Hi all!
Charlie Giancarlo, Pure Storage CEO, explains in an interview what’s PSTG special sauce in storage business.
This link is from Tinker who posted on NPI board, but the original post have politic issues not suitable for Saul’s board. Sorry Tinker!
https://www.networkworld.com/article/3238830/storage/the-fut…
Thanks to Saul and all contributors for what I have learned and enriched because of this board. I wish I could contribute more!
Best!
María
How impressive!
Charlie Giancarlo has a famous last name that is the first name of Giancarlo Stanton, the probable new HR king in baseball.
I also couldn’t help but note the name…the Yankees are just about sickening, and should probabltmy win 2 of the next 3 World Series with Judge and Giancarlo abusing baseballs.
I really liked the bit at the end of the article, where he (also as a former Cisco guy) indirectly referenced Arista:
Network pros wouldn’t run a software-defined network (SDN) with legacy management tools, and storage managers shouldn’t be managing flash with software built for legacy disks.
The analogy of compute being a 3-legged stool with CPUs/GPUs as one leg (cough, NVDA, cough), networking as another leg (can you say Arista Networks), amd storage as the other leg (seems that Giancarlo had career options as former CTO of Cisco and picked Pure here).
Needless to say, I am pleased that I have NVDA, ANET, and PSTG in my portfolio. Those 2 respectively are currently #1, #3 (almost tied at #2), and #14-ish (with likely bumping to #7 or higher as of Feb. 16th options exercising). The more I learn of Pure, the more I like it going forward.
Charlie Giancarlo, Pure Storage CEO, explains in an interview what’s PSTG special sauce in storage business.
Is anyone worried about this quote:
Giancarlo: The aspect of our solution that customers like most is our Evergreen subscription pricing.
If PURE’s advantage is mostly in its pricing model, then it has no moat.
Yes, later on, he talks about hot swapping/upgrading with their intent to do that perpetually, but that’s also a calculated risk for the company and may, in fact, slow down their advancement of the technology, since anything new hardware wise means a “free to the customer” upgrade cost.
He then compares managing flash arrays to running SDNs with legacy management tools, which doesn’t seem quite appropriate to me. SDN is about the management, but flash is, at its core, just a different hardware instantiation of the same logic, whether it be deduplication or replication. Yeah, I can imagine a few things being different, such as hardware failure prediction, but is that a kind of advantage that will last?
NetApp and EMC got caught napping with flash, but is hard for them, or anyone else for that matter, to catch up technologically? And if PURE’s advantage is pricing, won’t they have a tough road in front of them?
I’m not saying PURE is a bad investment, just trying to understand before I invest.
Smorgasbord,
You can install Pure products in the enterprise with an instruction sheet that is on a business card, and in fact is on a business card.
It is taking what Apple did and what SaaS has done to dramatically improve customer experience and value and bringing it forward to the storage world.
Pure products are specifically architectured to be that simple, and to enable Evergreen, and yet be as sophisticated as any enterprise product out there (and this is still improving as more and more functions are put out there).
No one comes close to the duplex ratios that PSTG does.
And to meet what PSTG does EMC and NTAP would need to totally change the architectures of its products, its software, and its business models. They will not be able to do so without totally upheaving their businesses. They are built for 5 year forklifts and selling all new storage, all the time.
I agree, I don’t know if this Evergreen model, where they get 3 year service contracts and only sell incremental data storage, that also comes with additional 3 year service contracts, is a viable business model or not.
However, they have reached $1 billion in sales faster than any data company in history, and they are cash flow positive from operations and nearly free cash flow positive (if not FCF positive) at present. Thus evidence that the business model can indeed work. They are shooting for 12 to 15% net margins in the current tax environment.
The share price only has stagnation and failure and inability to turn sustainably profitable built into it. And maybe that will be the case in the end. Their new Blade product is clearly prechasm and it is very possible that no real place for it will be found in the market. That product is a brilliant risk, and not all brilliant risks work out in the end.
So you are correct, but there are also reasons to think you are not correct. The share price agrees with you, which creates a rare opportunity of an actually performing very well, with high growth potential company, disrupting an industry, that is not priced in consistent with its narrative and with its business execution.
