Search for: All records

As AI continues to grow, modern applications are becoming more data- and compute-intensive, driving the development of specialized AI chips to meet these demands. One example is AMD's AI Engine (AIE), a dedicated hardware system that includes a 2D array of high-frequency very-long instruction words (VLIW) vector processors to provide high computational throughput and reconfigurability. However, AIE's specialized architecture presents tremendous challenges in programming and compiler optimization. Existing AIE programming frameworks lack a clean abstraction to represent multi-level parallelism in AIE; programmers have to figure out the parallelism within a kernel, manually do the partition, and assign sub-tasks to different AIE cores to exploit parallelism. These significantly lower the programming productivity. Furthermore, some AIE architectures include FPGAs to provide extra flexibility, but there is no unified intermediate representation (IR) that captures these architectural differences. As a result, existing compilers can only optimize the AIE portions of the code, overlooking potential FPGA bottlenecks and leading to suboptimal performance. To address these limitations, we introduce ARIES, an agile multi-level intermediate representation (MLIR) based compilation flow for reconfigurable devices with AIEs. ARIES introduces a novel programming model that allows users to map kernels to separate AIE cores, exploiting task- and tile-level parallelism without restructuring code. It also includes a declarative scheduling interface to explore instruction-level parallelism within each core. At the IR level, we propose a unified MLIR-based representation for AIE architectures, both with or without FPGA, facilitating holistic optimization and better portability across AIE device families. For the General Matrix Multiply (GEMM) benchmark, ARIES achieves 4.92 TFLOPS, 15.86 TOPS, and 45.94 TOPS throughput under FP32, INT16, and, INT8 data types on Versal VCK190 respectively. Compared with the state-of-the-art (SOTA) work CHARM for AIE, ARIES improves the throughput by 1.17x, 1.59x, and 1.47x correspondingly. For ResNet residual layer, ARIES achieves up to 22.58x speedup compared with optimized SOTA work Riallto on Ryzen-AI NPU. ARIES is open-sourced on GitHub: https://github.com/arc-research-lab/Aries. 

Data centers have been relying on renewable energy integration coupled with energy efficient specialized processing units and accelerators to increase sustainability. Unfortunately, the carbon generated from manufacturing these systems is be- coming increasingly relevant due to these energy decarbonization and efficiency improvements. Furthermore, it is less clear how to mitigate this aspect of embodied carbon. As workloads continue to evolve over each hardware generation we explore the tradeoffs of fabricating new application-tuned hardware compared with more general solutions such as Field Programmable Gate Arrays (FPGAs). We also explore how REFRESH FPGAs can amortize embodied carbon investments from previous generations to meet the requirements of future generations workloads. 

Smart phones have revolutionized the availability of computing to the consumer. Recently, smart phones have been aggressively integrating artificial intelligence (AI) capabilities into their devices. The custom designed processors for the latest phones integrate incredibly capable and energy efficient graphics processors (GPUs) and tensor processors (TPUs) to accommodate this emerging AI workload and on-device inference. Unfor- tunately, smart phones are far from sustainable and have a substantial carbon footprint that continues to be dominated by environmental impacts from their manufacture and far less so by the energy required to power their operation. In this paper we explore the possibility of reversing the trend to increase the dedicated silicon dedicated to emerging application workloads in the phone. Instead we consider how in-memory processing using the DRAM already present in the phone could be used in place of dedicated GPU/TPU devices for AI inference. We explore the potential savings in embodied carbon that could be possible with this tradeoff and provide some analysis of the potential of in- memory computing to compete with these accelerators. While it may not be possible to achieve the same throughput, we suggest that the responsiveness to the user may be sufficient using in- memory computing, while both the embodied and operational carbon footprints could be improved. Our approach can save circa 10–15kgCO2e. 

