Ozone (O3): An Out-of-Order Flash Memory Controller Architecture

Nam, Eyee Hyun; Kim, Bryan S.; Eom, Hyeonsang; Min, Sang Lyul

doi:10.1109/tc.2010.209

Cited by 64 publications

(35 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Entry-level flash storage devices, such as memory cards and USB thumb drives, commonly adopt synchronized channels. Independent channels are a popular design option of many commodity solid-state disks and have been studied in many prior researches [Agrawal et al 2008;Nam et al 2011;Seong et al 2010;Park et al 2012]. The experiments in this study ignore host read requests, because serving read requests does not trigger garbage collection and it is much faster than serving write requests.…”

Section: Methodsmentioning

confidence: 99%

“…Many commodity solid-state disks refuse to sacrifice their super-fast read performance, which has been the iconic advantage of flash-storage devices, and choose static channel binding, like RAID-0 style striping. Many prior studies have been conducted on the static channel binding of page data [Agrawal et al 2008;Nam et al 2011;Seong et al 2010;Park et al 2012]. With static channel binding, channels can be viewed as substorage devices that adopt their own instances of flash-translation layer for space management.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Reducing asynchrony in channel garbage-collection for improving internal parallelism of multichannel solid-state disks

Chang

Wen

2014

ACM Trans. Embed. Comput. Syst.

View full text Add to dashboard Cite

Solid-state disks use multichannel architectures to boost their data transfer rates. Because realistic disk workloads have numerous small write requests, modern flash-storage devices adopt a write buffer and a set of independent channels for better parallelism in serving small write requests. When a channel is undergoing garbage collection, it stops responding to inbound write traffic and accumulates page data in the write buffer. This results in contention for buffer space and creates idle periods in channels. This study presents a channel-management strategy, called garbage-collection advancing, which allows early start of garbage collection in channels for increasing the overlap among channel activities of garbage collection and restoring the balance of buffer-space usage among channels. This study further introduces cycle filling, which is a version of garbage-collection advancing tailored for the operation model of flash planes. Experimental results show that the proposed methods greatly outperformed existing designs of multichannel systems in terms of response and throughput. We also successfully implemented the proposed methods in a real solid-state disk and proved their feasibility in real hardware. ACM Reference Format:Li-Pin Chang and Chen-Yi Wen. 2013. Reducing asynchrony in channel garbage-collection for improving internal parallelism of multichannel solid-state disks.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Reducing asynchrony in channel garbage-collection for improving internal parallelism of multichannel solid-state disks

Chang

Wen

2014

ACM Trans. Embed. Comput. Syst.

View full text Add to dashboard Cite

show abstract

“…Modern SSDs are designed to exploit the parallelism of multiple units of the storage medium. Their device architectures allow outstanding I/O requests to be pipelined and reordered for better scheduling [Nam et al 2011] and to be dispatched to nonconflicting memory chips in parallel [Chang and Kuo 2002;Lee et al 2010;Yoo et al 2011]. Fusion-IO is known to embed as many flash chips as possible into their SSD products to benefit from the parallelism [Fusion-IO 2014a].…”

Section: Hardware Technology Trendsmentioning

confidence: 99%

“…The device I/O interface implemented in an SSD as well as in an HDD limits the benefit of accessing multiple memory chips in parallel. Although a single large I/O request consisting of small contiguous requests enhances the parallelism inside a device [Agrawal et al 2008;Akel et al 2011;Chen et al 2009;Nam et al 2011], a random access workload cannot take advantage of merged I/O requests. Command queueing and interrupt aggregation help the block I/O subsystem to amortize per-request overhead by overlapping descending and ascending I/O paths, respectively [Dees 2005;Intel and Seagate 2003;Yu et al 2010], but the optimizations rely on an intermediate hardware controller that adds an additional latency to an I/O request.…”

Section: Problem 2: Low Random Throughputmentioning

confidence: 99%

“…Semi-conductor technology has introduced nonvolatile memory to system researchers in academia and industry, opening a gate for new storage devices with the new storage media [Agrawal et al 2008;Chen et al 2009;Nam et al 2011]. The new storage devices are able to transfer data from/to an OS without necessarily moving physical sensors back and forth, achieving orders of magnitude improvements over HDDs in both latency and throughput.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Optimizing the Block I/O Subsystem for Fast Storage Devices

Shin

et al. 2014

ACM Trans. Comput. Syst.

Self Cite

View full text Add to dashboard Cite

Fast storage devices are an emerging solution to satisfy data-intensive applications. They provide high transaction rates for DBMS, low response times for Web servers, instant on-demand paging for applications with large memory footprints, and many similar advantages for performance-hungry applications. In spite of the benefits promised by fast hardware, modern operating systems are not yet structured to take advantage of the hardware's full potential. The software overhead caused by an OS, negligible in the past, adversely impacts application performance, lessening the advantage of using such hardware. Our analysis demonstrates that the overheads from the traditional storage-stack design are significant and cannot easily be overcome without modifying the hardware interface and adding new capabilities to the operating system.In this article, we propose six optimizations that enable an OS to fully exploit the performance characteristics of fast storage devices. With the support of new hardware interfaces, our optimizations minimize per-request latency by streamlining the I/O path and amortize per-request latency by maximizing parallelism inside the device. We demonstrate the impact on application performance through well-known storage benchmarks run against a Linux kernel with a customized SSD. We find that eliminating context switches in the I/O path decreases the software overhead of an I/O request from 20 microseconds to 5 microseconds and a new request merge scheme called Temporal Merge enables the OS to achieve 87% to 100% of peak device performance, regardless of request access patterns or types. Although the performance improvement by these optimizations on a standard SATA-based SSD is marginal (because of its limited interface and relatively high response times), our sensitivity analysis suggests that future SSDs with lower response times will benefit from these changes. The effectiveness of our optimizations encourages discussion between the OS community and storage vendors about future device interfaces for fast storage devices. Portions of this work appeared in the 4th USENIX Workshop on Hot Topics in Storage and File Systems [Yu et al. 2012] and in Y. J. Yu's Ph.D. dissertation [Yu 2012].

show abstract

A High-Speed Parallel Accessing Scheduler of Space-Borne Nand Flash Storage System

et al. 2020

Lecture Notes in Electrical Engineering

View full text Add to dashboard Cite

Ozone (O3): An Out-of-Order Flash Memory Controller Architecture

Cited by 64 publications

References 26 publications

Reducing asynchrony in channel garbage-collection for improving internal parallelism of multichannel solid-state disks

Reducing asynchrony in channel garbage-collection for improving internal parallelism of multichannel solid-state disks

Optimizing the Block I/O Subsystem for Fast Storage Devices

A High-Speed Parallel Accessing Scheduler of Space-Borne Nand Flash Storage System

Contact Info

Product

Resources

About