Normally the market is correct. But in this instance, and I am usually quite resistant to buying anything where the share price dictates failure, but here the share price is also considered expensive by some, and for the rest, I do not see the negatives coming to pass.
In particular, generic white box software will, I think, have higher cost of ownership, and maybe by quite the margin, over the proprietary solutions offered by Pure, thus they are not competing on price, but value of the whole product to its customers. Which is similar actually to ANET, but lets not compare the two any closer as they are different animals without question despite some similarities in thought and execution.
Tinker
Is anyone worried about this quote:
Giancarlo: The aspect of our solution that customers like most is our Evergreen subscription pricing.
Hi Smorgasbord, what he’s doing there is reminding us that what Pure is doing now is selling renewal revenue. That doesn’t worry me at all.
Saul
Tinker,
I agree, I don’t know if this Evergreen model, where they get 3 year service contracts and only sell incremental data storage, that also comes with additional 3 year service contracts, is a viable business model or not.
On the one hand, many people assume that just because there is a (sort of) pay-as-you licensing model that it’s a sign of how easy it is to replace the product.
On the other hand, there are a lot of benefits to having such a model that lock customers even more while ensuring that revenue and profitability gets maximized.
Pure’s Evergreen licensing is probably really simple and most likely not negotiated very much. The elimination (or at least better control) of greedy behavior will keep more value with Pure when transacting business. If Pure also controls the development and level of technology offered, they can service a more standard product, making the solution much more cost effective and customer focused.
Virtually all of our successful Saul companies have such a relatively simple licensing structure.
If companies are successful using the product, they are more likely to need more storage, which in turn means more money to Pure.
DJ
Tinker and Smorgasborg,
I really appreciate this kind of dialogue. I was an IT guy for many years (networking not storage.
Good pro/con discussion. As the final decision maker for purchases, I really liked contracts like Evergreen that lowered my CapEx and made expansion more predictable. I see this as a positive for PSTG. I have a 1/2 position in PSTG and hoping for a pull back with this market so I can fill out the rest. Regardless, long PSTG.
Regards,
Jim
https://blog.purestorage.com/contrasting-pure-netapps-philos…
This is from a pure blog so the source is the source. However, I believe the blog is accurate.
In the end what this is an example of, as is the entire whole product put out by Pure, of how Pure simplifies and makes a product that just works, and works well, and simplifies the technology so you have all the things you really need working and not the rest. Making storage a service or utility.
In this example it shows multi facets of how pure is so simple to use and how even NTAP creates unnecessary complexities.
W Pure you know what you are getting w out having to think 
so hard.
There are other examples such as w EMC who use marketing terms to exaggerate attributes of their products that create similar unnecessary complexities.
In other blogs it is pointed that even beyond you know what you are getting w pure, but also how less labor intensive it is to run a pure data center.
In many respects, again, similar in characteristics to Arista.
Some example of what differentiates Pure, and also makes Pure, despite selling at a premium it’s equipment, the total cost of ownership less than even generic white boxes could do, and still be far more convenient w much better customer service and likely uptime as well.
Comes from ill spent weekend digging through pure stuff that was made compelling due to current valuation of stock.
Tinker
In relation to the interview I think there are some important points the CEO brings up for Pure Storage
“Magnetic is a bottleneck as the world moves to real-time analytics — creating lots of opportunities for flash to enable new applications in the analytics space.”
A thesis for investing in Pure is that they have a superior product that is necessary for the future compute mega trend. In order for deep learning and AI to achieve its potential the data that users posess has to be stored in a way that efficiently and in a way that meets the demand that these new speedy processors require. As Tinker and others have convincingly demonstrated (I believe) Pure achieves this and with the added benefit of being minimally disruptive of on going operations and being so easy to use.
“My prediction is that we will see a 50 percent decline in flash pricing in about three years and a 75 percent decline in six years”
“But the important thing to realize is that flash doesn’t actually have to be cheaper than magnetic disk for the full transition to happen — it just has to be cheap enough for folks to be able to justify the conversion in terms of saving power, space, cooling, and manpower, and for folks to understand the business upside in moving to flash. … That ROI equation already works for the vast majority of important business applications today, and we think the pivot for unstructured data applications will happen much faster than people anticipate.”
