6 Reasons to Consider NVMe

Are you evaluating your organization’s future infrastructure purchases? If so, are you wondering if Non-Volatile Memory Express (NVMe) should be included into the product conversation? Rest assured, you have a lot of company, as more and more IT Directors and CTOs are questioning whether or not to deploy NVMe. NVMe was designed to take advantage of the unique properties available with pipeline-rich, random access, and memory-based storage.

As a result, I get questions about NVMe from customers daily:

"Is it really that fast?"

"Is there really that much reduction in latency?"

"What about integration? Are manufacturers getting behind the technology from a price/volume/availability perspective?"

Up to this point, the cost and limited availability of ready platforms to adopt NVMe have presented some industry challenges. And as a result, as some customers plan for greater adoption of NVMe based solid state drives (SSD), I have to ask the question - is now the right time for your organization to deploy NVMe?

Ultimately your organization’s ability to deploy NVMe is a function of your use case, technical and business requirements. Some of our customers are just not ready yet, but then some are beginning to successfully deploy NVMe, and that alone is a strong indicator for the readiness of the technology.

Here are my Top 6 reasons why I think now is a good time for your organization to consider NVMe.

1. The cost (time) of getting access to data and information is reduced. I/O intensive applications that require high throughput and low latency will finally get the proper storage medium to host that type of transient data.
2. “Big Data” and OLTP relational database platforms with intensive workloads can benefit from NVMe and the elimination of storage as a bottleneck. This bottleneck reduction can empower organizations to make fast, real-time data based decisions.
3. Those that also have to backup or replicate data within compliance windows (aka backup windows) will now have an option to utilize NVMe to meet those tight timeframes. Data is growing, but the backup window is constant.
4. High-Performance Computing (HPC) research initiatives looking for a value-add can benefit from NVMe’s speed and performance with faster job completion, enhanced cluster utilization, with 6x performance improvement over SATA for a modest 40-50% price increase, that is a 4x TCO improvement over SATA SSDs, or $0.004 to $0.007/IOP.
5. Allows drop-in acceleration for certain workloads. It’s simply identifying the hot spot and dropping it into the server. Plus, users can conduct “in place”, or “fleet upgrades”, depending on what server platform is chosen.
6. Lowers TCO for virtualization environments by increasing VM density. Hypervisor environments are often partitioned to support certain workloads and other considerations like IOPs or latency, often times increasing both OpEx (staff, time) or CapEx (infrastructure). NVMe makes managing virtualization clusters easier, by more readily supporting multi-tenant applications, databases, and heterogeneous workloads (higher density and consolidation).

As we continue to see flash adoption rates trend upward, what separates NVMe is that it is the first platform designed from the ground up to support flash or memory based storage. NVMe based SSDs presents a compelling technology option as it will continue to displace SATA with its superior TCO. NVMe is a clear win for anyone purchasing flash, or fast spinning disk, for a performance benefit.

Ready to investigate NVMe on your own, or have questions? Reach out to us to see if this is the right solution for your current needs and environment. You might find that the answer is yes -- that the time to implement NVMe is now.