SolidFire v Pure

Investment Analysis Clubs Saul’s Investing Discussions
1

After great failure and possibly incompetence on the part of NTAP (with their prior management regime) NTAP gave up and for nearly $900 million bought SolidFire, which is a company very similar to Pure. Formed in 2010, and has built from the ground up flash based storage.

SolidFire, like Pure, puts out their own blogs/marketing pieces that compare their product against the competition.

This is SolidFire comparing their storage against Pure. Now this is 2014, and Pure has certainly made many strides since then. One can only assume SolidFire did as well. And $900 million is not chickenfeed, but clearly SolidFire was less than half the size of Pure, despite starting out from the same place as Pure, than Pure was a the time NTAP bought them.

But, back in 2014, this is how SolidFire compared itself to PURE:

http://docs.media.bitpipe.com/io_12x/io_121281/item_1091481/…

They identify several places, particularly at scale and redundancy, where SolidFire is superior. If accurate it would support the proposition that SolidFire more targeted larger enterprises than Pure, which focused on small to mid-size enterprises and managed to make some good in-roads to large scale enterprise. If true, this would be a good match with NTAP who has many large enterprise customers, and giving NTAP a real and built from scratch flash platform to sell.

This still creates problems for NTAP. First, NTAP’s ONTAP software is not optimized for flash, and this makes a big difference, as NTAP’s prior flash failures (and they have been failures, and many) have proven. Will NTAP support two operating systems now, or try to converge the two systems? Difficult to do either way, and if it is done will the software remain optimized for flash?

Second is sales force. NTAP’s salesforce hunts NTAP products for dinner, and switching to Fireblade is like telling a eskimo to stop hunting walrus or polar bear and do some deer hunting, or farming pigs. Both produce great food, but it is a total change of life. Will NTAP’s salesforce transition properly to a flash first world?

At some point, probably yes.

But these are challenges.

From Pure’s perspective, if SolidFire is properly integrated into NTAP’s product line and NTAP does a great job with it, Pure cannot argue it is the only large scale built from scratch flash product. The will make it more difficult to sell into NTAP accounts.

Whether or not NTAP has done this well, I do not know as yet. Just bringing up the market intelligence on what NTAP is now able to offer without having to go in the field and test for myself.

Tinker

16 Likes
2

Second is sales force. NTAP’s salesforce hunts NTAP products for dinner, and switching to Fireblade is like telling a eskimo to stop hunting walrus or polar bear and do some deer hunting, or farming pigs. Both produce great food, but it is a total change of life. Will NTAP’s salesforce transition properly to a flash first world?

Whether or not I can determine if it is pertinent to the thesis, I had to give applause to a wonderful thought.

A.J.

5 Likes
3

…NTAP’s salesforce hunts NTAP products for dinner, and switching to Fireblade is like telling a eskimo to stop hunting walrus or polar bear and do some deer hunting, or farming pigs. Both produce great food, but it is a total change of life. Will NTAP’s salesforce transition properly to a flash first world?

I know nothing about NTAP or SolidFire but have lived through a few unsuccessful acquisitions.

When a company acquires another company or technology and doesn’t integrate it well (technically or commercially) it typically leads to big discounts, especially where there are improperly set incentives, which, in this case, could help expedite the commoditization of flash and hurt Pure’s margins or win rate.

DJ

2 Likes
4

First, NTAP’s ONTAP software is not optimized for flash

Sorry, have to say this is a false statement. Mainly because your statement is too broad and encompasses every possible use case, but also because it just isn’t true. Can’t say any more than that.

Will NTAP support two operating systems now, or try to converge the two systems? Difficult to do either way, and if it is done will the software remain optimized for flash?

No point feeding the ‘one versus many being better’ debate, but I will say that SolidFire remains an independent development operation, and there don’t appear to be any immediate plans to consolidate. Nor would I expect there to be, or even need to be. The SolidFire line serves a very different market and set of use cases from the NetApp FAS/AFF and E-Series lines of business. Which, again, is why comparing a Pure box to a NetApp FAS box is pretty much equivalent to comparing an EMC Celerra to an Amazon AWS instance – sure, they both offer storage, but the use cases are wildly different.

