 4 cores per cluster
- RISC-V 64-bit Linux capable scalar/vector processing for standalone AI or Accelerator Control and Assist at the Edge
Block Diagram
Benefits
- Performance: SiFive Intelligence Extensions enable enhanced functionality and acceleration instructions for matrix computation so that the X288 processor can achieve a best-in-class 4.5 TFLOPs (bfloat16 MatMul) or 9.2 TOPS (INT8 MatMul) for AI/ML computation acceleration.
- Power Efficiency: The X288 exhibits superior power efficiency due to its ability to turn off various sections of the chip when not in use.
- Longevity: RISC-V architecture is a safe bet for long-lived ecosystem support due to open standard ISA
Specifications
Identity
Compute
Files
Note: some files may require an NDA depending on provider policy.
Provider
Learn more about CPU IP core
The Future of AI is Modular: Why the SiFive-NVIDIA Milestone Matters
X100 - Securing the System - RISC-V AI at the Edge
Building the Future of AI on Intelligent Accelerators
RISC-V for Infrastructure: For Now, It’s All About the Developer
SiFive HiFive: The Vital Role of Development Boards in Growing The RISC-V Ecosystem + HiFive Premier P550 Update
Frequently asked questions about CPU IP cores
What is Multi-core capable RISC-V processor with vector extensions?
Multi-core capable RISC-V processor with vector extensions is a CPU IP core from Sifive, Inc. listed on Semi IP Hub.
How should engineers evaluate this CPU?
Engineers should review the overview, key features, supported foundries and nodes, maturity, deliverables, and provider information before shortlisting this CPU IP.
Can this semiconductor IP be compared with similar products?
Yes. Buyers can compare this product with similar semiconductor IP cores or IP families based on category, provider, process options, and structured technical specifications.