Straight from the horse's mouth: It's October 7 for Pure's IPO Pure Storage has told us something quite interesting: "Pure Storage will be listing shares for our initial public offering at the New York Stock Exchange on Wednesday, 7 October." One of the most eagerly awaited initial public offerings will take place on Wednesday next week with several companies anxious to see how well the …
This post has been deleted by its author
It is not differentiated. Yet more commodity crap with code wrapped around it. Not different enough from any other ssd array out there that compresses or dedupes. Performance is as good as the rest of the ssd arrays. Price is the same. Looking at their balance sheet damn they are spending a lot of money. It was the same with violin- trying to scale up the business costs a lot. Except violin also had the costs of making hardware from scratch. Violin didnt have the competition at that time though. Pure does have competition and a lot of it. Every customer that I speak with who is evaluating flash is testing tech from 3 or more vendors. And if I were a customer I would buy my flash from a big company not a startup that's begging for money. HP's, IBMs, EMCs, HDS flash is all decent.
Annoyingly a huge part of pure's pitch is the fuzzy math behind $/GB. That is a sales strategy that allows them to price on potential usable post-dedupe which means budgetary estimates are wacky. Many customers confide in me that they don't like that.
Personally I also dont like their always on compress/dedupe. That adds a big latency bump that they compensate for with DRAM... just like a spinning disk array. Sigh. Whats the point of an architecture that places cache in front of slow ssd enclosed flash? EMC DSSD and other SSDless arrays are the real future and pures ssd architecture will soon be obsolete. The flash chips cost the same in or out of an ssd- so why slow your flash down? Because of $/GB? The business case for all flash primary storage is now based on TCO but theres enough flash tech in the pipe where the cost difference will be negligible.
Anyways, boring tech in boring old storage in boring old data center infrastructure. Yawn. Their code and nice logos slapped on commodity harware is not worth a billion bucks. Now lets say however they had developed Spark... Well they wouldn't have. Pure obviously doesn't have the braintrust. Just business degrees at work here folks.
It seams like you never used their product and you don't understand how it works (at least you got it wrong) So why are you commenting?
Pure only uses DRAM cache because of a NAND flash problem called read disturb, where if you read the same cell many times you get errors. So Pure uses a very small amount of the controllers DRAM to cache the very highly used data. This is done not to increase performance, but to increase lifetime of the flash media.
Most of the controllers DRAM is used for meta data.
"Pure obviously doesn't have the braintrust. Just business degrees at work here folks" - which company does have the braintrust then and how are their people different from pure's?
Also if "you" invest in pure at the IPO and think they are going to make you a lot of money within the next 2-3 years, then you have not done your homework. Pure is spending a lot of money on RD and building out the company, they need to do so, if they don't then they are dead in the water.
Moving customers away from their incumbent storage vendor is very hard. In the early days, low latency and all of the other flash benefits eased the process. Now every major league player has a flash offering, albeit many of them suck, so Pure is up against EMC on (what customers see as) an even playing field and the default decision is don't change vendor. It's going to be hard.
16-32GB of ram is not negligible. Who cares about read distrub. It is not an issue with enterprise all flash arrays.
Which companies have the braintrust? Frankly I think most of the brains in tech are in AI and cloudscale/distributed computing. If you are only talking about infrastructure though well it is just a commodity with negligible differences between the options.
2-3years. So medium term investment?? In a startup with lots of competition?? That has to spend gobs of cash to even keep afloat?? That is competing with EMC AND other tier1 incumbent vendors that have equivalent products??? And you are telling to fund R&D and sales operations in the hope that Pure will successfully grow the business and refine their commodity technology year after year??
Yah I am not going to buy. I don't think the street will either but who knows I'm obviously not a high frequency trading algorithm.
Or am I? ;)
It is very simple since I work in the space and I have competed against Pure about 100 times over the last few years. This IPO will easily halve in market cap 6-9 months out just like NMBL did. I am not going to get into the why here. I am a very versed trader (third generation trader, with 25 years of trading experience). I could get into the why for a long time but I rather get into the "what" and save the non-finance/non-market readers a lot of time.
Here is what I am going to do with $50,000.
About 1 month out, PSTG will become marginable. Shortly after, smart money will create a huge short inventory. 3-5 months out from the IPO date, you will see +20% of the float short if available. I plan on getting more greedy given this is the short of a lifetime. About 6 weeks after the IPO, I am going to buy out of the money puts, 40-60% out of the money depending on how the IPO goes the first 6 weeks. I will buy some contracts dated 6 months out and other contracts about 9 months out, so I can get 2-3 reports in during my contract length. Yep, $50,000 of out of the money puts, about as risky of an investment you can make.That is how much I believe PSTG will be a poor IPO due to its undeserved valuation and bleed-cash-for-growth business model. It will struggle to maintain a $1bil market cap 9 months out as a public company that cannot behave as it behaved to get to this point.
It is all about sustaining revenue growth for NMBL. If NMBL cannot continue growth, and show their FC AFA can continue to penetrate enterprise storage, then the current $1.9bil valuation may be very rich.
That is the Pure Storage issue. Their growth is flattening. They had hyper-growth in mid-market VDI but that use case is merely a fraction of Enterprise storage, so that growth is not sustainable. Pure's win rate is decelerating as well. They have successfully dragged traditional storage reps (who work at their competition) into their conversation, and those traditional storage reps now have to drag their flash peer into the deal to compete against Pure, and many of them have better arrays than Pure. Yep. That is what the street will not understand. Pure's success in VDI has forced the Big 6/7 to play that game, and they are playing it more often, and they are beating Pure more often now that they are forced to play the game (versus selling the same old storage equipment that Pure would often win against). Pure does not have agnostic middle technologies that talk to other data sets managed by other vendors. They have to partner with companies to address the "we are not another data silo" issue often raised by storage buyers. And that dependency does not look strong to many storage buyers who are conditioned to buy from the incumbent, who by the way, have these technologies. They might meet with Pure and POC Pure, but they will meet with EMC, IBM, HP, Hitachi, Dell, etc and often not buy from Pure. Companies like EMC will also give away xtremeIO for free or peanuts, or let the box sit around for 6-12 months during the POC to derail Pure, assuming no urgent compelling event.
Pure also has another 20 new entrants in the AFA space that they are forced to compete with, apart from the Big 6/7. And some of these have similar data reduction technologies. So, you see, the game has changed. They cannot just mop up in every mid market VDI deal like they did in 2012-14. And thus, they will have flatter growth and reports showing they are still burning cash.
Unlike NMBL thought, their IPO may not have an initial skyrocket for the first few months. NMBL was an easy short if you know the space. And if you know the space well, $1.9bil could still be expensive if they do not continue to execute over the next few quarters. When NMBL gets cheap enough, NTAP will likely want to to buy them because they have won over a lot of NTAP customers and NTAP needs to stop the bleed. I use the "want" versus "should" or "will be able to".
I shorted NMBL at the right time. PSTG is next.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