It’s no different than suggesting that NetApp’s AltaVault operating system should be incorporated into ONTAP – that just doesn’t make sense business-wise, to me, because they are not used interchangeably from a use case perspective. (You can think of AltaVault as a “Data Domain 2.0”-type product, which NetApp acquired from Riverbed as part of the SteelStore acquisition.)

TL;DR: If you think that because NetApp’s SolidFire product line isn’t merged into ONTAP that NetApp fails, you’re probably not looking at all the pieces correctly. Same with a blanket comparison from Pure to something from any vendor that isn’t use at least mostly case-equivalent.

2 Likes
5

<<<TL;DR: If you think that because NetApp’s SolidFire product line isn’t merged into ONTAP that NetApp fails, you’re probably not looking at all the pieces correctly. Same with a blanket comparison from Pure to something from any vendor that isn’t use at least mostly case-equivalent.>>>

Prior to NTAP acquiring the company, SolidFire was less than half the size of Pure. I.e, it was losing to Pure, or at least the market segment it chose to sell into was not rewarding it as well as the one Pure was selling into. It seems to me, from looking at their blog, that SolidFire was targeting the larger enterprise sector more than the small to medium size businesses that Pure was more focused on.

However, absent NTAP’s sales force and incumbent customer base, it is likely that SolidFire will continue to lose to Pure, as it appears it was doing so prior to being acquired (less than $900 million when Pure was worth more than $3 billion) unless NTAP brings more synergies.

It will indeed help NTAP keep its customer base against Pure, but will it enable NTAP to expand beyond its customer base in competition with Pure?

I mean, the company was bought at a premium to its value (or it would not have sold, and thus comparing $900 million to $3 billion may give SolidFire too much benefit), and the company sold out because it could not thrive on its own, and perhaps not even survive on its own, or, like Pure, it would have stayed independent and gone IPO, following Pure at a much higher valuation.

As such, without looking at the product line in-depth, a lot can be ascertained by simply looking at the businesses and valuations of both companies and what lead to a low ball by out of SolidFire, at least compared to what Pure is valued at.

I therefore understanding what you are saying, but history indicates that SolidFire, on its own, is not necessarily a great competitor to Pure.

History also indicates that NTAP has attempted multiple fixes to its Flash problem, and has failed (miserably) on multiple occasions. NTAP bought the company out of desperation, and if NTAP cannot add synergies to it, it is likely the historical results will continue moving forward.

I do see the acquisition as very valuable to during up its own customer base, making it more difficult for Pure to sell into it, which is an extremely valuable thing for NTAP, and definitely will make life tougher for Pure. However, I don’t see it being that accretive to growing beyond its own customer base vs. Pure. At least from historical analysis when SolidFire was losing in the marketplace to the extent it had to give up its independence at a discount.

There is a reason that Pure (formed at about the same time as Solidfire) is still independent and Solidfire was bought at with such a discount compared to Pure’s valuation.

Anyways, I do understand how Pure is now facing more competition than it ever has in its brief history. That Pure’s product does not address all storage areas and needs, but I can also see that SolidFire is not some elixir that suddenly will cause the SolidFire product line to start winning in the market place (outside of the NTAP customer base) just because it is now under the NTAP umbrella.

Tinker

2 Likes
6

Network Appliance makes my case for me in this interview as to why the bought Solidfire:

https://www.theregister.co.uk/2016/09/01/netapp_founder_dave…

They admit that the new storage paradigm makes ONTAP engineers want to throw up.

That NTAP was losing sales, and that SolidFire is for people who hate storage and don’t want to deal with all the complications and crap that come with it. I.e, they just want it to work. And that SolidFire is selling into a “distinct” market rom what NTAP otherwise sells into.

Basically, all the same reasons that PURE indicates why you should buy their product, and none of what NTAP’s legacy products have, according to NTAP’s own words by inference, having anything to do with the new “distinct” market that NTAP can now sell into with SolidFire.