While Vision Transformers (ViTs) have shown consistent progress in computer vision, deploying them for real-time decision-making scenarios (< 1 ms) is challenging. Current computing platforms like CPUs, GPUs, or FPGA-based solutions struggle to meet this deterministic low-latency real-time requirement, even with quantized ViT models. Some approaches use pruning or sparsity to reduce model size and latency, but this often results in accuracy loss. To address the aforementioned constraints, in this work, we propose EQ-ViT, an end-to-end acceleration framework with novel algorithm and architecture co-design features to enable real-time ViT acceleration on AMD Versal Adaptive Compute Acceleration Platform (ACAP). The contributions are four-fold. First, we perform in-depth kernel- level performance profiling & analysis and explain the bottlenecks for existing acceleration solutions on GPU, FPGA, and ACAP. Second, on the hardware level, we introduce a new spatial and heterogeneous accelerator architecture, EQ-ViT architec- ture. This architecture leverages the heterogeneous features of ACAP, where both FPGA and artificial intelligence engines (AIEs) coexist on the same system-on-chip (SoC). Third, On the algorithm level, we create a comprehensive quantization-aware training strategy, EQ-ViT algorithm. This strategy concurrently quantizes both weights and activations into 8-bit integers, aiming to improve accuracy rather than compromise it during quanti- zation. Notably, the method also quantizes nonlinear functions for efficient hardware implementation. Fourth, we design EQ- ViT automation framework to implement the EQ-ViT architec- ture for four different ViT applications on the AMD Versal ACAP VCK190 board, achieving accuracy improvement with 2.4%, and average speedups of 315.0x, 3.39x, 3.38x, 14.92x, 59.5x, 13.1x over computing solutions of Intel Xeon 8375C vCPU, Nvidia A10G, A100, Jetson AGX Orin GPUs, and AMD ZCU102, U250 FPGAs. The energy efficiency gains are 62.2x, 15.33x, 12.82x, 13.31x, 13.5x, 21.9x. 

Embodied carbon has been widely reported as a significant component in the full system lifecycle of various computing systems green house gas emissions. Many efforts have been undertaken to quantify the elements that comprise this embodied carbon, from tools that evaluate semiconductor manufacturing to those that can quantify different elements of the computing system from commercial and academic sources. However, these tools cannot easily reproduce results reported by server vendors' product carbon reports and the accuracy can vary substantially due to various assumptions. Furthermore, attempts to determine green house gas contributions using bottom-up methodologies often do not agree with system-level studies and are hard to rectify. Nonetheless, given there is a need to consider all contributions to green house gas emissions in datacenters, we propose SCARIF, the Server Carbon including Accelerator Reporter with Intelligence-based Formulation tool. SCARIF has three main contributions: (1) We first collect reported carbon cost data from server vendors and design statistic models to predict the embodied carbon cost so that users can get the embodied carbon cost for their server configurations. (2) We provide embodied carbon cost if users configure servers with accelerators including GPUs, and FPGAs. (3) By using case studies, we show that certain design choices of data center management might flip by the insight and observation from using SCARIF. Thus, SCARIF provides an opportunity for large-scale datacenter and hyperscaler design. We release SCARIF as an open-source tool at https://github.com/arc-research-lab/SCARIF. 

