Our AT40KAL coprocessor series FPGAs offer from 5K to 50K usable gates and are designed for high-density, compute-intensive Digital Signal Processing (DSP) and other fast logic designs. The AT40KAL family resolves the logic vs. SRAM tradeoff by providing fast, flexible and distributed 10 ns FreeRAM without using valuable logic resources. Structured logic functions, including variable array multipliers, can be implemented directly in core cells without using any busing resources, providing dramatic improvements in speed, utilization, power and system cost. This FPGA technology is also available as an embedded core in our Field Programmable System Level Integrated Circuits (FPSLIC) devices, which combine a 5–50K AT40KAL FPGA with up to 36K of SRAM and a 25 MHz AVR® microcontroller (MCU).
A key feature of the AT40KAL family of FPGAs is their Cache Logic capability, which enables you to reprogram part of the FPGA without losing register data, while the remainder of the FPGA continues to operate without disruption. This is ideal for building adaptive filters, variable coefficient multipliers and other designs where the data path can be changed to increase system performance.
Offering massive register counts (1,024 to 6,400 registers), 5V AT40K FGPAs are designed to speed up processor-based system performance while reducing power consumption, part count and costs, making them ideal for use as re-configurable DSP coprocessors.
The AT40Kxx Coprocessor Series FPGAs can be designed using industry-standard EDA tools such as Mentor’s Precision Synthesis for VHDL and Verilog on a PC platform. We offer Integrated Development System [IDS] software that enables place and route while enabling you to synthesize your designs using industry-standard synthesis tools.
We also offer a family of FPGA Serial Configuration EEPROMs in densities ranging from 256 Kb to 32 Mb. This unique family of products can be programmed using a standard programmer or reprogrammed in-system like our FPGAs.