Can Pure Storage’s increasingly complete portfolio, coupled with ever lower prices per gigabyte and the company’s evergreen approach, take the madness out of the storage market?
Pure is trying to do something special in the storage market. It is trying to sell something seen as a commodity as something very special. It has additionally chosen from its inception to focus only on flash, not hard drives. This strategy has worked well for the company. Things are definitely going well for Pure. Thus, at Pure//Accelerate, the company’s annual conference, representatives spoke with great confidence about the company’s products and services and its future.
Still, it’s always good to keep using two words in the storage market. After all, it is a crazy industry. We don’t say that, says Charlie Giancarlo, the CEO of Pure Storage. We spoke to him recently when he visited Pure’s offices in the Netherlands. How does he view the market and what are Pure’s plans for the coming years?
More coverage with //E portfolio
Let’s start with the second question. The answer to this can also be found in extensive form in the article we published following our visit to Pure//Accelerate a few months ago. Pure wants to make hard drives obsolete by 2025. Giancarlo puts it succinctly, “In 2.5 years, customers will only be thinking about flash and tape.” This, by the way, does not imply that Pure will be the only remaining player, but rather that all other established names will also switch from hard disks to flash (or tape, of course, but that is not so relevant in the context of this article).
To achieve this, there is still work to be done, by the way, including by Pure. For starters, it needs to start offering flash solutions for all the data for which it makes sense to do it, that is, which cannot be stored on tape. The announcement of FlashBlade//E early this year and FlashArray//E during Accelerate are very important steps in this regard, he points out. “Until //E we covered about 20 percent of all data (i.e. primary data) with our products and services,” according to Giancarlo. Then you can have a good story in terms of evergreen, sustainability and ease of use, you can never achieve the coverage you want to achieve. So that’s what the //E portfolio has to make possible.
From fragmented environments to storage platforms
Mind you, FlashBlade//E and FlashArray//E have only just arrived. So it’s not like Pure can serve that other 80 percent at this point. Giancarlo readily admits as much. But he also points out that we are at an inflection point. That is, the ground gained by flash over hard disks is going to become even more serious as of now. “Flash has long been catching up with hard disks in terms of price/performance. The growth it is experiencing in that is exponential and is now moving very fast,” Giancarlo said. In addition, he points to the fact that flash is a semiconductor. That makes more possible than the relatively antiquated technology that hard drives use, is the idea. We covered that, too, in our earlier article. Flash allows customers to move more toward storage platforms.
A move toward storage platforms sounds interesting. It’s also necessary actually, because the storage market is a special market. Giancarlo calls it a “crazy industry,” as we cited earlier. “It is very fragmented, yet the big vendors think storage is or is becoming a commodity,” he sums up. This while there is actually virtually no standard for storage. Giancarlo has a history in the networking world, where it’s all about TCP/IP or UDP. Within compute, VMware can be called a standard that everyone can get along with.
“Storage, however, does not inter-operate with what an organization already has in place,” Giancarlo points out. “Even if you buy everything from the same vendor, you’re probably still buying different systems, with different APIs,” he continues. Organizations have to try to make a whole out of these separate components themselves, which is not easy. “What was missing was a standardized and consolidated environment that works as a system, not as a collection of components,” he sums up the reason for Pure Storage’s existence.
At this point, it’s worth pausing to consider what it means when an organization moves more toward a system or platform in terms of storage. It’s not that the people responsible for storage are then no longer needed. “Storage architects are still needed, building environments according to a set of predetermined rules,” Giancarlo points out when asked. The added value of systems and platforms for storage is much more in the follow-up. That’s when policies come into play. They actually take over. A policy-driven storage environment is very difficult, if not impossible, to achieve if you don’t start from a platform.
The road to 2025
In general, Pure’s story certainly makes sense. We stated as much in our earlier article. There are some variables, but they seem to be moving in the right direction. In particular, it concerns the price per gigabyte. That is still one of the most important criteria for many buyers when buying storage systems and all that comes with it. You can have such a great proposition, if it is or looks too expensive, you’re in trouble.
This is where Pure is going to make great strides in the coming years, Giancarlo points out. He shows us an overview of the types of storage that Pure wants to take over from hard drives with their DFM-based flash systems. Tier 1 they have basically covered already, but Tier 2 has many tiers. At 20 cents per gigabyte, which they are already at with the //E portfolio, they have basically covered nearline storage. Next year Pure wants to go to 15 cents per gigabyte to go after JBOD, while by 2025 they’ll be at 10 cents per gigabyte. That will also put deep archive storage within reach. Once that is achieved, the move to hyperscalers can be made.
