Long-time StorageSearch readers like me know that Zsolt Kerekes writes a lot about SSD design. He frequently comments on what he calls symmetries. It is the notion that some SSD features/characteristics can yield excellent performance in certain scenarios but not in others. Symmetries can also be performance tradeoffs such as speed versus wear.
Kerekes argues that symmetries are a method to “comparatively describe or evaluate any type of SSD using any memory technology and any type of interface.” In an extensive piece, he describes 11 SSD symmetries including ones about read/write, applications, scalability and age. I think the article is a good resource for organizations considering SSD technology for their data center. The article raises many questions that should spur good customer/vendor discussion. Take a look.
What it means to me
Organizations considering SSD technology really need to do their homework about the features and benefits that are going to offer the most value and have the most positive performance effects on their applications. Flexibility, such as being able to integrate multiple memory technologies into a single box (like the K2-H), is going to be increasingly important to customers who want efficiency and customization options. This is especially true because there are many memory innovations coming on the near horizon.
My other take away from Kerekes symmetries discussion is how important software innovation is to the future of SSD SAN storage. Software can mitigate some symmetry tradeoffs as well as to deliver new enhancements such as increased data protection. This is one reason why we have invested so much into SPEAR and DataProtect.
Kerekes’ symmetries article is definitely worth a read for customers wanting to evaluate SSD features in the context of real-world scenarios.