Through our discussions and research it appears likely that Flash is the future of data storage across all markets. Great news for all creators of flash. Except, is it? Anybody excited to learn that Pure’s CEO believes what they sell will be for 50% less in three years and 75% less in six years. That answers the question as to whether flash will be commoditized in the coming years. According to Pure’s CEO, you betcha.
This is where Pure’s Evergreen strategy makes me happy. The pricing reduction will mean wider adoption and a larger TAM and CAP for PURE. Why go legacy when it’s not much more to get the the new hotness that will score your better productivity and efficiency. With Evergreen, Pure in a way will be protected from and even benefited by commodization of Flash. Recurring revenue from a service contract will protect them from declining Flash prices while competitors suffer. And when it’s time to ship their customers the newest updates the flash they are providing them will be of a lower value creating an increase in value of that contract. Happy customer and happier margins.
Darth,
That is about the most insightful case made for PURE that I have heard. Putting the commoditization of flash in perspective as it applies to the primary software and service company that Pure is.
Timeliness is everything in the stock market. At present companies like Nvidia and Arista are valued at at premium, as they should be. A company like SHOP may no longer be valued at a premium, but that is another thread.
Here you have a company that is priced to stagnate for the most part, and although stock markets are usually smarter than us, the market does often use rule of thumb heuristics. And unlike many companies where we hypothesize future advantages that are not presently evidence by any evidence other than narrative, these advantages for Pure are evidenced by real world facts.
To me it seems quite compelling for a long term hold after going over it all for the weekend. Whether or not I actually put lots of money behind that statement…(I may or may not have already done so, but will make a final decision either way anyways on Monday).
Tinker
Thanks Tinker,
Sometimes you get lucky with a company and they are doing something special that the market hasn’t appreciated yet. Burt’s articles on PSTG from seaking Alpha elude to Pure being one of these. I believe he title it something like that investors aren’t appreciating Pure’s potential yet. If our thesis is correct eventually they will. Only time will tell if Pure is indeed one of these companies, but when you’re lucky enough or smart enough to find them before the market it makes for that much better returns in the long term.
I bought my first individual stock in October of 2015. I had no idea what I was doing. I used a service many of my fellow workers were using that recommended where to place our 401k in mutual funds. The company running it just started a new service for individual stocks. Of their first 10 I picked two because I knew them. The first stock I ever bought was NVDA at $28 and change. The second stock a few seconds later was ATVI at about the same price. Then a few others not worth mentioning as I didn’t own them for long and they’ve likely done not much since. How’s that luck for you, two of the best performing stocks in the entirety of the market on luck alone. The point is that if you have those two companies in your news feed any given few weeks it will immediately be apparent why those are good investments. The next installment the company said to sell those stocks because I guess the charts weren’t any good and we should replace them with other non mention worthy stocks. The market wasn’t recognizing ATVI or NVDA. YET. A few months later I found the Fool and soaked up everything and now I’ve found NPI and Sauls board. Maybe my Yahoo feed doesn’t have much on PSTG. YET. But here, people are finding info on companies with great stories.
Here is a British communication company on why their customers should choose Pure. It is a familiar rehash for those on the aforementioned boards.
“But the important thing to realize is that flash doesn’t actually have to be cheaper than magnetic disk for the full transition to happen — it just has to be cheap enough for folks to be able to justify the conversion in terms of saving power, space, cooling, and manpower, and for folks to understand the business upside in moving to flash.
I think you are missing the main point. The big sales advantage of SSD has been for years, even when it was much more expensive than rotating rust, that it was fast. As the price comes down … it isn’t necessary for SSD to be cheaper than rotating rust, merely that the premium is small enough that one doesn’t mind paying it for the speed.
What continues to bother me about the claims for Pure is that it is still networked and physics tells me that no matter what you do at the end of the cable and no matter how many parallel cables you provide, that true random access is going to be dramatically slower than SSDs which are in the box with the processor (being immensely cheaper, as well). For huge volumes of poorly structured data this may not matter, but for relational databases, it is almost certainly going to be a triumph of standard practice over performance.
In reading the interview from Pure’s CEO I figured there had to be a reason why he would be almost excited for the declining cost of his product. The obvious is that it will lead to the “cross over” to Flash array in storage. Flash is inarguably better than legacy storage. What’s more debatable is that Pure’s is the best in class. Although taken as whole, with their Flash Arrays, Flashblade, Software, and Service model they are a front runner for that title.
