 |
 |
 |

Request more info on this core
PDF Datasheets
Certification

The x1 and x4 versions of this core have passed the PCI-SIG PCI Express
compliance tests and appear on their Integrators
List.
PCIe Simulation Model
Register for a free download to simulate the CPXP-EP Endpoint Controller functioning in a sample system.
PCIe PHYs
The core is designed to work with any 16-bit PIPE-compliant physical layer (PHY). Internal testing and customer experiences have verified this with PHYs from several sources, including:
Altera, ARM, Genesys Logic, Jmicron, NXP, Snowbush, and Xilinx.
Please contact us for more details.
PCIe Development Boards
We conduct physical interoperability testing with systems and boards from a variety of trusted vendors, including Agilent, Altera, Dini Group, Hardi (now Synplicity) and Genesys Logic. Please contact us for more details.
PCIe Chipset Interoperability
While the core is rigorously designed to conform to the specification, we verify this through real-world testing with popular host chipsets, including :
• Intel I915
• Intel I945
• Intel I965
• Intel I975
• nVidia nForce 430
• nVidia nForce 550
• nVidia nForce 570 SLI
• nVidia nForce 590 SLI
• nVidia MPC55 Tritium
• nVidia MPC61
• nVidia C55 Tritium
• VIA PT890Pro
• SIS 986
• SIS 671FX
Please contact us for more details.
PCIe Core Links
Technical Article
News Releases
06/04/07
CAST Offers PCI Express Model, Highlights APS 32-bit Processor
Cores at DAC 2007
04/11/07
CAST Selects Avery Design Systems for PCI Express Verification
IP
06/13/05
CAST Announces PCI Express Endpoint Controller IP Core
Presentation
|
 |

|
 |
|
CPXP-EP
PCI Express Endpoint Controller Core
On this page: Description
| Implementation Results | Features
| Applications | Block Diagram
| Functional Description | Support
| Verification | Deliverables
Implements a PCI Express endpoint controller that
is compliant with PCI Express Base specification 1.1, including the Transaction,
Data Link, and Physical protocol layers.
The scalable and flexible core has a modular architecture and a high-performance,
low-latency design. It supports multiple device link widths to better
match the bandwidth needs of specific applications—x1 (single lane)
and x4 (four lane)—and offers bi-directional data rates from 250MB/s
(x1) to 1GB/s (x4). It supports most advanced PCI Express capabilities,
including message signaled interrupts, multiple virtual channels, advanced
error reporting, end-to-end cyclic redundancy check, and power management
features. Multi-lane versions of the core support lane reversal and polarity
inversion.
The core has an Application Interface (AIF) layer that makes integration
significantly easier than the transaction layer packet (TLP) interface
typically provided. This AIF includes a DMA controller and provides a
higher-level connection between the core logic and the user’s application.
Interfaces to standard system buses such as Wishbone or AMBA AHB are available.
The external connection interface from the core conforms to the Intel®
PIPE specification, ensuring compatibility with any 16-bit PIPE-compliant
physical layer (PHY). The core has been successfully used with PHYs from
multiple vendors.
The synchronous, latch-free core design was rigorously verified for
compliance with the PCI Express specification, and has passed PCI-SIG
certification. The core has been tested for interoperability with multiple
motherboards using chipsets from various vendors, and is in use by multiple
customers.
See representative implementation results (each
in a new pop-up window):

Features
- Compliant with PCI Express Base Specification 1.1
- Implements Transaction, Data Link, and Physical protocol layers in
hardware
- Supports x1 and x4 link widths
- Offers a data rate of 2.5 Gbps per lane
- Supports up to eight Virtual Channels
- Supports lane reversal and polarity inversion
- PCI Configuration space type 0 header
- MSI capability support
- End-to-end cyclic redundancy code (ECRC) generation and checking support
- Advanced Error Reporting capability support
- Configurable TLP data payload size, from 128B to 4kB
- Configurable size for the Transmit Retry and Receive data buffers
- Modular architecture
- Synchronous design
- 64-bit internal datapath at 125MHz •
Integration
- Application Interface (AIF) for easier system integration using industry
standard bus interfaces (e.g., Wishbone, AMBA); handles up to 8 DMA
channels
- Conforms to standard PIPE interface for compatibility with any 16-bit
PIPE-compliant PHY
Certification
- This core has passed the PCI-SIG PCI Express compliance tests, and
appears on their Integrators List.
Applications
PCI Express is rapidly being adopted for a variety of interconnection
applications, including:
- Network servers
- Graphics and multimedia
- Communications and mobile product
- Industrial, automotive, and other embedded systems
Block Diagram

Functional Description
The core is divided into four modules responsible for the Configuration
Space, Transaction Layer, Data Link Layer, and Physical Layer MAC.
Configuration Space
Provides a Configuration space register file type 0. In addition to
the mandatory functions, numerous extended capabilities are also supported.
Transaction Layer module
Responsible for the assembly and disassembly of transaction layer packets.
The transaction layer supports four address spaces: Configuration space,
Memory space, I/O space and Message Space. Power management services are
supported.
Data Link Layer module
Responsible for link management, data protection and integrity checking,
retry and power management services.
Physical Layer MAC module
Implements the logical sub-block of the physical layer. The module is
responsible for link training and status monitoring, link width negotiation,
lane order negotiation, lane polarity reversal control and power management
implementation.
Application Interface Module

Provides a higher-level interface that makes system integration easier
than working at the TLP level. Includes a DMA controller and other functions
for straightforward interfacing of the core to standard SoC buses. AMBA
AHB and Wishbone versions are available.
The Completion controller automatically handles read and write requests
and sends completions if needed.
The configurable DMA controller provides up to eight independent DMA
channels.
Support
The core as delivered is warranted against defects for ninety days from
purchase. Thirty days of phone and email technical support are included,
starting with the first interaction. Additional maintenance and support
options are available.
Verification
The core has been verified using Avery's PCI-Xactor verification tool, and has passed
PCI-SIG certification. Interoperability has been verified with Intel,
nVidia, VIA, and SIS chipsets, and PHYs from GeneSys Logic, Altera, and
XIlinx have been successfully used. The core is in use by multiple customers.
Deliverables
The core is available in ASIC (synthesizable HDL) or FPGA (netlist)
forms, and includes everything required for successful implementation:
- ASIC cores: HDL RTL source of the CPXP-EP and CPXP-EP-AIF
FPGA cores: Post-synthesis EDIF netlist of the CPXP-EP and CPXP-EP-AIF
- Sophisticated HDL Testbench including external FIFOs, buffers, models
of interfaces, and the core
- Simulation script, vectors, expected results, and comparison utility
- Synthesis or place and route script
- Comprehensive user documentation
On this page: Description
| Implementation Results | Features
| Applications | Block Diagram
| Functional Description | Support
| Verification | Deliverables
Download PDF datasheets for more info: ASIC
| Altera | Xilinx
|