When you read the quick interview the sense you get is that NTAP still hates flash, thinks it is not “real” storage, but they were forced to acquire it or else, and figured “what do we have to lose.”

Since SolidFire was losing in the marketplace to Pure before it was acquired by NTAP, and with NTAP’s attitude towards it, I find it difficult to believe that outside of its own customer base (and who knows how heavily its sales force will focus on it even there) that SolidFire will not be more successful than it previously was.

This includes selling into the cloud, AI, and these new use cases that are growing rapidly and that are outside of the expertise of NTAP with ONTAP.

Tinker

3 Likes
7 
So, hlygrail works for NetApp, but is looking into buying shares of Pure if the share price gets just a bit lower?

[http://discussion.fool.com/will-it-not-require-emc-or-ntap-or-hp...](http://discussion.fool.com/will-it-not-require-emc-or-ntap-or-hp-to-32976377.aspx)
1 Like
8

I do not know where Hlygrail works, and if I did, it would be none of our business. I think he provided us with valuable perspective.

As I have dug deeper however, as you might get the gist from my recent NPI entries as well, NTAP really is not only not very good at Flash storage, they, culturally, really hate it. It is too simple, too untechnical, and the customers who want it hate to talk shop and get into the nitty gritty of storage. Which is a typical engineer perspective, and I don’t blame them for it. i know I hated computers moving from machine language and assembler to GUI and high level languages so that even the common man could effectively and with sophistication operate the things. I loved it when hexadecimal was a language that no one else in not just the room, but the entire institution, but myself understood. It was great.

I am not an insider, and of course cannot know, but when one looks at the history of NTAP, and why and how they bought SolidFire, one begins to learn that NTAP is competing against PURE, without any synergies other than being under the NTAP umbrella with a large installed user based and larger sales force they can provide (assuming they focus on it) with a product that Pure easily beat in the marketplace.

It also becomes apparent that NTAP seems to resent having to move into the “distinct” segment of storage with customers who just want it to work. And it is very telling that the founder, instead of talking enthusiastically about AFAs, and how they will improve the customer experience for storage solutions, says that “what do we have to lose,” now we can address both segments the market.

Really? Nothing about how this new advance in technology can improve your entire product line for all of your customers? Instead, “what do we have to lose!?!”

It begs the question, if this is coming from the top, just how much focus will the company actually put into its smaller AFA sales in lieu of its large but much slower growing legacy sales, that is engineering culture greatly enjoys better.

I will take the laser focused organization concerned with customer experience and not customers who like to look under the hood with the engineers, any day.

Tinker

5 Likes
9

They identify several places, particularly at scale and redundancy, where SolidFire is superior.

Hmm.

SolidFire uses a bit for bit redundant data copy scheme they call “Helix.” Every byte that gets written is stored on two different SSDs and they mix up the backup copies between SSDs.

Pure has a parity-based scheme they call “Raid HA.” This uses parity bits to enable reconstructing data in case of failure.

SolidFire’s approach is like an SSD tuned version of Raid 1. The main downside is that you have to buy twice as much storage as the size of the data you’re storing. The upside is that if an SSD goes down, there are copies of everything so you can continue to access the data at the same speed as before.

Pure’s approach is like an SSD tuned version of Raid 6. Dual parity bits are spread out among all the SSDs, so you can lose two SSDs units and still not lose any data. Typically, when a unit fails you replace it right away, which starts a rebuild process. The dual parity means that a second unit can fail during the rebuild process and you’re still OK.

There is a possibility that a second SSD failing in a SolidFire array could mean loss of data, since some of the data could have been written only to both units. However, SSDs typically don’t have mechanical failure, so it’s rare for an entire SSD to abruptly fail, which is unfortunately the common failure mechanism in hard drives. So, having two complete SSDs fail in an array is really unlikely.

I don’t know the technical details of Raid HA, but it appears that Pure is able to use that technology to repair single bit failures in SSDs on the fly, and even treat low performance reads as a failure, thus remapping data elsewhere in the array.

