Mixture-of-Experts (MoE) model architecture has emerged as a promising solution for scaling transformer models efficiently, offering sparse activation that reduces computational costs while increasing model capacity. However, as MoE models scale, they need to be distributed across GPU devices, thus face critical performance bottlenecks due to their large memory footprint. Expert parallelism distributes experts across GPUs, however, faces key challenges including an unbalanced token routing and expert activation, resulting in communication tail latency and processing inefficiencies. While existing solutions address some of these issues, they fail to resolve the dual challenges of load imbalance and communication skew. The imbalance in token processing load across experts causes uneven processing times on different GPUs, while communication skew between GPUs leads to unbalanced inter-GPU data transfers. These factors degrade the performance of MoE models by increasing tail latency and reducing overall throughput. To address these limitations, we propose an Integer Linear Programming (ILP) formulation to optimize expert placement by jointly considering token load, communication, and computation costs. We exploit the property that there is a token routing dependency across layers, where tokens routed to a specific expert in one layer are likely to be routed to a limited set of experts in the subsequent layer. Our solution, MoETuner, offers an optimal expert-to-GPU assignment that minimizes inter-GPU token routing costs and balances token processing across devices, thereby reducing tail latency and end-to-end execution time. Experimental results demonstrate 9.3% and 17.5% of end-to-end speedups for single-node and multi-node inference respectively, showcasing the potential of our ILP-based optimization for offering expert parallel solutions for next-generation MoEs.

Distributed execution of deep learning training involves a dynamic interplay between hardware accelerator architecture and device placement strategy. This is the first work to explore the co-optimization of determining the optimal architecture and device placement strategy through novel algorithms, improving the balance of computational resources, memory usage, and data distribution. Our architecture search leverages tensor and vector units, determining their quantity and dimensionality, and on-chip and off-chip memory configurations. It also determines the microbatch size and decides whether to recompute or stash activations, balancing the memory footprint of training and storage size. For each explored architecture configuration, we use an Integer Linear Program (ILP) to find the optimal schedule for executing operators on the accelerator. The ILP results then integrate with a dynamic programming solution to identify the most effective device placement strategy, combining data, pipeline, and tensor model parallelism across multiple accelerators. Our approach achieves higher throughput on large language models compared to the state-of-the-art TPUv4 and the Spotlight accelerator search framework. The entire source code of PHAZE is available at https://github.com/msr-fiddle/phaze.

Recommendation models rely on deep learning networks and large embedding tables, resulting in computationally and memory-intensive processes. These models are typically trained using hybrid CPU-GPU or GPU-only configurations. The hybrid mode combines the GPU's neural network acceleration with the CPUs' memory storage and supply for embedding tables but may incur significant CPU-to-GPU transfer time. In contrast, the GPU-only mode utilizes High Bandwidth Memory (HBM) across multiple GPUs for storing embedding tables. However, this approach is expensive and presents scaling concerns. This paper introduces Hotline, a heterogeneous acceleration pipeline that addresses these concerns. Hotline develops a data-aware and model-aware scheduling pipeline by leveraging the insight that only a few embedding entries are frequently accessed (popular). This approach utilizes CPU main memory for non-popular embeddings and GPUs' HBM for popular embeddings. To achieve this, Hotline accelerator fragments a mini-batch into popular and non-popular micro-batches. It gathers the necessary working parameters for non-popular micro-batches from the CPU, while GPUs execute popular micro-batches. The hardware accelerator dynamically coordinates the execution of popular embeddings on GPUs and non-popular embeddings from the CPU's main memory. Real-world datasets and models confirm Hotline's effectiveness, reducing average end-to-end training time by 2.2x compared to Intel-optimized CPU-GPU DLRM baseline.

Training recommendation models pose significant challenges regarding resource utilization and performance. Prior research has proposed an approach that categorizes embeddings into popular and non-popular classes to reduce the training time for recommendation models. We observe that, even among the popular embeddings, certain embeddings undergo rapid training and exhibit minimal subsequent variation, resulting in saturation. Consequently, updates to these embeddings lack any contribution to model quality. This paper presents Slipstream, a software framework that identifies stale embeddings on the fly and skips their updates to enhance performance. This capability enables Slipstream to achieve substantial speedup, optimize CPU-GPU bandwidth usage, and eliminate unnecessary memory access. SlipStream showcases training time reductions of 2x, 2.4x, 1.2x, and 1.175x across real-world datasets and configurations, compared to Baseline XDL, Intel-optimized DRLM, FAE, and Hotline, respectively.

