Pure Storage Inc (PSTG) 2019 Q1 法說會逐字稿

Good afternoon.

My name is Chris, and I will be your conference operator today.

At this time, I would like to welcome everyone to the Pure Storage Q1 Fiscal 2019 Earnings Conference Call.

(Operator Instructions) Thank you.

Matt Danziger, Head of Investor Relations, you may begin your conference.

Thank you, and good afternoon.

Welcome to the Pure Storage Q1 Fiscal 2019 Earnings Conference Call.

Joining me today are our CEO, Charlie Giancarlo; our CFO, Tim Riitters; our President, David Hatfield; and our VP of Tech Strategy, Matt Kixmoeller.

Before we begin, I would like to remind you that during this call, management will make forward-looking statements, which are subject to various risks and uncertainties.

These include statements regarding: competitive, industry and technology trends; our strategy, positioning and opportunity; our current and future products; business and operations, including our operating model; growth prospects; and revenue and margin guidance for future periods.

Any forward looking statements that we make are based on assumptions as of today, and we undertake no obligation to update them.

Our actual results may differ materially from the results predicted, and reported results should not be considered as an indication of future performance.

A discussion of risks and uncertainties relating to our business is contained in our filings with the SEC, and we refer you to these public filings.

During this call, we will discuss non-GAAP measures in talking about the company's performance, and reconciliations to the most directly comparable GAAP measures are provided in our earnings press release and slides.

Also, beginning this quarter, we adopted the new revenue accounting standard, ASC 606, and all of the current quarter financial results, our financial outlook and historical results and comparison to historical results are stated in accordance with this new standard.

Please see our earnings slides for more information.

This call is being broadcast live on the Pure Storage Investor Relations website and is being recorded for playback purposes.

An archive of the webcast will be available on the IR website for at least 45 days and is the property of Pure Storage.

With that, I'll turn the call over to our CEO, Charlie Giancarlo.

Thank you, Matt.

Good afternoon, everyone, and thanks for joining us on today's earnings call.

Q1 was a strong quarter for Pure, and our fiscal year is off to a good start.

I will begin the call with an overall summary of our first quarter results and provide the market context that is guiding our innovation strategy.

Hat will then provide a go-to-market and partner update.

And finally, Tim will give a detailed review of our financials and our outlook.

Revenue for the quarter was $256 million, up 40% when compared to the same period a year ago.

Gross margins were 66.3%, and operating margins were negative 6%.

Not only did we exceed both our revenue and profit guidance ranges, we also achieved another quarter of positive free cash generation.

At Pure, we are driven by our desire to help our customers achieve more with their data.

This focus, in combination with our industry leading Evergreen business model, our product innovation and our focus on our customer success, is the foundation of everything we do.

It has served us well enabling Pure to be a leader in the market adoption of important new technologies empowering us to shape new data center architectures.

Over the last several decades, data centers were built around monolithic application stacks supporting large application environments and relatively small amounts of data.

Think about your PC 15 years ago where 1/3 or more of your hard disk was consumed by the operating system and the applications and only a tiny amount by your data.

The same was true in data centers.

Virtualization freed the software from the hardware but large specialized application stacks remained dominant.

Today, the exact opposite is true.

Modern distributed application environments are created, scaled and reduced in VMs and containers on the fly across the data center and even between clouds.

Datasets, on the other hand, have grown exponentially, making it extremely difficult to copy, move and safeguard them.

This fundamental change in the scale of apps and data in combination with advances in new technologies like GPUs, flash storage and high-speed networks are enabling new data-centric architectures that reduce complexity, increase performance and reliability while decreasing costs.

Public clouds already leverage data-centric architectures at massive scale.

They have learned that optimized storage compute and fast converged networks create ultimate flexibility to offer the range of services that they provide.

Pure has been at the forefront of democratizing this approach just as we've done with all flash and NVMe so that private and hybrid clouds can do the same as public clouds.

