NVMe was developed to support the higher demands of NAND flash storage and increased bandwidth of the PCIe^® interface. 

Early PCIe SSDs often used proprietary firmware that could limit scalability. NVMe emerged as a standardized and highly scalable storage protocol to address those issues.  

NVMe SSDs make use of parallel, low-latency data paths. This means they can handle demanding workloads with a smaller infrastructure footprint compared to SAS and SATA SSDs.

This kind of low latency is important wherever NVMe is used, but it’s especially critical in data center environments. By using fewer storage devices and CPU cycles to achieve the same result, enterprises maximize their return on storage infrastructure, leading to a reduction in space and power consumption.  

PCIe (peripheral component interconnect express) is the interface standard that NVMe is built on. PCIe slots may be used for graphics cards, SSDs, or HDDs. 

PCIe slots come in several sizes — such as PCIe x1, x4, x8, and x16 — that reflect the number of lanes available for data transfer. NVMe’s ability to take advantage of multiple lanes helps drive its fast transfer speeds.