Speaker: Anahita Shayesteh
NVMe offers a standards based high-performance interface to flash storage. NVMe SSDs delivers low-latency and high throughput storage and are now available from multiple vendors. In this talk, we outline the system performance benefits when these devices are used for local and remote storage. We will also present the system performance improvements when these SSDs are used in Big Data frameworks and enterprise RDBMS application workloads. We will conclude by exploring the opportunities and challenges presented by these devices to storage sub-systems and server architecture.
Anahita Shayesteh is a senior system and performance architect at Samsung Memory Solutions Lab in San Jose, California. Previous to that she was a research scientist at Intel Labs in Santa Clara. She graduated from UCLA in 2006 finishing her PhD in Computer Architecture under Prof. Glenn Reinman. She has worked on various topics in architecture including CPU and GPU design as well as Cache and Interconnect design. Her current focus is on performance analysis of datacenter applications, especially as related to the memory hierarchy and storage systems.
Hosted by Professor Jason Cong
REFRESHMENTS at 3:45 pm, SPEAKER at 4:15 pm
Date(s) - Oct 06, 2015
4:15 pm - 5:45 pm