One example in Q1 that highlights the approach that customers are beginning to adopt is a leading SaaS company.

Their systems built entirely on server-based direct attached storage infrastructure ran head first into the challenges of this design: expenses and isolated data islands, high complexity and spiraling costs.

They made a multi-million-dollar first purchase from Pure to implement a data-centric architecture moving off of direct attached storage.

After evaluating multiple vendors, they chose Pure for our ease of use, active cluster technology, container support and the ability to provide six nines of uptime, allowing them to meet their service level agreements with their customers and free their engineers from time-consuming capacity and growth planning.

Our latest NVMe innovations allow us to supplant and improve upon traditional direct attached storage in both traditional and modern workloads, something we'll talk more about at our Pure//Accelerate User Conference, which starts tomorrow.

We believe that the direct attached storage market represents a significant opportunity to Pure.

The scale-out approach of shared storage is beginning to show early signs of adoption by enterprises, and Pure's technology is helping customers to embrace the strategy.

We continue to be a pioneer of innovation.

We were ahead of the market on all flash.

We were ahead of the market on NVMe, and we're doing it again with direct attached storage replacements and AI.

We're looking forward to sharing even more of our innovations with our customers, partners and analysts at our user conference, Accelerate.

And with that, I'll turn the call over to Hat.

Thanks, Charlie.

Our first quarter indeed got off to a tremendous start.

Q1 showcased the strength in our go-to-market strategy as our teams executed well across 3 key dimensions: first, in acquiring new and expanding our footprint with our customers on a global scale; second, we deepened our alliance and channel partnerships; and third, we continued to demonstrate that we have a highly differentiated technology platform that provides competitive advantage to our customers in this data-centric world.

The combination of these factors drove continued momentum in the first quarter.

Customer growth was strong.

In the quarter, we added nearly 300 new customers, up approximately 45% from last year.

As we've indicated in the past, our focus is continuing to drive upmarket, and we were very pleased with our growth in the Global 2000 largest government and health care organizations and the top 1000 cloud companies.

Repurchase rates remained steady and predictable with approximately 70% of our business coming from existing customers, and win rates against competitors remained strong year-over-year.

We also saw notable contributions from both EMEA and APJ with each significantly outpacing the overall company growth rate.

In conjunction with our solid sales execution, we strengthened and expanded key alliance and channel partnerships across our go-to-market ecosystem this quarter.

First, our partnership with NVIDIA continues to thrive in the field and at a corporate level.

In the quarter, we announced our joint offering, AIRI, the industry's first AI-ready converged infrastructure solution.

AIRI enables customers to be up and running with AI and deep learning applications in hours compared to the weeks and months required to build and scale alterative solutions that had a 10x larger physical footprint and offer 1/10 the performance density.

AIRI delivers better performance and will save money compared to the public cloud.

Let me share one customer example from the quarter that has taken this leap into AI and machine learning.

PAIGE.

AI, one of our first AIRI customers, is enabling pathologists to be more efficient, researchers to be more quantitative and patients to be more confident with cancer diagnosis.

Pathology is key in cancer diagnosis, and the majority of this work is typically done manually with methods developed more than a century ago.

With access to one of the largest tumor data archives, PAIGE.

AI turned to Pure and NVIDIA to deploy a deep learning infrastructure that removed the challenging bottleneck of ingesting and analyzing millions of images into the AI system.

With the data bottleneck removed, the system's potential is furthered, facilitating more accurate and timely diagnosis and subsequent treatment.

The market for AI is showing promise and interest from enterprises across industries.

Recent analysis from IDC estimates that storage for AI workloads will be a $10 billion market by 2022, representing a 5-year CAGR of more than 35%.

Gartner also estimates that 80% of enterprises will deploy AI by 2020.

We are excited about the opportunities ahead in this area with our highly differentiated capabilities, enabling our customers to derive more insights for more of their data.

