Enhancing design space exploration by extending CPU/GPU specifications onto FPGAs

The design cycle for complex special-purpose computing systems is extremely costly and time-consuming. It involves a multiparametric design space exploration for optimization, followed by design verification. Designers of special purpose VLSI implementations often need to explore parameters, such as optimal bitwidth and data representation, through time-consuming Monte Carlo simulations. A prominent example of this simulation-based exploration process is the design of decoders for error correcting systems, such as the Low-Density Parity-Check (LDPC) codes adopted by modern communication standards, which involves thousands of Monte Carlo runs for each design point. Currently, high-performance computing offers a wide set of acceleration options that range from multicore CPUs to Graphics Processing Units (GPUs) and Field Programmable Gate Arrays (FPGAs). The exploitation of diverse target architectures is typically associated with developing multiple code versions, often using distinct programming paradigms. In this context, we evaluate the concept of retargeting a single OpenCL program to multiple platforms, thereby significantly reducing design time. A single OpenCL-based parallel kernel is used without modifications or code tuning on multicore CPUs, GPUs, and FPGAs. We use SOpenCL (Silicon to OpenCL), a tool that automatically converts OpenCL kernels to RTL in order to introduce FPGAs as a potential platform to efficiently execute simulations coded in OpenCL. We use LDPC decoding simulations as a case study. Experimental results were obtained by testing a variety of regular and irregular LDPC codes that range from short/medium (e.g., 8,000 bit) to long length (e.g., 64,800 bit) DVB-S2 codes. We observe that, depending on the design parameters to be simulated, on the dimension and phase of the design, the GPU or FPGA may suit different purposes more conveniently, thus providing different acceleration factors over conventional multicore CPUs.

Ultra-Compact and Robust FPGA-Based PUF Identification Generator

Physically Unclonable Functions (PUFs), exploit inherent manufacturing variations and present a promising solution for hardware security. They can be used for key storage, authentication and ID generations. Low power cryptographic design is also very important for security applications. However, research to date on digital PUF designs, such as Arbiter PUFs and RO PUFs, is not very efficient. These PUF designs are difficult to implement on Field Programmable Gate Arrays (FPGAs) or consume many FPGA hardware resources. In previous work, a new and efficient PUF identification generator was presented for FPGA. The PUF identification generator is designed to fit in a single slice per response bit by using a 1-bit PUF identification generator cell formed as a hard-macro. In this work, we propose an ultra-compact PUF identification generator design. It is implemented on ten low-cost Xilinx Spartan-6 FPGA LX9 microboards. The resource utilization is only 2.23%, which, to the best of the authors' knowledge, is the most compact and robust FPGA-based PUF identification generator design reported to date. This PUF identification generator delivers a stable range of uniqueness of around 50% and good reliability between 85% and 100%.

Constructive Synthesis of Memory-Intensive Accelerators for FPGA From Nested Loop Kernels

Field-programmable gate arrays are ideal hosts to custom accelerators for signal, image, and data processing but de- mand manual register transfer level design if high performance and low cost are desired. High-level synthesis reduces this design burden but requires manual design of complex on-chip and off-chip memory architectures, a major limitation in applications such as video processing. This paper presents an approach to resolve this shortcoming. A constructive process is described that can derive such accelerators, including on- and off-chip memory storage from a C description such that a user-defined throughput constraint is met. By employing a novel statement-oriented approach, dataflow intermediate models are derived and used to support simple ap- proaches for on-/off-chip buffer partitioning, derivation of custom on-chip memory hierarchies and architecture transformation to ensure user-defined throughput constraints are met with minimum cost. When applied to accelerators for full search motion estima- tion, matrix multiplication, Sobel edge detection, and fast Fourier transform, it is shown how real-time performance up to an order of magnitude in advance of existing commercial HLS tools is enabled whilst including all requisite memory infrastructure. Further, op- timizations are presented that reduce the on-chip buffer capacity and physical resource cost by up to 96% and 75%, respectively, whilst maintaining real-time performance.

Programmable high-performance IIR filter chip

The paper presents a state-of-the-art commercial demonstrator chip for infinite impulse response (IIR) filtering. The programmable IIR filter chip contains eight multiplier/accumulators that can be configured in one of five different modes to implement up to a 16th-order IIR filter. The multiply-accumulate block is based on a highly regular systolic array architecture and uses a redundant number system to overcome problems of pipelining in the feedback loop. The chip has been designed using the GEC Plessey Semiconductors CLA 78000 series gate array, operates on 16-bit two's complement data and has a clock speed of 30 MHz. Issues such as overflow detection and design for testability have also been addressed and are described.

High-radix systolic modular multiplication on reconfigurable hardware

The overall aim of the work presented in this paper has been to develop Montgomery modular multiplication architectures suitable for implementation on modern reconfigurable hardware. Accordingly, novel high-radix systolic array Montgomery multiplier designs are presented, as we believe that the inherent regular structure and absence of global interconnect associated with these, make them well-suited for implementation on modern FPGAs. Unlike previous approaches, each processing element (PE) comprises both an adder and a multiplier. The inclusion of a multiplier in the PE means that the need to pre-compute or store any multiples of the operands is avoided. This also allows very high-radix implementations to be realised, further reducing the amount of clock cycles per modular multiplication, while still maintaining a competitive critical delay. For demonstrative purposes, 512-bit and 1024-bit FPGA implementations using radices of 2(8) and 2(16) are presented. The subsequent throughput rates are the fastest reported to date.