On the role of burst buffers in leadership-class storage systems
The largest-scale high-performance (HPC) systems are stretching parallel file systems to
their limits in terms of aggregate bandwidth and numbers of clients. To further sustain the …
their limits in terms of aggregate bandwidth and numbers of clients. To further sustain the …
Label-free tissue scanner for colorectal cancer screening
The current practice of surgical pathology relies on external contrast agents to reveal tissue
architecture, which is then qualitatively examined by a trained pathologist. The diagnosis is …
architecture, which is then qualitatively examined by a trained pathologist. The diagnosis is …
Parallel I/O and the metadata wall
SR Alam, HN El-Harake, K Howard… - Proceedings of the sixth …, 2011 - dl.acm.org
Large HPC installations typically make use of parallel file systems that adhere to POSIX I/O
conventions, and that implement a separation of data and metadata in order to maintain high …
conventions, and that implement a separation of data and metadata in order to maintain high …
Triple-A: A non-SSD based autonomic all-flash array for high performance storage systems
Solid State Disk (SSD) arrays are in a position to (as least partially) replace spinning disk
arrays in high performance computing (HPC) systems due to their better performance and …
arrays in high performance computing (HPC) systems due to their better performance and …
Middleware-firmware cooperation for high-speed solid state drives
State-of-the-art solid state drives (SSDs) equip multiple internal resources such as NAND
flash chips, data buses, and cores to take advantage of internal parallelism and provide fast …
flash chips, data buses, and cores to take advantage of internal parallelism and provide fast …
An infinite dimensional model for a single server priority queue
We consider a Markovian single server queue in which customers are preemptively
scheduled by exogenously assigned priority levels. The novelty in our model is that the …
scheduled by exogenously assigned priority levels. The novelty in our model is that the …
A scale-out enterprise storage architecture
A robust enterprise SSD design should provide scalable throughput and storage capacity by
integrating (up to thousands) flash chips in a scale-out fashion. However, the current" …
integrating (up to thousands) flash chips in a scale-out fashion. However, the current" …
NAND flash memory-based SSD controller using full page-associative block recycling mechanism
M Nishi, K Matsuda, R Inada… - 2011 3rd IEEE …, 2011 - ieeexplore.ieee.org
A 36GB SSD controller enabling write endurance of 2 PB and 32K internal IOPS under
random 4 KB page write workload is presented that offers 10 years of lifetime for random …
random 4 KB page write workload is presented that offers 10 years of lifetime for random …
Case-Specific Lifetime Management Strategies for Long-Lasting Flash Devices
W Choi - 2021 - etda.libraries.psu.edu
NAND flash memory-based devices have limited lifetime, which should be carefully and
effectively managed for a longer use. This is because flash lifetime is non-renewable …
effectively managed for a longer use. This is because flash lifetime is non-renewable …
[PDF][PDF] Evaluation of a flash storage filesystem on the cray XE-6
J Srinivasan, RS Canon - Procedia Computer Science, 2013 - cug.org
• Flash storage is a scarce resource and cannot be used for long-‐term storage, or even for
much longer than the duration of the job• I/O acceleration can be explicitly requested by the …
much longer than the duration of the job• I/O acceleration can be explicitly requested by the …