Vendor: Baya Systems Category: Coherency

Configurable Cache Coherent Fabric IP

Scalable Coherent Connectivity for Advanced SoCs and Chiplet Systems

Overview

Scalable Coherent Connectivity for Advanced SoCs and Chiplet Systems

Advanced SoCs and chiplet-based systems need coherent fabrics that can efficiently connect CPUs, accelerators, memory, I/O, and peer-to-peer traffic while maintaining data consistency across the system. As coherent traffic scales across more agents, memories, and chiplets, the fabric must balance bandwidth, latency, cache coherency, QoS, and physical implementation requirements.

Baya’s Configurable Cache Coherent Fabric IP provides a scalable Network-on-Chip for systems where major traffic is coherent. It supports AMBA CHI interfaces for fully coherent agents, AXI interfaces for non-coherent and I/O-coherent agents, and ACE-Lite connectivity for I/O-coherent devices, enabling coherent, non-coherent, and peer-to-peer traffic to share a unified fabric architecture.

Configured and optimized through FabricStudio, the fabric gives teams flexible topology control, cache coherence controllers, built-in system-level cache, QoS, RAS, and physically aware implementation capabilities. It helps engineering teams build high-bandwidth, low-latency coherent interconnects that scale across advanced SoCs and multi-chiplet systems while reducing design risk and accelerating time to market.

 

Key features

Scalable Cache Coherent Architecture

Build cache-coherent fabrics across large SoCs and multi-chiplet systems with support for up to 256 chiplets, 128 HN slices, and flexible coherent topology control.

Flexible CHI, AXI, and ACE-Lite Connectivity

Connect fully coherent CHI agents, AXI-based system IP and memory controllers, ACE-Lite I/O-coherent devices, PCIe controllers, and non-coherent fabric through a unified system fabric.

System-Level Cache and Coherency Features

Support up to 512 MB of system-level cache, configurable snoop filters, UMA/NUMA/sub-NUMA modes, atomics, cache maintenance operations, and advanced coherency features.

Bandwidth Management and QoS

Use weighted bandwidth allocation, flow priority, virtual channel isolation, route selection, input rate limiting, and bandwidth/latency monitoring to meet system-level performance targets.

Physically Aware, Deployment-Ready Design

Enable modular implementation with advanced tiling, hierarchical design generation, RAS, debug, trace, ECC protection, runtime repair, and PPA optimizations for scalable physical design.

Specifications

Identity

Part Number
Coherent Fabric
Vendor
Baya Systems
Type
Silicon IP

Files

Product Sheet Download

Note: some files may require an NDA depending on provider policy.

Provider

Baya Systems
HQ: USA

Learn more about Coherency IP core

How Cache Coherency Simplifies AI Software

This article explains how hardware-managed cache coherency can simplify AI software by coordinating shared data across CPUs, accelerators and chiplets. The article highlights the role of coherent and non-coherent interconnects in building scalable, efficient AI SoCs, while presenting Arteris’ Ncore, FlexNoC and FlexGen IP as complementary solutions for managing data movement.

SoC design: When a network-on-chip meets cache coherency

Many people have heard the term cache coherency without fully understanding the considerations in the context of system-on-chip (SoC) devices, especially those using a network-on-chip (NoC). To understand the issues at hand, it’s first necessary to understand the role of cache in the memory hierarchy.

Frequently asked questions about Coherency Interconnect IP cores

What is Configurable Cache Coherent Fabric IP?

Configurable Cache Coherent Fabric IP is a Coherency IP core from Baya Systems listed on Semi IP Hub.

How should engineers evaluate this Coherency?

Engineers should review the overview, key features, supported foundries and nodes, maturity, deliverables, and provider information before shortlisting this Coherency 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.

×
Semiconductor IP