Learned indexes use machine learning models to learn the mappings between keys and their corresponding positions in key-value indexes. These indexes use the mapping information as training data. Learned indexes require frequent retrainings of their models to incorporate the changes introduced by update queries. To efficiently retrain the models, existing learned index systems often harness a linear algebraic QR factorization technique that performs matrix decomposition. This factorization approach processes all key-position pairs during each retraining, resulting in compute operations that grow linearly with the total number of keys and their lengths. Consequently, the retrainings create a severe performance bottleneck, especially for variable-length string keys, while the retrainings are crucial for maintaining high prediction accuracy and in turn, ensuring low query service latency. To address this performance problem, we develop an algorithm-hardware co-designed string-key learned index system, dubbed SIA. In designing SIA, we leverage a unique algorithmic property of the matrix decomposition-based training method. Exploiting the property, we develop a memoization-based incremental training scheme, which only requires computation over updated keys, while decomposition results of non-updated keys from previous computations can be reused. We further enhance SIA to offload a portion of this training process to an FPGA accelerator to not only relieve CPU resources for serving index queries (i.e., inference), but also accelerate the training itself. Our evaluation shows that compared to ALEX, LIPP, and SIndex, a state-of-the-art learned index systems, SIA-accelerated learned indexes offer 2.6x and 3.4x higher throughput on the two real-world benchmark suites, YCSB and Twitter cache trace, respectively.

Federated Learning (FL) allows machine learning models to train locally on individual mobile devices, synchronizing model updates via a shared server. This approach safeguards user privacy; however, it also generates a heterogeneous training environment due to the varying performance capabilities across devices. As a result, straggler devices with lower performance often dictate the overall training time in FL. In this work, we aim to alleviate this performance bottleneck due to stragglers by dynamically balancing the training load across the system. We introduce Invariant Dropout, a method that extracts a sub-model based on the weight update threshold, thereby minimizing potential impacts on accuracy. Building on this dropout technique, we develop an adaptive training framework, Federated Learning using Invariant Dropout (FLuID). FLuID offers a lightweight sub-model extraction to regulate computational intensity, thereby reducing the load on straggler devices without affecting model quality. Our method leverages neuron updates from non-straggler devices to construct a tailored sub-model for each straggler based on client performance profiling. Furthermore, FLuID can dynamically adapt to changes in stragglers as runtime conditions shift. We evaluate FLuID using five real-world mobile clients. The evaluations show that Invariant Dropout maintains baseline model efficiency while alleviating the performance bottleneck of stragglers through a dynamic, runtime approach.

Recommendation models are commonly used learning models that suggest relevant items to a user for e-commerce and online advertisement-based applications. Current recommendation models include deep-learning-based (DLRM) and time-based sequence (TBSM) models. These models use massive embedding tables to store a numerical representation of item's and user's categorical variables (memory-bound) while also using neural networks to generate outputs (compute-bound). Due to these conflicting compute and memory requirements, the training process for recommendation models is divided across CPU and GPU for embedding and neural network executions, respectively. Such a training process naively assigns the same level of importance to each embedding entry. This paper observes that some training inputs and their accesses into the embedding tables are heavily skewed with certain entries being accessed up to 10000x more. This paper tries to leverage skewed embedded table accesses to efficiently use the GPU resources during training. To this end, this paper proposes a Frequently Accessed Embeddings (FAE) framework that exposes a dynamic knob to the software based on the GPU memory capacity and the input popularity index. This framework efficiently estimates and varies the size of the hot portions of the embedding tables within GPUs and reallocates the rest of the embeddings on the CPU. Overall, our framework speeds-up the training of the recommendation models on Kaggle, Terabyte, and Alibaba datasets by 2.34x as compared to a baseline that uses Intel-Xeon CPUs and Nvidia Tesla-V100 GPUs, while maintaining accuracy.

Modern machine learning workloads use large models, with complex structures, that are very expensive to execute. The devices that execute complex models are becoming increasingly heterogeneous as we see a flourishing of domain-specific accelerators being offered as hardware accelerators in addition to CPUs. These trends necessitate distributing the workload across multiple devices. Recent work has shown that significant gains can be obtained with model parallelism, i.e, partitioning a neural network's computational graph onto multiple devices. In particular, this form of parallelism assumes a pipeline of devices, which is fed a stream of samples and yields high throughput for training and inference of DNNs. However, for such settings (large models and multiple heterogeneous devices), we require automated algorithms and toolchains that can partition the ML workload across devices. In this paper, we identify and isolate the structured optimization problem at the core of device placement of DNN operators, for both inference and training, especially in modern pipelined settings. We then provide algorithms that solve this problem to optimality. We demonstrate the applicability and efficiency of our approaches using several contemporary DNN computation graphs.