RapidIO® 4.x BRC1 LP-Serial Memory Mapped Endpoint Core
Features
-
Implements a complete RapidIO Endpoint solution.
-
Compliant with Rev. 4.1 of the specification.
-
Implements RapidIO Error Management Extensions.
-
Supports 1x, 2x, and 4x link widths at BRC1 data rates.
-
Supports the High Availability/Redundant System Hardware (HARSH) profiles
-
Supports the MECS Time Synchronization Protocol.
-
AXI4 interfaces for target, initiator, and messaging functions.
-
Management Entity with integrated decoder for RapidIO maintenance transactions.
-
IP-XACT support for rapid integration.
-
Available for the following AMD and Intel FPGA families:
-
AMD Versal
-
Intel Agilex 7
-
Intel Agilex 5
-
-
Porting kit is available for ASIC targets.
Description
The BRC1 LP-Serial Memory Mapped Endpoint Core implements a complete RapidIO endpoint core capable of supporting link rates of up to 5.0 Gbps per lane. It includes shared memory and messaging logical layer functions. It also includes functions that allow the endpoint to be managed with or without a local processor. Figure 1 illustrates the overall architecture of the core.
The core integrates an IO Logical Layer Initiator
block that maps AXI4 transactions into RapidIO IO
Logical request packets. The IO Logical Layer Target
block maps RapidIO IO Logical request packets into
AXI4 transactions on the user interface. Together these
functions support the transparent, low latency sharing
of memory across the link.
The HARSH profile support provide a range of
features that support the creation of high availability
systems. This includes MECS timestamp synchronization
that coordinates timing functions across the system
with nano-second accuracy. The error management
functions provide detailed traceability of system level
errors along with optional error timestamping to assist
with error analysis.
RapidIO®4.x BRC1 LP-Serial Memory Mapped Endpoint Core