Is this your business? Claim it to manage your IP and profile
Engineered specifically for image processing applications, Dillon Engineering's 2D FFT core provides a two-dimensional transformation with efficiency and speed. This core is indispensable in scenarios where extensive data interactions involve both dimensions of the image data, necessitating a meticulous approach to manage throughput, scaling, and memory utilization. The architecture supports both on-chip and external memory transposes, ensuring optimized data handling for peak performance. By leveraging robust scaling options and continuous processing capabilities, Dillon Engineering's 2D FFT core adeptly manages complex data sets with the agility and speed necessitated by modern image processing tasks. Its implementation seamlessly interfaces with various memory configurations, solidifying its place as a versatile and scalable solution for complex signal processing tasks.
The Pipelined FFT core from Dillon Engineering is designed for seamless, continuous data processing, executing FFT calculations at the rate of one point per clock cycle. Particularly beneficial for applications that prioritize efficient memory use, the pipelined structure of this core is ideal for ASICs or scenarios where reducing memory footprint is essential. The core supports radial-2 length selections and offers variable runtime length options, which greatly enhances flexibility and adaptability in live processing applications. With advanced decimation techniques and optional buffering for normal order I/O, it sustains high data transformation speeds while minimizing resource use, ensuring optimal results even within constrained design parameters.
The UltraLong FFT core from Dillon Engineering is engineered for transforming massive data lengths that overshoot the internal memory capacities of FPGAs and ASICs. By partitioning the FFT process, results are stored in external memories, necessitating multiple transformations that require three transpose operations and additional processing stages. This broadens the applications of FFTs beyond conventional limits, particularly where large data throughput is essential while retaining efficiency in logic usage. The UltraLong FFT core is devised with an adaptable architecture, supporting various memory strategies and resource-sharing options. Its performance is optimized based on external memory bandwidth, with SRAM technologies typically offering superior throughput while SDRAM solutions allow for the longest FFT stretches. Dillon Engineering ensures these cores maximize efficiency aligned with the available memory architectures at hand, guaranteeing suitable solutions for voluminous data processing needs.
Designed with ASIC applications in mind, the Load Unload FFT core offers a minimal memory footprint suitable for environments prioritizing reduced ASIC area. This core incorporates a load-process-unload cycle that efficiently handles single data sets while offering configuration flexibility. The core is capable of processing fixed or floating-point math and provides run-time selections for FFT length and directional operations, heightening its adaptability in dynamic processing environments. By utilizing minimal memory and butterfly configurations, it stands as an optimal choice where area constraints dictate cost-effective solutions without performance sacrifices. With options to include input buffers, the Load Unload FFT core accommodates continuous data applications smoothly, maintaining robustness even when deployed in resource-challenged scenarios.
The Mixed Radix FFT core developed by Dillon Engineering enables the processing of diverse FFT lengths through the combination of radix-2, 3, 5, and 7 factors, making it ideal for applications requiring non-standard FFT lengths. This versatility is critical in fields like LTE OFDM, where specific frequency bin spacing is crucial. The core is capable of managing a wide assortment of data transformations, ensuring compatibility with various system architectures. By supporting both fixed and floating-point arithmetic, the core facilitates a tailored approach to each project, encompassing internal and external memory configurations to boost parallel processing performance. Mixed Radix FFT architectures are scalable, ensuring that users can optimize for performance and memory usage concurrently, tailored to their specific implementation scenarios.
Dillon Engineering’s Parallel FFT utilizes a comprehensive parallel architecture, optimizing efficiency for short FFT lengths ranging from 4 to 64 points. As one of the fastest and most power-efficient FFT architectures available, it is designed to maximize computational speed, handling gigasample data rates of over 25 GSPS for hefty, real-time challenges. This core architecture minimizes complexity by employing constant twiddle factors, which in turn reduces the logic usage required. The fully parallel, asynchronous pipeline structure supports ultra-high throughput, making it exceptional for low-latency applications. This adaptability extends to various data lengths, ensuring that users can leverage the full potential of the architecture according to their specific FFT requirements.
The AES Crypto core by Dillon Engineering offers comprehensive encryption and decryption solutions meeting FIPS 197 standards. Equipped to handle a wide range of data processing needs, this core supports multiple encryption modes including ECB, CBC, OFB, and others, allowing for dynamic changes in encryption keys without affecting throughput. It features a scalable architecture from HDL or targeted EDIF, adaptable to FPGA and ASIC environments, designed with efficiency in mind, offering high throughput and area performance trade-offs. This adaptability renders it suitable for various applications requiring robust security measures seamlessly integrated within existing workflows, leveraging the speedy encryption processes tailored to any application requirements.
Dillon Engineering's Floating Point Library core provides extensive IEEE 754 compliant modules optimally designed for precision and performance. By supporting a full spectrum of floating-point calculations with parametric adjustments using ParaCore Architect, each module can be tailored to specific application needs. This library is well-equipped for FPGA and ASIC deployments, controlled through compile-time parameters allowing for a balance between precision and resource allocation. Adaptable across single and double precision scenarios, the library facilitates complex implementations requiring nuanced calculations, fully supporting the integration of precise arithmetic operations while maintaining logical efficiency.
Join the world's most advanced semiconductor IP marketplace!
It's free, and you'll get all the tools you need to evaluate IP, download trial versions and datasheets, and manage your evaluation workflow!
To evaluate IP you need to be logged into a buyer profile. Select a profile below, or create a new buyer profile for your company.
