Finish Faster With Xilinx DSP Design Solutions: Corporate Headquarters European Headquarters Japan Asia Pacific

Download as pdf or txt
Download as pdf or txt
You are on page 1of 2

Finish Faster with Xilinx DSP Design Solutions

Xilinx System Generator for DSP

Pre-verified Signal Processing Algorithms as IP Cores

The industrys premier solution for FPGA-based DSP design enables


you to:

Xilinx and partners provide a range of DSP IP cores that are


optimized for speed and cost.

Generate high-performance DSP algorithms in FPGAs from a


high level executable specification in Simulink

FEC: Reed-Solomon, Viterbi, TCCs, and others

Accelerate simulations by orders of magnitude using your target


hardware in the loop with Simulink or ModelSim

Math functions: CORDIC, Multiplier, MACs, and others


Video IPCompression scaling and others

Import VHDL and Verilog modules directly into Simulink using a


ModelSim co-simulation interface

Todays signal-processing systems

Industry-standard DSP connectivity with Serial RapidIO, PCI,


and EMIF

demand increasingly higher

FFTs, Filters, and others

Specify state machine and other logic using MATLAB code that is
automatically compiled into RTL HDL

performance and flexibility.


Xilinx Virtex-4 FPGAs are ideally
suited for high-performance signalprocessing tasks traditionally serviced
by an ASIC or ASSP. They allow you to
create high-performance DSP engines
that can boost the performance of
your programmable DSP system by
performing complementary
co-processing functions in digital
communications, video/imaging, and
other applications. The Virtex-4 FPGA
family is the newest and most powerful
addition to the Xilinx XtremeDSP
solution, providing blazing DSP performance with unrivalled economy. With up
to 512 XtremeDSP slices operating at

DSP Services and Support for Virtex-4 FPGAs

500 MHz, these devices can implement


DSP Education Services

DSP Support Services

DSP Design Services

Reduce your time-to-knowledge with public,


private, and online courses including:

Ensure the success of your DSP project with


award-winning technical support, including:

Reduce your project risk by allowing our


engineers to help you with:

DSP Design Flow (three-day course)

Industrys best support web site

System architecting

DSP Implementation Techniques for Xilinx


FPGAs (three-day course)

Free DSP hotline support

FPGA implementation

Platinum DSP hotline support

IP core modification

Titanium on-site AE support

Turnkey system design

complex tasks such as:


Hundreds of IF-to-baseband down
conversion channels
128X chip-rate processing for spreadspectrum systems
High-definition H.264 and MPEG-4

Take the Next Step

encode/decode algorithms

Learn more about achieving blazing DSP performance with unrivalled economy. Visit us online at www.xilinx.com/virtex4

Corporate Headquarters

European Headquarters

Japan

Asia Pacific

Xilinx, Inc.
2100 Logic Drive
San Jose, CA 95124
Tel: (408) 559-7778
Fax: (408) 559-7114
Web: www.xilinx.com

Xilinx, Ltd.
Citywest Business Campus
Saggart,
Co. Dublin
Ireland
Tel: +353-1-464-0311
Fax: +353-1-464-0324
Web: www.xilinx.com

Xilinx, K.K.
Shinjuku Square Tower 18F
6-22-1 Nishi-Shinjuku
Shinjuku-ku, Tokyo
163-1118, Japan
Tel: 81-3-5321-7711
Fax: 81-3-5321-7765
Web: www.xilinx.co.jp

Xilinx, Asia Pacific


Unit 1201, Tower 6, Gateway
9 Canton Road
Tsimshatsui, Kowloon,
Hong Kong
Tel: 852-2-424-5200
Fax: 852-2-494-7159
E-mail: [email protected]

The XtremeDSP solution accelerates

The Programmable Logic CompanySM

All three Virtex-4 platforms offer XtremeDSP capabilities. Choose the device that provides
the optimal ratio of DSP performance for your unique application.
Virtex-4 SX devices offer the most cost-effective implementation of ultra-highperformance DSP functionality, with the highest ratio of XtremeDSP slices
up to 512 slices delivering up to 256 GMACS performance
Virtex-4 LX devices offer ample XtremeDSP slices and add more logic, memory,
and I/O resources
Virex-4 FX devices add embedded PowerPC processors and RocketIO multi-gigabit
transceivers

Easiest-to-use Design Solutions for FPGA-based DSP


Xilinx and its partners provide complete solutions for rapid DSP development and
implementation.
Reduce design time with System Generator for DSP
Implement fast, highly optimized algorithms with a rich DSP IP library
Bring products to market faster with award-winning technical support and DSP services

property cores, and design services.


verifying, and deploying your DSP algorithms

PN 0010801

Optimized Performance and Cost for Your DSP Applications

superior devices, design tools, intellectual

and systems in FPGAs.

