Machine learning solutions are rapidly adopted to enable a variety of key use cases, from conversational AI assistants to scientific discovery. As the adoption of machine learning models becomes increasingly prevalent, the associated lifecycle carbon footprint is expected to increase, including both *operational carbon* from training and inference and *embodied carbon* from AI hardware manufacturing. We introduce CATransformers, the first carbon-aware co-optimization framework for Transformer-based models and hardware accelerators. By integrating both operational and embodied carbon into early-stage design space exploration, CATransformers enables sustainability-driven model architecture and hardware accelerator co-design that reveals fundamentally different trade-offs than latency- or energy-centric approaches. Evaluated across a range of Transformer models, CATransformers consistently demonstrates the potential to reduce total carbon emissions --by up to 30\% -- while maintaining accuracy and latency. We further highlight its extensibility through a focused case study on multi-modal models. Our results emphasize the need for holistic optimization methods that prioritize carbon efficiency without compromising model capability and execution time performance. The source code of CATransformers is available at https://github.com/facebookresearch/CATransformers.

Training and deploying large-scale machine learning models is time-consuming, requires significant distributed computing infrastructures, and incurs high operational costs. Our analysis, grounded in real-world large model training on datacenter-scale infrastructures, reveals that 14~32% of all GPU hours are spent on communication with no overlapping computation. To minimize this outstanding communication latency and other inherent at-scale inefficiencies, we introduce an agile performance modeling framework, MAD-Max. This framework is designed to optimize parallelization strategies and facilitate hardware-software co-design opportunities. Through the application of MAD-Max to a suite of real-world large-scale ML models on state-of-the-art GPU clusters, we showcase potential throughput enhancements of up to 2.24 × for pretraining and up to 5.27 × for inference scenarios, respectively.

As the development of large-scale Generative AI models evolve beyond text (1D) generation to include image (2D) and video (3D) generation, processing spatial and temporal information presents unique challenges to quality, performance, and efficiency. We present the first work towards understanding this new system design space for multi-modal text-to-image (TTI) and text-to-video (TTV) generation models. Current model architecture designs are bifurcated into 2 categories: Diffusion-and Transformer-based models. Our systematic performance characterization on a suite of eight representative TTI/TTV models shows that after state-of-the-art optimization techniques such as Flash Attention are applied, Convolution accounts for up to 44% of execution time for Diffusion-based TTI models, while Linear layers consume up to 49 % of execution time for Transformer-based models. We additionally observe that Diffusion-based TTI models resemble the Prefill stage of LLM inference, and benefit from 1.1-2.5x greater speedup from Flash Attention than Transformer-based TTI models that resemble the Decode phase. Since optimizations designed for LLMs do not map directly onto TTI/TTV models, we must conduct a thorough characterization of these workloads to gain insights for new optimization opportunities. In doing so, we define sequence length in the context of TTI/TTV models and observe sequence length can vary up to 4x in Diffusion model inference. We additionally observe temporal aspects of TTV workloads pose unique system bottlenecks, with Temporal Attention accounting for over 60 % of total Attention time. Overall, our in-depth system performance characterization is a critical first step towards designing efficient and deployable systems for emerging TTI/TTV workloads.

Deep learning recommendation systems serve personalized content under diverse tail-latency targets and input-query loads. In order to do so, state-of-the-art recommendation models rely on terabyte-scale embedding tables to learn user preferences over large bodies of contents. The reliance on a fixed embedding representation of embedding tables not only imposes significant memory capacity and bandwidth requirements but also limits the scope of compatible system solutions. This paper challenges the assumption of fixed embedding representations by showing how synergies between embedding representations and hardware platforms can lead to improvements in both algorithmic- and system performance. Based on our characterization of various embedding representations, we propose a hybrid embedding representation that achieves higher quality embeddings at the cost of increased memory and compute requirements. To address the system performance challenges of the hybrid representation, we propose MP-Rec — a co-design technique that exploits heterogeneity and dynamic selection of embedding representations and underlying hardware platforms.

