Related papers: LFTL: A multi-threaded FTL for a Parallel IO Flash…
We present the most interesting elements of the correctness specification of BilbyFs, a performant Linux flash file system. The BilbyFs specification supports asynchronous writes, a feature that has been overlooked by several file system…
Autoregressive large language models (LLMs) deliver strong performance but require inherently sequential decoding, leading to high inference latency and poor GPU utilization. Speculative decoding mitigates this bottleneck by using a fast…
We propose the design and an implementation of a bulk-parallel external memory priority queue to take advantage of both shared-memory parallelism and high external memory transfer speeds to parallel disks. To achieve higher performance by…
We present LearnedFTL, a new on-demand page-level flash translation layer (FTL) design, which employs learned indexes to improve the address translation efficiency of flash-based SSDs. The first of its kind, it reduces the number of double…
The demise of Moore's Law and Dennard Scaling has revived interest in specialized computer architectures and accelerators. Verification and testing of this hardware depend heavily upon cycle-accurate simulation of register-transfer-level…
As LLMs and foundation models scale, checkpoint/restore has become a critical pattern for training and inference. With 3D parallelism (tensor, pipeline, data), checkpointing involves many processes, each managing numerous tensors of varying…
This paper describes a novel design of a threshold logic gate (a binary perceptron) and its implementation as a standard cell. This new cell structure, referred to as flash threshold logic (FTL), uses floating gate (flash) transistors to…
Block devices in computer operating systems typically correspond to disks or disk partitions, and are used to store files in a filesystem. Disks are not the only real or virtual device which adhere to the block accessible stream of bytes…
Lattice Boltzmann method (LBM) is a promising approach to solving Computational Fluid Dynamics (CFD) problems, however, its nature of memory-boundness limits nearly all LBM algorithms' performance on modern computer architectures. This…
Hardware peripherals such as GPUs and FPGAs are commonly available in server-grade computing to accelerate specific compute tasks, from database queries to machine learning. CSPs have integrated these accelerators into their infrastructure…
The performance of data intensive applications is often dominated by their input/output (I/O) operations but the I/O stack of systems is complex and severely depends on system specific settings and hardware components. This situation makes…
In recent years, emerging storage hardware technologies have focused on divergent goals: better performance or lower cost-per-bit. Correspondingly, data systems that employ these technologies are typically optimized either to be fast (but…
Parallel programming remains a daunting challenge, from the struggle to express a parallel algorithm without cluttering the underlying synchronous logic, to describing which devices to employ in a calculation, to correctness. Over the…
High parallel framework has been proved to be very suitable for graph processing. There are various work to optimize the implementation in FPGAs, a pipeline parallel device. The key to make use of the parallel performance of FPGAs is to…
The standardization of NVMe Zoned Namespaces (ZNS) in the NVMe 2.0 specification presents a unique new addition to storage devices. Unlike traditional SSDs, where the flash media management idiosyncrasies are hidden behind a flash…
RAPID-LLM is a unified performance modeling framework for large language model (LLM) training and inference on GPU clusters. It couples a DeepFlow-based frontend that generates hardware-aware, operator-level Chakra execution traces from an…
Recently, 3D SSDs are widely adopted in PCs, data centers, and cloud storage systems. To increase capacity, high bit-density cells, such as Triple-Level Cell (TLC), are utilized within 3D SSDs. However, due to the inferior performance of…
While Transformers and other sequence-parallelizable neural network architectures seem like the current state of the art in sequence modeling, they specifically lack state-tracking capabilities. These are important for time-series tasks and…
Shared memory multiprocessors come back to popularity thanks to rapid spreading of commodity multi-core architectures. As ever, shared memory programs are fairly easy to write and quite hard to optimise; providing multi-core programmers…
High-energy physics experiments are currently recording large amounts of data and in a few years will be recording prodigious quantities of data. New methods must be developed to handle this data and make analysis at universities possible.…