Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

Purchasing flash storage: Types and prices explained

Purchasing flash storage can be an involved process. Buyers need to understand their priorities when it comes to storing data and familiarize themselves with the types of flash storage offered by vendors. Dennis Martin, founder and president of computer industry analyst firm Demartek, spoke with Senior News Director Dave Raffo about some items to consider when making a flash storage purchase decision.

Single-level cell (SLC) flash was one of the earliest types of flash storage, but it has been improved upon by multi-level cell (MLC) and, most recently, triple-level cell (TLC) flash. Previously, SLC had been considered to be the flash storage type that offered the greatest amount of endurance. While MLC and TLC flash were generally less expensive for an equal amount of storage, their lack of endurance made them more likely to be used in consumer products, making SLC the main force in enterprise arrays. However, developments in MLC and TLC have made these options far cheaper than SLC and have provided increased endurance. Recently, SLC has begun to lose its place in the market. Martin said SLC hasn't received much attention lately, and noted that it's generally more expensive than other types of flash storage. He also said any type of all-flash array will likely work well, regardless of the flash technology used. "If you really want SLC," he said, "you can get it, and you'll pay more for it, but any of these all-flash technologies work really well, actually."

But for Martin, the types of flash storage used weren't as important as the features provided by the storage array. When buying a flash storage array, he would "look for the same kinds of things you look for in any kind of enterprise array" -- features such as resiliency, uptime, warranties, deduplication, compression, replication, disaster recovery and snapshots.

It's also important to understand the terms used with flash prices, such as effective capacity. The compression and deduplication capabilities of a flash array can increase the amount of data that can be stored, essentially increasing the capacity of the storage device. "They'll typically say, here's the average dedupe and/or compression ratio that we advertise, let's say it's 6:1," Martin explained. "The price is based on that sort of capacity. So instead of paying, let's say if it was 6:1, instead of paying $600, if it's 6:1 you might be only [paying] $100 per gigabyte."

Dollars-per-IOPS quantifies how effective a flash array is at performing I/O operations. Faster flash arrays will likely cost less when it comes to dollars-per-IOPS.

When it comes to buying a flash array, it's important to know what qualities you're looking for. "If you're buying for capacity," Martin said, "you might want to look more at the effective capacity price. But if you're buying for performance, then dollar-per-IOPS is a good way to go." In terms of performance, flash will generally outperform other types of storage. If overall capacity is the most important metric, hard drives and tape are generally more cost effective.

Editor's note:

This is the second part of a two-part interview about flash storage. The first part dives into the different features of flash storage arrays and compares flash storage to hard drive storage.

View All Videos

Transcript - Purchasing flash storage: Types and prices explained

Hello, my name is Dave Raffo, I'm the senior news director for TechTarget's Storage Media sites. Joining me today to discuss solid-state deployment options is Dennis Martin, founder and president of Demartek, an industry analyst frm with its own hands-on testing lab. Welcome Dennis.

Dennis Martin: Thanks, Dave.

If a company is just getting started with all-flash storage, what are the alternatives for using flash in an enterprise storage system?

Martin: There are lots of choices with companies and brand names. But the main thing is that there are flash drives, flash modules under various names, and there are NVMe (nonvolatile memory express) implementations inside of storage systems. There are a lot of choices, but [they are] fast and do a good job.

What features should a storage buyer look for when shopping for an all-flash storage system?

Martin: I would look for the same kinds of things  you look for in any kind of enterprise array: resiliency, uptime, good warranties, that sort of thing. I would also look for things like dedupe and compression, replication, any kind of disaster recovery features, snapshots -- the same things you would look for in a storage array.

What about the different types of media for flash. We hear about MLC and SLC, and now we have 3D NAND TLC drives. What do shoppers need to look for and what's different with  those types of media?

Martin: I don't see a lot of news about SLC lately because the price points are better with MLC. And now that TLC has been introduced, we're starting to see bigger capacities with TLC NAND. I would look at what are the other features? What does it need? If you really want SLC, of course you can get it, and you'll pay more for it, but any of these all-flash technologies work really well, actually.

In your testing, do you see a big difference in performance between all-flash arrays from different vendors?

Martin: We haven't put a lot of them head-to-head against each other, running the exact same tests. All I can say is they all do well.

What do vendors mean when they talk about dollars-per-gigabyte of effective capacity on an all-flash array?

Martin: Effective capacity with flash has to do with [the following]: Here's the raw capacity, but then they're doing things like dedupe and compression all the time. That means you get compression and dedupe built in. If you have a 100 gigabytes of data -- depending on the kind of data -- you might accidentally take up 50 gigabytes or maybe 30 or 20, it depends on the ratios. They're selling the effective capacity. They'll typically say, here's the average dedupe and/or compression ratio that we advertise, let's say it's 6:1, or 7:1 or something of that nature. So then ... the price is based on that sort of capacity. So instead of paying, let's say if it was 6:1, instead of paying $600, if it's 6:1 you might be only [paying] $100 per gigabyte. That's just a hypothetical number of course.

We also hear vendors quote dollars-per-IOPS pricing for flash. What do you consider the most important pricing metrics for an all-flash array?

Martin: I would take a combination of both effective capacity and dollars-per-IOPS. The main thing when you talk about dollars-per-IOPS is that you're comparing the performance and the price for flash versus the performance and price for hard drives. So hard drives, of course, are cheaper just in raw capacity, but in terms of dollar-per-IOPS it's completely upside down; the flash is much more efficient that way. If you're buying for capacity, you might want to look more at the effective capacity price; but if you're buying for performance, then dollar-per-IOPS is a good way to go.

Thanks Dennis.

Martin: Thanks Dave.

That wraps up today's session on solid-state deployment options with Dennis Martin. I hope you found this informative, thanks for watching.

+ Show Transcript

Join the conversation

3 comments

Send me notifications when other members comment.

By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy

Please create a username to comment.

What plays the biggest role when selecting among the different flash storage types?
Cancel
Nice article that helps clear up some of the confusion for those new to flash storage.
Cancel
Thanks so much, Todd. Glad you liked the article.
Cancel

-ADS BY GOOGLE

SearchSolidStateStorage

SearchCloudStorage

SearchDisasterRecovery

SearchDataBackup

Close