Coarse-Grained Reconfigurable Architectures (CGRAs) have been thoroughly explored as a promising solution for accelerating compute-intensive applications, offering a balance between flexibility and energy efficiency. Recently, CGRA designs have tried to handle arbitrary complex code constructs, often resulting in increased architectural complexity and inefficient use of Processing Elements (PEs), in particular for Address Generation Instructions (AGIs).This paper introduces Mage, a Decoupled Access-Execute (DAE) CGRA specifically optimised for Static Control Programs (SCPs), which are well-suited for DAE-based acceleration. By leveraging an SCP-tailored Address Generation Unit for affine access patterns computation, Mage maximises PE utilisation for data processing. Compared to other State-of-the-Art DAE CGRAs, Mage reduces area occupation by up to 5.7x while ensuring high area efficiency, reaching 5701.7 MOPs/mm2.

The increasing complexity of real-life applications demands constant improvements of microprocessor systems. One of the most frequently adopted microprocessor design scheme is the von Neumann architecture. Central Processing Unit (CPU performs computations and communicates with memory in a constant exchange of information. This unceasing motion of data between these two components became a significant performance bottleneck. A lot of power, energy, and computational time are wasted in this communication. With Beyond von Neumann Computing (BvNC paradigms, calculations are performed inside or very close to a memory array. BvNC approaches are proposed in the literature, mainly based on modifications of existing memories, enabling simple computations. Others exploit emerging technologies to both store and compute data, using analog operations. In this work we follow a different approach, where computational units are placed close to memory cells, improving versatility and performance. We propose a Hybrid-SIMD architecture made of memory and computing elements in an interleaved structure. Hybrid-SIMD can be used both as a low density memory and as SIMD accelerator. We insert our design in a classical von Neumann system based on a RISC-V processor, and we estimate its impact, demonstrating its capability to improve speed reducing at the same time energy consumption.

Conventional integrated circuits' design uses one layer to place logic gates and many additional layers to route interconnections. This design technique is built around the constraints of MOSFET transistors. To further improve the performance of integrated circuits, it is necessary to go beyond this limitation and to design true 3-D circuits. Although this possibility is difficult to implement with transistor technology, perpendicular nanomagnet logic (pNML) intrinsically enables the design of 3-D devices. It is very low-power consumption and offers the possibility to be integrated in the back end of traditional fabrication processes. These characteristics make pNML an ideal candidate to implement low-power coprocessors. In this paper, we demonstrate the possibilities offered by pNML technology by designing a 3-D coprocessor for the summed-area table, one of the most common algorithms used in image processing. We demonstrate the effectiveness of the design and the technology itself by comparing the performance with transistor implementations. The 3-D design makes it possible to obtain a small circuit footprint. Overall, the results presented here are a great step forward toward the design of 3-D coprocessors in pNML technology.

In the post-CMOS scenario, field coupled nanotechnologies represent an innovative and interesting new direction for electronic nanocomputing. Among these technologies, nanomagnet logic (NML) makes it possible to finally embed logic and memory in the same device. To fully analyze the potential of NML circuits, design tools that mimic the CMOS design-flow should be used for circuit design. We present, in this paper, the latest and improved version of Torino Politecnico Nanotechnology (ToPoliNano), our design and simulation framework for field coupled nanotechnologies. ToPoliNano emulates the top-down design process of CMOS technology. Circuits are described with a VHSIC hardware description language netlist and layout is then automatically generated considering in-plane NML (iNML) technology. The resulting circuits can be simulated and performance can be analyzed. In this paper, we describe several enhancements to the tool itself, like a circuit editor for custom design of field coupled nanodevices, improved algorithms for netlist optimization and new algorithms for the place and route of iNML circuits. We have validated and analyzed the tool by using extensive metrics, both by using standard circuits and ISCAS'85 benchmarks. This contribution highlights the improvements of ToPoliNano, which is now a innovative and complete tool for the development of iNML technology.

Nanomagnet logic (NML) is one of the most novel solutions studied as complementary technology to CMOS transistors. Information propagation involves only a change in spin orientation, no charge movement is present. Since the basic element is a nanomagnet, NML circuits have no stand-by power consumption and the ability to mix logic and memory in the same device. While CMOS is a multilayer technology, until now NML is confined to one single physical layer. The consequence is that circuit area grows exponentially due to interconnections overhead. In this paper, we present an innovative solution that drastically reduces the area wasted for interconnection wires relying on the properties of domain walls (DWs). We mix DWs and NML technologies in a unique DW logic (DWL) solution that exploits the advantages of both technologies. The proposed solution is technologically compatible with up-to-date fabrication processes. All the results here presented for the NML logic blocks and the DWs interconnections and their combination are obtained through rigorous micromagnetic simulations. Moreover, we implemented as a case study an high performance adder (Pentium 4 adder) and evaluated its features with increasing parallelism and compared with the simple NML implementation in order to explore the potential of DWL technology at circuit and architectural level. The reduction in circuit area corresponds to a notable reduction in both the latency and power consumption. The improvements in NML technology are shown by both the remarkable performance improvement and new possibilities offered by this novel solution.

