Related papers: Address Translation Design Tradeoffs for Heterogen…
Artificial intelligence (AI) application domains consist of a mix of tensor operations with high and low arithmetic intensities (aka reuse). Hierarchical (i.e. compute along multiple levels of memory hierarchy) and heterogeneous (multiple…
To deliver high performance in power limited systems, architects have turned to using heterogeneous systems, either CPU+GPU or mixed CPU-hardware systems. However, in systems with different processor types and task affinities, scheduling…
Hardware specialization is becoming a key enabler of energyefficient performance. Future systems will be increasingly heterogeneous, integrating multiple specialized and programmable accelerators, each with different memory demands.…
A large language model (LLM) is one of the most important emerging machine learning applications nowadays. However, due to its huge model size and runtime increase of the memory footprint, LLM inferences suffer from the lack of memory…
Edge computing processes data where it is generated, enabling faster decisions, lower bandwidth usage, and improved privacy. However, edge devices typically operate under strict constraints on processing power, memory, and energy…
Recent studies have extensively explored NPU architectures for accelerating AI inference in on-device environments, which are inherently resource-constrained. Meanwhile, transformer-based large language models (LLMs) have become dominant,…
Transformer-based large language models (LLMs) have demonstrated exceptional capabilities in sequence modeling and text generation, with improvements scaling proportionally with model size. However, the limitations of GPU memory have…
On-chip communication infrastructure is a central component of modern systems-on-chip (SoCs), and it continues to gain importance as the number of cores, the heterogeneity of components, and the on-chip and off-chip bandwidth continue to…
The exponential growth of data-intensive machine learning workloads has exposed significant limitations in conventional GPU-accelerated systems, especially when processing datasets exceeding GPU DRAM capacity. We propose MQMS, an augmented…
Heterogeneous multi-core architectures combine a few "host" cores, optimized for single-thread performance, with many small energy-efficient "accelerator" cores for data-parallel processing, on a single chip. Offloading a computation to the…
Hardware accelerators, such as those based on GPUs and FPGAs, offer an excellent opportunity to efficiently parallelize functionalities. Recently, modern embedded platforms started being equipped with such accelerators, resulting in a…
Large inter-GPU all-reduce operations, prevalent throughout deep learning, are bottlenecked by communication costs. Emerging heterogeneous architectures are comprised of complex nodes, often containing $4$ GPUs and dozens to hundreds of CPU…
Computational workloads composing traditional transformer models are starkly bifurcated. Multi-Head Attention (MHA) and Grouped-Query Attention are memory-bound due to low arithmetic intensity, while FeedForward Networks are compute-bound.…
We present a new least-privilege-based model of addressing on which to base memory management functionality in an OS for modern computers like phones or server-based accelerators. Existing software assumptions do not account for…
As custom hardware accelerators become more prevalent, it becomes increasingly important to automatically generate efficient host-driver code that can fully leverage the capabilities of these accelerators. This approach saves time and…
Accelerator-based heterogeneous architectures, such as CPU-GPU, CPU-TPU, and CPU-FPGA systems, are widely adopted to support the popular artificial intelligence (AI) algorithms that demand intensive computation. When deployed in real-time…
The emergence of new, off-path smart network cards (SmartNICs), known generally as Data Processing Units (DPU), has opened a wide range of research opportunities. Of particular interest is the use of these and related devices in tandem with…
Memory latency, bandwidth, capacity, and energy increasingly limit performance. In this paper, we reconsider proposed system architectures that consist of huge (many-terabyte to petabyte scale) memories shared among large numbers of CPUs.…
Neural architectures and hardware accelerators have been two driving forces for the progress in deep learning. Previous works typically attempt to optimize hardware given a fixed model architecture or model architecture given fixed…
The hardware/software boundary in modern heterogeneous multicore computers is increasingly complex, and diverse across different platforms. A single memory access by a core or DMA engine traverses multiple hardware translation and caching…