Large-scale Graph Convolutional Network (GCN) inference on traditional CPU/GPU systems is challenging due to a large memory footprint, sparse computational patterns, and irregular memory accesses with poor locality. Intel’s Programmable Integrated Unffied Memory Architecture (PIUMA) is designed to address these challenges for graph analytics. In this paper, a detailed characterization of GCNs is presented using the Open-Graph Benchmark (OGB) datasets to determine the viability of PIUMA as a potential solution to GCN scalability. First, the extent of sparse matrix dense matrix multiplication (SpMM) as a performance driver for GCN on CPU and GPU is explored, offering a methodology for predicting GCN behavior as a function of dataset characteristics. Second, an SpMM kernel optimized for PIUMA is described and investigated for sensitivity to system parameters including memory bandwidth, latency, and thread count. SpMM scalability on PIUMA is demonstrated, while the scalability limitations of a Xeon-optimized SpMM implementation are discussed. Finally, GCN performance is compared on PIUMA versus a Xeon CPU system and Ampere GPU system, showing impressive results on PIUMA for largescale datasets.

Neural personalized recommendation models are used across a wide variety of datacenter applications including search, social media, and entertainment. State-of-the-art models comprise large embedding tables that have billions of parameters requiring large memory capacities. Unfortunately, large and fast DRAM-based memories levy high infrastructure costs. Conventional SSD-based storage solutions offer an order of magnitude larger capacity, but have worse read latency and bandwidth, degrading inference performance. RecSSD is a near data processing based SSD memory system customized for neural recommendation inference that reduces end-to-end model inference latency by 2× compared to using COTS SSDs across eight industry-representative models.

Deep learning recommendation systems must provide high quality, personalized content under strict tail-latency targets and high system loads. This paper presents RecPipe, a system to jointly optimize recommendation quality and inference performance. Central to RecPipe is decomposing recommendation models into multi-stage pipelines to maintain quality while reducing compute complexity and exposing distinct parallelism opportunities. RecPipe implements an inference scheduler to map multi-stage recommendation engines onto commodity, heterogeneous platforms (e.g., CPUs, GPUs). While the hardware-aware scheduling improves ranking efficiency, the commodity platforms suffer from many limitations requiring specialized hardware. Thus, we design RecPipeAccel (RPAccel), a custom accelerator that jointly optimizes quality, tail-latency, and system throughput. RPAccel is designed specifically to exploit the distinct design space opened via RecPipe. In particular, RPAccel processes queries in sub-batches to pipeline recommendation stages, implements dual static and dynamic embedding caches, a set of top-k filtering units, and a reconfigurable systolic array. Compared to previously proposed specialized recommendation accelerators and at iso-quality, we demonstrate that RPAccel improves latency and throughput by 3 × and 6 ×.

Neural personalized recommendation is the cornerstone of a wide collection of cloud services and products, constituting significant compute demand of cloud infrastructure. Thus, improving the execution efficiency of recommendation directly translates into infrastructure capacity saving. In this paper, we propose DeepRecSched, a recommendation inference scheduler that maximizes latency-bounded throughput by taking into account characteristics of inference query size and arrival patterns, model architectures, and underlying hardware systems. By carefully optimizing task versus data-level parallelism, DeepRecSched improves system throughput on server class CPUs by 2× across eight industry-representative models. Next, we deploy and evaluate this optimization in an at-scale production datacenter which reduces end-to-end tail latency across a wide variety of recommendation models by 30%. Finally, DeepRecSched demonstrates the role and impact of specialized AI hardware in optimizing system level performance (QPS) and power efficiency (QPS/watt) of recommendation inference. In order to enable the design space exploration of customized recommendation systems shown in this paper, we design and validate an end-to-end modeling infrastructure, DeepRecInfra. DeepRecInfra enables studies over a variety of recommendation use cases, taking into account at-scale effects, such as query arrival patterns and recommendation query sizes, observed from a production datacenter, as well as industry-representative models and tail latency targets.