In addition to NVIDIA, or partnership with Cisco continues to demonstrate outstanding momentum.

Together, we believe that our FlashStack solution is the fastest growing converged infrastructure offering in the market.

Activity with the Cisco sales teams and our joint partners continues to be robust, and we were showcased at their annual Partner Connection conference last month.

We demonstrated how FlashBlade enhances performance for Cisco and NVIDIA deep learning neural networks as well as for Oracle data warehousing.

These demonstrations, in addition to some exciting announcements at Accelerate, illustrate how Cisco and Pure continue our commitment to create solutions for modernized multi-cloud customer requirements.

In summary, I have never been more thrilled about the momentum in our business.

Since the first generation of FlashArray and the introduction of our unique software platform, we've offered a fundamentally different business model and customer experience with our Evergreen Storage subscription.

Enterprises across the globe have taken advantage of Evergreen to protect and future proof their investments without having to perform disruptive and risky tech refreshes as their data continues to explode.

We've already delivered multiple generations of performance improvement, storage density and software nondisruptively, and in 2018, we will reset the bar for all others yet again as we extend FlashArray//X across our portfolio.

This land-and-expand business model and subscription to innovation combined with our software platform makes it easy for customers to remain with Pure, which we've seen reflected in our highly predictable repeat purchase rates.

Stay tuned as we will be making many new announcements at Accelerate on our products and services.

2018 is shaping up to be our best year yet, and we're only just getting started.

With that, I'll now turn the call over to Tim.

Thanks, Hat.

Q1 was a great start to the year for Pure as we exceeded both our revenue and operating margin guidance.

Before I dive into the specifics, I'll make my usual note that the gross margin, operating margin, OpEx, net income and free cash flow numbers I will use are non-GAAP unless otherwise noted.

Reconciliations of these non-GAAP metrics to their GAAP comparables as well as our full Q1 results and presentation are available on our website at investor.purestorage.com.

Additionally, as a reminder, our results and growth rates on this call are under the new revenue recognition standard, ASC 606.

For comparability, we have updated all historical periods in our investor deck and provided separate schedules and reconciliations in the appendix to help reconcile between 606 and 605 for historical periods.

Total revenue in Q1 exceeded our guidance range and grew 40% year-over-year to $255.9 million.

Product revenue grew 37% year-over-year to $195.4 million, and support subscription revenue grew 52% year-over-year to $60.5 million.

Geographically, 72% of sales came from the United States and 28% came from international markets.

One of our objectives we highlighted at the beginning of the year was to grow our international presence, and we did that this quarter, increasing our international revenue to 28% from 21% in the year-ago quarter.

Q1 total gross margins were 66.3%, coming in at the high end of our guided range and stable with last quarter.

We've now been operating within our target model range of between 63% and 68% for 10 consecutive quarters.

Product gross margins decreased 0.2 point quarter-on-quarter at 66.3%, driven by higher indirect COGS as a percentage of revenue given Q1's seasonal dynamics.

Support subscription gross margins were 66.3%, up 0.9% sequentially, driven by a continued increase in amortization of ongoing support contracts as a result of our larger installed base, continued solid execution in our support organizations and timing of renewal bookings during the quarter.

Turning to operating margins.

Our operating loss in the quarter was negative 6% or negative $15.3 million and compares favorably to the negative 13.9% or negative $25.3 million in the year-ago quarter.

This represents an 8 point improvement from last year and a 5 point improvement over the midpoint of our guidance.

This notable overperformance was driven by a combination of factors, including: number one, a strong performance in our top line; number two, better-than-expected product gross margins; and number three, modestly lower-than-expected operating expenses.

While we have consistently driven both growth and operating leverage in the business and we expect these trends to continue, we would not anticipate similar outperformance on operating margin next quarter.

Our net loss for the quarter was negative $16.2 million or negative $0.07 per share.

This compared to negative $24.3 million or negative $0.12 per share in the year-ago period.