What I’ve learned is that I have much to learn about the differences between hard disk-based storage and flash storage in terms of failure protection and performance. I do worry about Pure’s always de-duped, always encrypted model for performance, but perhaps there’s something special with flash (or it could just be the large redundant NV Ram caching) going on.

As for NTAP selling flash, it all depends on how the company decides to structure commissions. They could easily put an additional bounty on SolidFire flash products if they thought that strategic.

Overall, I’m getting more intrigued with Pure than I had originally thought.

8 Likes
10

NTAP still hates flash, thinks it is not “real” storage, but they were forced to acquire it or else, and figured “what do we have to lose.”
a song heard over and over, a veritable chorus of most companies reactions when confronted by innovation.
I have no experience of mass storage but on my home devices I will never buy another HDD. Like BEV compared to ICE, SSD is just better , quicker responding, eventually more reliable.
BTW researched out that 50% of the non accident reasons to junk a car were ICE related, and that does not include automatic transmission failure.

PSTG at this stage seems less powerful and less convincing as an investment compared to ANET and NVDA. OTOH, it is earlier in development and a lot cheaper.

I get a laugh out of analysts waiting for NVDA to get back to the price where it was a year or two ago. Not going to happen. Absent a Big Bear, when other fast growers will be cheaper too. Probably cheaper than NVDA, which will still have a better CAP than most.

1 Like
11

Where I work is only relevant to the extent that the knowledge and information being shared is accurate and sound. To that end, I’ll again note that I’ve been working in the storage industry for “a long time” (you can interpret that however you want, but it’s probably longer than a majority of you have even been investing in the market).

To be clear, I’ve never suggested to anyone “hey, go buy 5000 shares of NTAP and ignore PSTG!” I have only tried to provide accurate information where misinformation was popping up, some of which may not be well-known (but none of which would be considered private or intellectual property), to help you properly evaluate the space in, around and between the two companies. Clearly, that bothers some people for some reason, but it doesn’t make it untrue.

So, hlygrail works for NetApp, but is looking into buying shares of Pure if the share price gets just a bit lower?

Only for the sake of everyone else being able to actually gain some intelligent utility from that comment, let’s assume I do work at said company. Odds are pretty high that I’d already be enrolled in that company’s employee share purchase plan, and am probably already buying those shares at a MINIMUM of 15% discount to the marketplace throughout the year. And let’s assume I don’t suck as an employee, so I might also be getting stock options or RSUs every year or two. Over time, don’t you think that might add up to a pretty big pile?

And let’s assume most reasonable employees, not to mention investors, try not to over-concentrate into their own employer’s stock? (Enron or MCI Worldcom, anyone??) Wouldn’t it be prudent to look elsewhere? And given employment history and understanding of the landscape, don’t you think there might be some equivalent ability to well-understand what other vendors might be doing and where the opportunity might exist?

And if that isn’t clear enough, my personal goals dictate largely where I invest. A potential investment IN one company is not necessarily a vote AGAINST another – except for some ethical rules against companies that willfully hurt people, I will invest anywhere I think I can beat the market because that helps me achieve my goals.

Hope that makes understanding the information I’ve shared very clear, but for anyone who can’t read properly, a TL;DR summary: I would recommend buying both NTAP and PSTG if you’re a long-term (5+ years) investor.

14 Likes
12

Where I work is only relevant to the extent that the knowledge and information being shared is accurate and sound. To that end, I’ll again note that I’ve been working in the storage industry for “a long time” (you can interpret that however you want, but it’s probably longer than a majority of you have even been investing in the market).

I would recommend buying both NTAP and PSTG if you’re a long-term (5+ years) investor.

I didn’t intend to imply anything negative by my attempts to piece together where your perspectives are coming from, and I appreciate the additional insight you’ve provided.

I was guessing there might be some instances where investing in a direct competitor might be frowned upon by a present employers, but if there are no restrictions associated with that, using knowledge of the industry seems like a good idea.

I’ll try to take a closer look at NTAP, since I haven’t previously.

1 Like