Related papers: Milliscale: Fast Commit on Low-Latency Object Stor…
Attention-based large language models (LLMs) have transformed modern AI applications, but the quadratic cost of self-attention imposes significant compute and memory overhead. Dynamic sparsity (DS) attention mitigates this, yet its hardware…
Existing methods for training LLMs on long-sequence data, such as Tensor Parallelism and Context Parallelism, exhibit low Model FLOPs Utilization as sequence lengths and number of GPUs increase, especially when sequence lengths exceed 1M…
Split learning (SL) enables collaborative training of large language models (LLMs) between resource-constrained edge devices and compute-rich servers by partitioning model computation across the network boundary. However, existing SL…
One-stage object detectors such as SSD or YOLO already have shown promising accuracy with small memory footprint and fast speed. However, it is widely recognized that one-stage detectors have difficulty in detecting small objects while they…
Cloud object storage such as AWS S3 is cost-effective and highly elastic but relatively slow, while high-performance cloud storage such as AWS ElastiCache is expensive and provides limited elasticity. We present a new cloud storage service…
The Knights Landing (KNL) is the codename for the latest generation of Intel processors based on Intel Many Integrated Core (MIC) architecture. It relies on massive thread and data parallelism, and fast on-chip memory. This processor…
Superpixel segmentation can be used as an intermediary step in many applications, often to improve object delineation and reduce computer workload. However, classical methods do not incorporate information about the desired object.…
Efficient neural network backbones for mobile devices are often optimized for metrics such as FLOPs or parameter count. However, these metrics may not correlate well with latency of the network when deployed on a mobile device. Therefore,…
Computation-Enabled Object Storage (COS) systems, such as MinIO and Ceph, have recently emerged as promising storage solutions for post hoc, SQL-based analysis on large-scale datasets in High-Performance Computing (HPC) environments. By…
Persistent memory provides high-performance data persistence at main memory. Memory writes need to be performed in strict order to satisfy storage consistency requirements and enable correct recovery from system crashes. Unfortunately,…
Diffusion Transformers, particularly for video generation, achieve remarkable quality but suffer from quadratic attention complexity, leading to prohibitive latency. Existing acceleration methods face a fundamental trade-off: dynamically…
Many managed key-value and NoSQL databases - such as Amazon DynamoDB, Azure Cosmos DB, and Google Cloud Firestore - enforce strict maximum item sizes (e.g., 400 KB in DynamoDB). This constraint imposes significant architectural challenges…
Transformers are widely used for their ability to capture data relations in sequence processing, with great success for a wide range of static tasks. However, the computational and memory footprint of their main component, i.e., the Scaled…
High-quality articulated 3D assets are indispensable for embodied AI and physical simulation, yet 3D generation still focuses on static meshes, leaving a gap in "sim-ready" interactive objects. Most recent articulated object creation…
Pyramidal networks are standard methods for multi-scale object detection. Current researches on feature pyramid networks usually adopt layer connections to collect features from certain levels of the feature hierarchy, and do not consider…
Cloud-based outsourced Location-based services have profound impacts on various aspects of people's lives but bring security concerns. Existing spatio-temporal data secure retrieval schemes have significant shortcomings regarding dynamic…
Objective Caml is a famous dialect of the ML family languages. It is well-known for its performance as a compiled programming language, notably thanks to its incremental generational automatic memory collection. However, for historical…
Many ML applications and products train on medium amounts of input data but get bottlenecked in real-time inference. When implementing ML systems, conventional wisdom favors segregating ML code into services queried by product code via…
Efficient long-context understanding and reasoning are increasingly vital for large language model (LLM) applications such as multi-turn dialogue and program analysis. However, the core self-attention mechanism scales quadratically with…
Erasure codes have been widely considered a promising solution to enhance data reliability at low storage costs. However, in modern geo-distributed storage systems, erasure codes may incur high data access latency as they require data…