The weighted average shares used for the per share calculations were 224 million in Q1 and 206 million for the year-ago period.

Total headcount at the end of the period was more than 2,300 employees compared to more than 2,100 employees as of our prior quarter-end and up more than 500 over the same time a year ago.

Turning to the balance sheet.

We ended Q1 with cash and investments of $1.1 billion, an increase of $500 million mainly driven by cash proceeds from our convertible debt issuance this quarter, plus free cash flow generation from the business.

Also note that we also repurchased $20 million of our common stock in conjunction with this debt offering.

Free cash flow was positive $8.6 million for the quarter after removing the impact of our employee stock purchase plan.

We are pleased to deliver another quarter of positive free cash flow generation, our first Q1 ever of positive free cash flow for Pure.

Had we included the impact from ESPP, our free cash flow would have been negative $3.7 million.

As we have previously noted, going into Q2, we will continue to observe seasonal dynamics both in our P&L and in our cash flow.

On the P&L side, we expect a notable increase in revenue and continued investment in our business, including our Accelerate user conference.

Secondly, as it relates to cash flow, we expect Q2 to be the lowest cash generation quarter in our business consistent with prior years as cash collections tend to be seasonally low following our seasonally low Q1 revenue quarter.

With that, let's turn to our guidance.

For our second quarter, we expect revenues in the range of between $296 million and $304 million, a $300 million midpoint; non-GAAP gross margins in the range of between 63.5% and 66.5%; and non-GAAP operating margin in the range of between negative 7% and negative 3%.

For the full year, we expect revenues in the range of between $1.320 billion and $1.370 billion, $1.345 billion at the midpoint representing a $10 million raise at the midpoint from the guide we provided during our last earnings call; non-GAAP gross margin in the range of between 63.5% and 66.5%; and non-GAAP operating margin in the range of between 0 and positive 4%.

In closing, we are pleased with the strong start to our year and look forward to sharing more information on the road ahead, including some exciting product developments with you at our Accelerate event later this week.

With that, we'll now open the call for questions.

Operator?

(Operator Instructions) Your first question comes from the line of Wamsi Mohan with Bank of America Merrill Lynch.

I was wondering around the guidance.

Your Q2 guide at the midpoint implies 17% quarter-on-quarter growth.

Last year's seasonality was quite a bit stronger.

Just wondering how we should think of seasonality as we go through the course of this year, and I have a follow-up.

Yes.

Wamsi, this is Tim.

In terms of growth, now that we're in this sort of nice solid 30% plus growing, we're looking more at the year-on-year compares, and so the midpoint that we offered up here in the guidance was 34% at the mid by my calculations.

That's what I would be thinking about as you think about your modeling going forward, is sort of model it off a year-on-year basis for the next several quarters if that makes sense.

And so would you expect that like second half versus first half growth on a year-on-year basis to be relatively consistent?

Relatively consistent, it'll be down a little bit.

Obviously, we just put up a 40%, a very, very strong Q1 performance.

And so just by math, Q2 will -- or, sorry, second half will be a little bit lower, but I think that's how I would think about modeling it.

And the other question I had was on the competitive side.

Have you seen any changes on the competitive side both from HPE, from NetApp, and from EMC?

And could you just comment on any impact that you have baked in from NAND price declines in your revenue guide?

This is Hat.

So first, we haven't seen anything really discernible in terms of a difference in the competitive landscape.

We are pleased with our performance.

Recent IDC market share stats have us outpacing the market by 3x.

So I think that's a great metric.

Win rates continue to be strong [against] all the competitors, and we believe that our innovation gap is widening.

There's a bunch of announcements that we'll share later this week that we feel just continues to expand the moat.

On the NAND, do you want to add -- hit that one?

Yes.

On the NAND, Wamsi, what we've historically said is that both a tight NAND supply and a sort of more loose NAND supply both help us from a differentiation perspective.