Of course, all these predictions are based on the expectation that there are and will continue to be enough raw materials to make that much flash storage. We don’t really have a good idea of that at the moment, although people at Pure don’t seem too worried about it.
Price/performance alone is not enough
Price/performance is definitely very important. But Giancarlo also sees that in this crazy industry it is not enough. He explains with wonder in his voice that the storage industry is “the only industry that won’t be persuaded if something performs 10x better.” Once a party is a customer of Pure, they don’t have to convince them to stay a customer, if we look at the customer retention rates. Those are very high. For prospects, the situation is different. They almost always say that every supplier tells the same story. And that it therefore seems like it doesn’t matter all that much what you buy.
The reason for the above customer attitude, again, is that storage is seen as a commodity, Giancarlo confirms. “Very much so,” he replies when we ask him if this is the reason. “Most organizations also place the responsibility for storage at relatively very low levels within the company,” he notes. The provisioning and procurement of storage has basically gone the same way for about 20-30 years. “It’s never of strategic importance, always for specific workloads,” according to him.
In that respect, it’s similar to how the security market has been functioning (or not functioning, actually) for a very long time, we note. Giancarlo agrees, but with the addition that when looking for new security solutions, organizations do get convinced if something is 10x better than other propositions in the market. By the way, another parallel with the security industry is that there, too, a trend toward platforms is currently underway, but that aside.
Pure doesn’t only want to talk about storage. Therefore, the company focuses more and more on what it calls the Digital Experience. This experience consists of a collection of SLAs. These include SLAs around uptime, buffer capacity, performance, zero planned downtime, energy efficiency and the most recent is ransomware recovery. The latter is an add-on; the first five are included as standard with Pure’s offering. “Expect a string of SLAs to come out,” Giancarlo reflects the ambitions in this area. This should ensure that Pure’s platform approach can even better demonstrate its added value. This also means that it will become easier to get over the commodity argument in conversations with prospects.
A final point we touch on in our conversation with Giancarlo is the topic of sustainability. After all, that’s another big advantage of flash systems over hard drives. During Accelerate, someone from Virgin Media O2 was on stage talking about a 98 percent reduction in energy consumption compared to what they had in place before they switched to Pure. Of course, such math depends on where you’re coming from and how old (read: inefficient) the equipment was. But it should be clear that significantly lower square footage and racks (due to the much higher density of DFM modules compared to hard drives), coupled with inherently lower power consumption can result in substantial savings in energy consumption.
However, sustainability is not equally important everywhere, Giancarlo points out. In the U.S., for example, it still doesn’t play a role for most customers. In Europe, on the other hand, it plays much more of a role. He is pretty critical of how the general approach to this anyway. There are various formulas to calculate an ESG score, but those are mostly checkmarks to be checked. “That is to some extent greenwashing,” he judges.
Still, especially in the area of IT, there are very hard numbers. You have a clear lifecycle of a product, divided into several scopes. Scope 1 emissions are those that are entirely under Pure’s control, Scope 2 emissions include building the products, where one also depends on other parties, and Scope 3 emissions are when customers actually use the products. You can put a positive spin on this through offsets, but that’s not a “true metric,” Giancarlo points out. “A true metric is how much you burn to begin with,” according to him. This is something that European customers in particular are increasingly concerned about, he says.
So on this point, Pure already has a pretty good story, as evidenced by the example of Virgin Media O2. But, of course, there is always room for improvement. Pure has traditionally worked exclusively with x86 hardware (read: Intel Xeon). Could Arm also be an option? The usually very spontaneous Giancarlo answers this very carefully, with a somewhat mysterious smile around his mouth: “We are constantly looking at other technologies.” What exactly this means, everyone must decide for themselves. He does say that the controllers in the FlashArrays and FlashBlades consume the most energy. So that’s where Pure can optimize more.
On to 2025
Based on our conversation with Giancarlo (and our visit to Pure//Accelerate earlier this year), we are very curious about 2025. Whether the choice by then will indeed only be between flash and tape, whether Pure will have achieved all the goals of the roadmap in the meantime, and whether Pure will be able to stick to its current offering to achieve them. In any case, it is clear that the market as a whole still has some steps to take in order to achieve this. That will be the deciding factor, we suspect. Pure seems ready, however, we dare say.