NanoMagnet logic (NML) is among the emerging technologies that might replace CMOS in the next decades. According to its physical characteristics, to better exploit the potential of this technology-and of other similar ones-the use of parallel architectures with regular layout that avoid long interconnection signals is advised. Systolic arrays (SAs) are among these architectures, being composed of a grid of equal processing elements that are locally interconnected. However, they are usually implemented to execute only a small set of algorithms, and for this reason, throughout the years, they have not been an appealing solution for CMOS. To seriously analyze the potentials of NML, complex architectures must be conceived, and their physical implementation explored considering realistic technological constraints. With the increasing complexity of NML circuits, two issues, then, are noticed: 1) the need for a regular structure arises, that at the same time helps to reduce the intrinsic pipelining nature of NML and can be configured to be used for several applications without developing a dedicated design for each algorithm and 2) the capability to synthesize, place and route NML circuits is fundamental to demonstrate the feasibility of the architecture in two important conditions: efficiently managing the complexity of the design and sticking to the characteristics that are technologically feasible at the time of writing. In this paper, we address these issues presenting a new reconfigurable SA that can be programmed to execute different algorithms, and we provide two examples to show its working principle. Moreover, the array is synthesized and simulated with the aid of the first real tool for nanotechnology circuits that we have conceived, Torino Politecnico Nanotechnology tool. The joint contribution at both the architectural and physical design levels gives a relevant step forward to the state of the art in the demonstration of this emerging technology potential.

The current trend for intensive computational architectures is to adopt massive parallelism, with several concurrent tasks performed simultaneously, as done for example in GPUs. This approach has many advantages, such as the reduced design time given by circuit replication and an increasing in computational speed without the need of higher frequency. It has however evidenced an important bottleneck in data exchange between memory and processor. We envisage a revolutionary path for the future relation between memory and logic in parallel processors, where a new type of architecture exploits the principle of caching to the limit. Our Logic-in-Memory (LIM) architecture mixes logic and memory in the same device, removing the bottleneck of other existing parallel solutions. The architecture we propose, here in its preliminary version, has an array organization and each element in the array is based on three blocks: a logic unit for processing, a smart memory block and a routing structure for inter block communication. In this article we show the benefits of this approach with an application example in the image processing field. We can achieve a 4X computational time reduction for an image processing algorithm (Summed Area Table) with respect to the best architecture present in the literature, even with a preliminary and not optimized version. Besides the adoption of massive parallelism to increase performance, new technologies to open the post-CMOS era are explored. Among them NanoMagnet Logic (NML) is particularly interesting for its ability to mix logic and memory in the same device. We present here the preliminary results of the NML implementation of the LIM architecture. We thus demonstrate that it is not only a good solution for a standard CMOS technology but can also exploit the potential of an emerging technology as NML.

In past years the most common way to improve computers performance was to increase the clock frequency. In recent years this approach suffered the limits of technology scaling, therefore computers architectures are shifting toward the direction of parallel computing to further improve circuits performance. Not only GPU based architectures are spreading in consideration, but also Systolic Arrays are particularly suited for certain classes of algorithms. An important point in favor of Systolic Arrays is that, due to the regularity of their circuit layout, they are appealing when applied to many emerging and very promising technologies, like Quantum-dot Cellular Automata and nanoarrays based on Silicon NanoWire or on Carbon nanotube Field Effect Transistors. In this work we present a systematic method to improve Systolic Arrays performance exploiting Pipelining and Input Data Interleaving. We tackle the problem from a theoretical point of view first, and then we apply it to both CMOS technology and emerging technologies. On CMOS we demonstrate that it is possible to vastly improve the overall throughput of the circuit. By applying this technique to emerging technologies we show that it is possible to overcome some of their limitations greatly improving the throughput, making a considerable step forward toward the post-CMOS era.