Printed in the U.S.A.

The 500 MHz XtremeDSP slice delivers unmatched versatility, efficiency, and performance.
Configure each XtremeDSP slice for over 40 DSP functions, such as multiply-accumulate,
multiply, addition, and multiplexing
Reduce DSP power consumption by 86% (57W/MHz) and save precious logic resources
for other tasks
Cascade multiple XtremeDSP slices at full system speed to build complex filters and
multi-precision functions

your products time-to-market through

This gives you the fastest means of designing,

2004 Xilinx Inc. All rights reserved. The Xilinx name is a registered trademark; Virtex-4, RocketIO, XtremeDSP, PicoBlaze, and MicroBlaze are trademarks; and The Programmable Logic Company is a service mark of Xilinx Inc. Power PC is a trademark of International
Business Machines Corporation in the United States, or other countries, or both. All other trademarks are the property of their owners.

XtremeDSP Slice Delivers Maximum Performance and Efficiency

Versatile 500 MHz XtremeDSP Slices

Integrated Hard and Soft Microprocessors

The Challenge: Implement high-performance DSP algorithms


more cost-effectively.

The Challenge: Complex control and RTOS implementation.

The Virtex-4 Solution: Up to 512 new XtremeDSP slices

The Virtex-4 Solution: A broad selection of 8- to 32-bit microprocessor systems and operating system support
(VxWorks, Integrity, Linux, etc.)

500 MHz throughput (256 GMACS overall performance)

Hard 32-bit IBM PowerPC 405 cores in FX platforms for implementing advanced
frameworks such as Software Communications Architectures (SCAs) for software-defined
radio applications

in 4VSX55
40+ arithmetic functions
1/7th the power compared to previous-generation FPGAs

Xilinx PicoBlaze and MicroBlaze soft microprocessors for control circuits

Directly cascadeable without loss in speed

Whether you are working with spread-spectrum, multi-carrier, or


narrowband communication systems, Virtex-4 FPGAs are the ideal choice.

Example: Cable Head-end System

Example: Wireless Base station

Spread Spectrum
(eg. 3GPP2)

Digital Radio Card


Spectrum
Channelization

Use the new XtremeDSP


slices to efficiently
implement:

Internet

Web &
Application
Servers
Remultipliers
(StatMuxes)

Digital Radio Functions


Baseband Functions

TV
Broadcast

Receivers

XtremeDSP slices are

DDC

Chip Rate
RACH Rx
Path Profiler
TCC
Viterbi

Polyphase Transform
Matched Filter

used in the receive path


(shown adjacent) and the
transmit path for
corresponding transmit
functions.

Block Parameters:
MPEG Encoder
Scaling, filtering
Motion estimation hierarchical
1/2 or 1/4 pel resolution
Deblocking
Variable length coder (VLC)
DCT

CMTS Router
Backplane (e.g. Fibre Channel)

Narrowband

Baseband Card
QAM
Adaptive EQ
Carrier/Timing
Recovery
Reed Solomon
Viterbi
De/Interleaver

Video Servers

Modulator

OFDM
MIMO
TCC
LDPC

Multi-Carrier

Compact Multi-Channel Designs using SRL16s


The Challenge: Keep cost and power down for multi-channel
signal-processing designs
The Virtex-4 Solution: Unique SRL16s enable you to achieve
very high compute density and make efficient use of logic slices.

Powerful Serial and Parallel Interfaces


The Challenge: Need to interface to DSP processors, memory, and other systems
The Virtex-4 Solution: Extremely flexible I/O interfaces

Single-Channel
Reed Solomon Encoder Replicated
16 times (without using SRL16s)

DSP Processors &


ADCs and DACs
Serial RapidIO
EMIF etc.
LVDS etc.

External Memories

System Interfaces

DRAM
DDR2, DDR
SDRAM, RLDRAM II
FCRAMII

Serial RapidIO

SRAM
QDRII, ZBT

PCI Express
PCI
HD-SDI
Aurora

Using a single SRL16, a 16-channel


Reed-Solomon encoder consumes only
10% of the silicon area compared to a
design that simply replicates a single
channel version 16 times.

Efficient 16-channel RS-Encoder


using a Single SRL16

You might also like