While it’s been 40 years since flash memory was invented by Toshiba, it’s been over the past two decades that we’ve begun to use it in our everyday lives: thumb drives and compact flash cards (the predecessors to SD cards) entered the market around the year 2000. SSDs, which were once large, refrigerator-sized solid-state storage devices attached to multi-million dollar supercomputers, have shrunk, implemented flash memory, and became cost-effective enough to be used in data centers and now as the internal storage device of choice for PCs and laptops over the past several years.
When SSDs moved in volume from the data center to the PC, they were viewed as straightforward hard drive replacements and used the SATA protocol to communicate with the PC motherboard and its components. Subsequently, wide adoption of the PCI bus, new storage form factors, increased memory density, and the faster NVMe protocol all combined to lead to the M.2 SSDs used as boot drives in PCs and laptops today.
The data throughput of these NVMe SSDs far exceeds the SATA technology that was popular in the late 2000s. Current NVMe SSDs (Gen 3) have a sequential read/write speed of about 3,200 MB/s, whereas SATA SSDs max out around 550 MB/s. Many high-performance applications such as AI, high-speed imaging, visualization, simulation, and high-throughput data capture can take advantage of the much faster NVMe SSD throughput speeds. For example, autonomous vehicle prototyping collects an enormous amount of data during testing. Traditional SATA-based storage is not able to keep up so those technologies are rapidly being replaced by NVMe SSDs.
The cost per GB for an NVMe SSD is rapidly approaching SATA SSDs, so it’s not long before they replace SATA technology in most applications. Along with wide adoption of NVMe SSDs comes the desire to make them easily removable so security-conscious organizations can remove or transport sensitive data as easily as they do with SATA hard drives and SSDs today.
Product Marketing Manager