Protein comparison is gaining importance year after year since it has been demonstrated that biologists can find correlation between different species, or genetic mutations that can lead to cancer and genetic diseases. Protein sequence alignment is the most computational intensive task when performing protein comparison. To speed-up alignment, dedicated processors that can perform different computations in parallel have been designed. Among them, the best performance has been achieved using systolic arrays (SAs). However, when the processing elements of the SA have an internal loop, performance could be highly reduced. In this paper, we present an architectural strategy to address this problem applying pipeline interleaving; this strategy is applied to an SA for Smith Waterman algorithm that we designed. Results encourage the adoption of pipeline interleaving for parallel circuits with loop-based processing elements. We demonstrate that important benefits in terms of higher operating frequency can be derived without so relevant costs as increased complexity, area, and power required.

In the post-CMOS scenario a primary role is played by the quantum-dot cellular automata (QCA) technology. Irrespective of the specific implementation principle (e.g., either molecular, or magnetic or semiconductive in the current scenario) the intrinsic deep-level pipelined behavior is the dominant issue. It has important consequences on circuit design and performance, especially in the presence of feedbacks in sequential circuits. Though partially already addressed in literature, these consequences still must be fully understood and solutions thoroughly approached to allow this technology any further advancement. This paper conducts an exhaustive analysis of the effects and the consequences derived by the presence of loops in QCA circuits. For each problem arisen, a solution is presented. The analysis is performed using as a test architecture, a complex systolic array circuit for biosequences analysis (Smith-Waterman algorithm), which represents one of the most promising application for QCA technology. The circuit is based on nanomagnetic logic as QCA implementation, is designed down to the layout level considering technological constraints and experimentally validated structures, counts up to approximately 2.3 milion nanomagnets, and is described and simulated with HDL language using as a testbench realistic protein alignment sequences. The results here presented constitute a fundamental advancement in the emerging technologies field since: 1) they are based on a quantitative approach relying on a realistic and complex circuit involving a large variety of QCA blocks; 2) they strictly are reckoned starting from current technological limits without relying on unrealistic assumptions; 3) they provide general rules to design complex sequential circuits with intrinsically pipelined technologies, like QCA; and 4) they prove with a real application benchmark how to maximize the circuits performance.

The interest in emerging nanotechnologies has been recently focused on nanomagnetic logic (NML), which has unique appealing features. NML circuits have very low power consumption and, because of their magnetic nature, maintain the information safely stored even without power supply. The nature of these circuits is much different from that of CMOS circuits. As a consequence, to better understand NML logic, complex circuits and not only simple gates must be designed. This constraint calls for a new design and simulation methodology. It should efficiently encompass manifold properties: 1) being based on commonly used hardware description language (HDL) in order to easily manage complexity and hierarchy; 2) maintaining a clear link with physical characteristics; and 3) modeling performance aspects such as speed and power, together with logic behavior. In this paper, we present a very-high-speed integrated circuits HDL (VHDL) behavioral model for NML circuits, which allows the evaluation of not only the logic behavior but also its power dissipation. It is based on a technological solution called “snake-clock.” We demonstrate this model using a case study which offers the right variety of internal substructures to test the method: a 4-bit microprocessor designed using asynchronous logic. The model enables a hierarchical bottom-up evaluation of the processor logic behavior, area, and power dissipation, which we evaluate using a benchmark division algorithm. The results highlight the flexibility and the efficiency of this model, as well as the remarkable improvements that it brings to the analysis of NML circuits.

In the years to come new solutions will be required to overcome the limitations of scaled CMOS technology. One approach is to adopt Nano-Magnetic Logic Circuits, highly appealing for their extremely reduced power consumption. Despite the interesting nature of this approach, many problems arise when this technology is considered for real designs. The wire is the most critical of these problems from the circuit implementation point of view. It works as a pipelined interconnection, and its delay in terms of clock cycles depends on its length. Serious complications arise at the design phase, both in terms of synthesis and of physical design. One possible solution is the use of a delay insensitive asynchronous logic, Null Convention Logic (NCL TM ). Nevertheless its use has many negative consequences in terms of area occupation and speed loss with respect to a Boolean version. In this article we analyze and compare different solutions: nanomagnetic circuits based on full NCL, mixed Boolean-NCL, and fully Boolean logic. We discuss the advantages of these logics, but also the issues they raise. In particular we analyze feedback signals, which, due to their intrinsic pipelined nature, cause errors that still have not found a solution in the literature. The innovative arrangement we propose solves most of the problems and thus soundly increases the knowledge of this technology. The analysis is performed using a VHDL behavioral model we developed and a microprocessor we designed based on this model, as a sound and realistic test bench.