NVMdurance launches white paper on automation of 3D flash trimming

Dublin, Ireland and San Jose, Calif – July 26th, 2016 – Irish technology company, NVMdurance, White paper 2016extending the life of solid state discs (SSDs), has today launched a white paper that provides detailed information on how machine learning can optimize the trade-offs between storage capacity, endurance and data retention of flash memory products.

The way that data is stored in flash memory devices can lead to wear and limitations in data retention with repeated erase and program cycles and even with multiple reading of cells. These issues become even more important as the number of bits per flash memory cell increases. The tuning of registers in 2D as well as 3D flash memory can be used to optimize flash wear, storage capacity and data retention.

The white paper, Machine Learning Enables Longer Life High Capacity SSDs, prepared by Coughlin Associates , explains about the trade-offs between storage capacity, endurance and data retention and how machine learning techniques such as that provided by NVMdurance can optimize these important flash memory characteristics.

“The NVMdurance technology can also avoid runtime impact and bad tail latency (long latencies in a distribution of operations) due to read retries,” said Tom Coughlin of Coughlin Associates and author of the paper. “This optimization can furthermore compound any endurance gain achieved by some other means, such as more powerful ECC or overprovisioning.”

Mr Coughlin continued: “These advantages make automated flash memory tuning, using machine learning, attractive and should result in this approach becoming pervasive in the industry, particularly as the industry moves to 3D flash. It provides a competitive advantage for those who implement it, and a disadvantage for those companies who don’t.”

“NVMdurance software compounds the endurance gain currently being achieved by other means,” said Pearse Coyle, NVMdurance CEO. “We are very pleased to have a sound evidential estimate as to the advantage our technology brings to the flash memory market.”