Related papers: DOLMA: A Data Object Level Memory Disaggregation F…
In-situ LLM inference on end-user devices has gained significant interest due to its privacy benefits and reduced dependency on external infrastructure. However, as the decoding process is memory-bandwidth-bound, the diverse processing…
Much sequential data exhibits highly non-uniform information distribution. This cannot be correctly modeled by traditional Long Short-Term Memory (LSTM). To address that, recent works have extended LSTM by adding more activations between…
Disaggregating resources in data centers is an emerging trend. Recent work has begun to explore memory disaggregation, but suffers limitations including lack of consideration of the complexity of cloud-based deployment, including…
Modern LLM serving is no longer homogeneous or monolithic. Production systems now combine disaggregated execution, complex parallelism, runtime optimizations, and stateful workloads such as reasoning, agents, and RL rollouts. Simulation is…
Distributed machine learning is becoming a popular model-training method due to privacy, computational scalability, and bandwidth capacities. In this work, we explore scalable distributed-training versions of two algorithms commonly used in…
Memory corruption vulnerabilities often enable attackers to take control of a target system by overwriting control-flow relevant data (such as return addresses and function pointers), which are potentially stored in close proximity of…
Resource disaggregation offers a cost effective solution to resource scaling, utilization, and failure-handling in data centers by physically separating hardware devices in a server. Servers are architected as pools of processor, memory,…
Dynamic resource management is essential for optimizing computational efficiency in modern high-performance computing (HPC) environments, particularly as systems scale. While research has demonstrated the benefits of malleability in…
Despite the promise of Vision-Language-Action (VLA) models as generalist robotic controllers, their robustness against perceptual noise and environmental variations in out-of-distribution (OOD) tasks remains fundamentally limited by the…
WebAssembly's (Wasm) monolithic linear memory model facilitates memory corruption attacks that can escalate to cross-site scripting in browsers or go undetected when a malicious host tampers with a module's state. Existing defenses rely on…
Training with larger mini-batches improves the convergence rate and can yield superior performance. However, training with large mini-batches becomes prohibitive for Large Language Models (LLMs), due to the large GPU memory requirement. To…
The configurable building blocks of current FPGAs -- Logic blocks (LBs), Digital Signal Processing (DSP) slices, and Block RAMs (BRAMs) -- make them efficient hardware accelerators for the rapid-changing world of Deep Learning (DL).…
We present Hierarchical Memory Matching Network (HMMN) for semi-supervised video object segmentation. Based on a recent memory-based method [33], we propose two advanced memory read modules that enable us to perform memory reading in…
Fine-tuning large-scale pre-trained models is inherently a resource-intensive task. While it can enhance the capabilities of the model, it also incurs substantial computational costs, posing challenges to the practical application of…
Parameter-Efficient Fine-Tuning (PEFT), particularly Low-Rank Adaptation (LoRA), has become a standard approach for adapting Large Language Models (LLMs) under limited compute. However, in continual settings where models are updated…
Disaggregated memory is promising for improving memory utilization in computer clusters in which memory demands significantly vary across computer nodes under utilization. It allows applications with high memory demands to use memory in…
Federated Learning (FL) has emerged as a de facto machine learning area and received rapid increasing research interests from the community. However, catastrophic forgetting caused by data heterogeneity and partial participation poses…
With the continuously increasing integration level, manycore processor systems are likely to be the coming system structure not only in HPC but also for desktop or mobile systems. Nowadays manycore processors like Tilera TILE, KALRAY MPPA…
We present MACLA, a framework that decouples reasoning from learning by maintaining a frozen large language model while performing all adaptation in an external hierarchical procedural memory. MACLA extracts reusable procedures from…
In this paper, we present Latest Object Memory Management (LOMM) for temporally consistent video instance segmentation that significantly improves long-term instance tracking. At the core of our method is Latest Object Memory (LOM), which…