I’m still trying to figure how much of their business is through Evergreen. Is it everything? Is that the only way you buy Pure? They use the word Evergreen maybe all of two times in each of the few earnings call I’ve been through. None of the analysts ask questions about it. If it is a small part of the equation that is problematic to the thesis. If it is substantial that is very good news and a compelling reason to take a bigger swing at Pure. Maybe it is just that’s how they sell their product so it doesn’t need to be highlighted.
From calls
“We are winning with our highly differentiated technology, business model and customer service.”
“Cloud continues to be our strongest segment, at more than 25% of revenues and one that we believe we can increasingly take share in given our speed, our simplicity, our dev ops capable automation, and our Evergreen subscription model”
Just a question I have and can’t seem to find a definitive answer without actually trying to buy something. Maybe Evergreen is it.
Other things I find interesting. Nvidia has apparently taken a meaningful partnership with Pure. Pure claims to be to storage what Nvidia is to Compute for AI. They are mentioned as an increasing partnership in particular Flashblade and Nvidia’s DGX-1. Several Autonomous car companies (also partnered with Nvidia) are using the Flashblade.
https://www.google.com/amp/s/www.pcworld.com/article/2929632…
This program is unique on the storage industry. Costs 8 to 10% per year of the underlying storage. Controllers are software defined storage.
Example here of an investment bank and how they have 30 IT guys to support the fork lift upgrades and it is one of the worse banes in storage IT mngnt
https://www.google.com/amp/s/www.infoworld.com/article/30873…
Updated Evergreen. It is not 100% though. It is up to 25% of the new additional storage bought, thus encouraging larger incremental purchases.
Indicates such programs by other companies are for one year and largely just marketing.
Further AFA market was growing at 50% per annum and estimated to be more than $3 billion in 2018. That gives, if accurate, Pure just less than 50% market share in AFA.
A few little tidbits.
Tinker
http://mobile.enterprisestorageforum.com/storage-hardware/al…
Here are some market share and market numbers. The numbers do not add up however.
First, market estimates have gone up 3x in a year from 2015 estimates.
Second Pure is given 12% market share but that number is whoafully off or the market numbers are off. Pure will do $1 billion in sales this year w what is 12% marketshare. That puts the AFA market at around $10 billion. But the article estimates 2020 AFA to only be $9 some billion.
Thus either marketshare numbers are wrong, present total market is wrong, Pure has gained incredible market share since early 2017, or the criteria they are using is faulty or is tossing in hybrids as well.
The article estimates that 50% of all data center storage will be AFA by 2020.
Article also talks about flash becoming a more mature product by this year.
If accurate Oure is #4 vendor. But does not seem accurate or again criteria used is not just pure AFA.
But interesting. Data storage is declining but flash is expanding rapidly.
Tinker
Tamhas,
I think pure’s Flashblade is what they would counter that point with.
http://www.fujitsu.com/id/Images/8.3.3%20FAC2017Track3_Brent…
No idea if it will succeed or not.
Interesting, it shows that the processing itself is 6% more efficient, but that it is 3005+ more efficient when the data download is taken into account. I do not understand this aspect of their slides.
Anyone?
Tinker
The numbers do not add up however.
First, market estimates have gone up 3x in a year from 2015 estimatesSecond Pure is given 12% market share but that number is whoafully off or the market numbers are off. Pure will do $1 billion in sales this year w what is 12% marketshare. That puts the AFA market at around $10 billion. But the article estimates 2020 AFA to only be $9 some billion.
There are 2 sources in this article.
The 7bn market value in 2017 growing by 40% with Pure at 12% market share is from 1 source del’ollio group or whatever it is. The 2020 estimate of 9bn is by Gartner Group.
If it was 7Bn in 2017 and growing at 40% then clearly the 9bn seems out. Having said that - Gartner Group is usually the go to guys for market sizing.
Ant
Darthtaco, perhaps I am missing it, but I don’t see an answer. I get that they have done a bunch of things to make accesses within the unit have very low latency, but if the application is at the other end of a cable, I don’t see how you get rid of the inherent latency in the cable.