Dense matrix multiply (MM) serves as one of the most heavily used kernels in deep learning applications. To cope with the high computation demands of these applications, heterogeneous architectures featuring both FPGA and dedicated ASIC accelerators have emerged as promising platforms. For example, the AMD/Xilinx Versal ACAP architecture combines general-purpose CPU cores and programmable logic with AI Engine processors optimized for AI/ML. An array of 400 AI Engine processors executing at 1 GHz can provide up to 6.4 TFLOPS performance for 32-bit floating-point (FP32) data. However, machine learning models often contain both large and small MM operations. While large MM operations can be parallelized efficiently across many cores, small MM operations typically cannot. We observe that executing some small MM layers from the BERT natural language processing model on a large, monolithic MM accelerator in Versal ACAP achieved less than 5% of the theoretical peak performance. Therefore, one key question arises:How can we design accelerators to fully use the abundant computation resources under limited communication bandwidth for end-to-end applications with multiple MM layers of diverse sizes? We identify the biggest system throughput bottleneck resulting from the mismatch between massive computation resources of one monolithic accelerator and the various MM layers of small sizes in the application. To resolve this problem, we propose the CHARM framework to composemultiple diverse MM accelerator architecturesworking concurrently on different layers within one application. CHARM includes analytical models which guide design space exploration to determine accelerator partitions and layer scheduling. To facilitate system designs, CHARM automatically generates code, enabling thorough onboard design verification. We deploy the CHARM framework on four different deep learning applications in FP32, INT16, and INT8 data types, including BERT, ViT, NCF, and MLP, on the AMD/Xilinx Versal ACAP VCK190 evaluation board. Our experiments show that we achieve 1.46 TFLOPS, 1.61 TFLOPS, 1.74 TFLOPS, and 2.94 TFLOPS inference throughput for BERT, ViT, NCF, and MLP in FP32 data type, respectively, which obtain 5.29\(\times\), 32.51\(\times\), 1.00\(\times\), and 1.00\(\times\)throughput gains compared to one monolithic accelerator. CHARM achieves the maximum throughput of 1.91 TOPS, 1.18 TOPS, 4.06 TOPS, and 5.81 TOPS in the INT16 data type for the four applications. The maximum throughput achieved by CHARM in the INT8 data type is 3.65 TOPS, 1.28 TOPS, 10.19 TOPS, and 21.58 TOPS, respectively. We have open-sourced our tools, including detailed step-by-step guides to reproduce all the results presented in this paper and to enable other users to learn and leverage CHARM framework and tools in their end-to-end systems:https://github.com/arc-research-lab/CHARM. 

Fast-evolving artificial intelligence (AI) algorithms such as large language models have been driving the ever increasing computing demands in today’s data centers. Heterogeneous computing with domain-specific architectures (DSAs) brings many opportunities when scaling up and scaling out the computing system. In particular, heterogeneous chiplet architecture is favored to keep scaling up and scaling out the system as well as to reduce the design complexity and the cost stemming from the traditional monolithic chip design. However, how to interconnect computing resources and orchestrate heterogeneous chiplets is the key to success. In this paper, we first discuss the diversity and evolving demands of different AI workloads. We discuss how chiplet brings better cost efficiency and shorter time to market. Then we discuss the challenges in establishing chiplet interface standards, packaging, and security issues. We further discuss the software programming challenges in chiplet systems. 

With the increase in the computation intensity of the chip, the mismatch between computation layer shapes and the available computation resource significantly limits the utilization of the chip. Driven by this observation, prior works discuss spatial accelerators or dataflow architecture to maximize the throughput. However, using spatial accelerators could potentially increase the execution latency. In this work, we first systematically investigate two execution models: (1) sequentially (temporally) launch one monolithic accelerator, and (2) spatially launch multiple accelerators. From the observations, we find that there is a latency throughput tradeoff between these two execution models, and combining these two strategies together can give us a more efficient latency throughput Pareto front. To achieve this, we propose spatial sequential architecture (SSR) and SSR design automation framework to explore both strategies together when deploying deep learning inference. We use the 7nm AMD Versal ACAP VCK190 board to implement SSR accelerators for four end-to-end transformer-based deep learning models. SSR achieves average throughput gains of 2.53x, 35.71x, and 14.20x under different batch sizes compared to the 8nm Nvidia GPU A10G, 16nm AMD FPGAs ZCU102, and U250. The average energy efficiency gains are 8.51x, 6.75x, and 21.22x, respectively. Compared with the sequential-only solution and spatial-only solution on VCK190, our spatial-sequential-hybrid solutions achieve higher throughput under the same latency requirement and lower latency under the same throughput requirement. We also use SSR analytical models to demonstrate how to use SSR to optimize solutions on other computing platforms, e.g., 14nm Intel Stratix 10 NX.