Without going into all the details, it really is software at its core that allows us to mix and match and use a variety of different NAND types and get more benefits through data reduction and deduplication and compression.

And so again, whether we're in a tight market or a loose market, we are always going to sort of be able to drive better gross margins, which is what you saw in the results again today.

Your next question comes from the line of Sherri Scribner with Deutsche Bank.

You mentioned you outpaced the market by 3x this quarter based on what you've seen.

I assume that's not 3x the all-flash array market.

What is your sense of the growth in the all-flash array market?

And how much do you think you outgrew the all-flash array market this quarter?

Well, look, I think the most interesting, most recent compare was the IDC Q4 numbers that came in and looking at the whole year for last year, and some interesting dynamics, I think, happened in Q4.

If you look in general both NetApp, EMC and HP all struggled in Q4, and we turned in a very strong quarter.

So I think the full year 2017 results are in, and I think we're feeling very good about the competitive compares.

Okay.

And then, Tim, I guess, if you look at the sales and marketing line, that number went down pretty significantly, at least more than I would have thought it would, but it seems like your guidance is implying that, that ticks back up again.

Can you maybe talk about the dynamics there?

And is that the right way to think about it?

Yes.

On the S&M line, so the overall dollar amount did go down.

And really, obviously, we've continued to add sales reps but the commission expense obviously follows the overall sequential decline in revenue, so Q4 tends to be a very high commission expense for us and then Q1 correspondingly lower, and that'll start climbing up through the course of the year.

And then we reminded folks in our prepared remarks, we have sort of big chunks of marketing spend throughout the year, most notably here coming up in this quarter, our Accelerate.

And so that's why you see S&M being relatively low but then climbing through the year again.

Your next question comes from the line of Mark Moskowitz with Barclays.

Two questions.

First, beyond conservatism, can you help us understand what could be some of the puts and takes in terms of why the current momentum is not lending to any sort of upside or rates, guidance for the full year?

Are there -- are those the new products or new go to market that could result in elongating sales cycles because of what's going to happen later this week at Accelerate?

So Mark, this is Tim.

First thing I would say is we did, in fact, raise our guidance for the year $10 million at the mid relative to what we offered up about 90 days ago.

So I think it really does speak to the momentum that we're seeing in the go-to-market engine as well as the innovation that we're seeing in the business as well.

So again, I would suggest that what you saw with our results today actually suggests that we are seeing a nice confidence and nice growing in the business with that raise.

Okay.

And then as far as the cloud, it is a significant piece of the revenue pie for Pure currently.

NetApp has made some pretty good inroads with both Azure and Amazon.

We get questions a lot from investors how we should think about Pure in terms of [tactically].

What are the opportunities ahead for Pure in terms of incremental penetration so that you can also be offering powered by Pure in the big cloud vendors?

Or is this still an area where you got to make up ground?

Yes.

Well, this is Charlie.

I think the most exciting aspect of our cloud business is over 30% of our business now is 2 cloud providers across the spectrum, SaaS, IaaS, PaaS, and that's been growing for us.

That's a very big market for us.

We're very excited about it.

So when the cloud vendors themselves are using Pure, that signals a lot to, I think, our private data center customers as well.

We are doing more and more to integrate with the cloud.

You should come to Accelerate to see what we're doing there, what we're announcing over the next few days.

But we're -- we feel very good about the cloud powering our business.

Also just add on from a technical point of view, if you look at the all-NVMe movement, going after DAS workloads and some of the wins that we highlighted in the overall earnings call today, we're just seeing a great uptick in some of the next-gen use cases within cloud providers as well.

So it's not only about going after those growing customers.

We're making sure we're aligned to the net new applications that are driving their growth.

Your next question comes from the line of Aaron Rakers with Wells Fargo.

This is Joe Quatrochi on for Aaron.

I want to dive a little bit deeper into the NVIDIA alignment.

So when NVIDIA reports like 550 DGX customers exiting 2017, how do we think about Pure's penetration into that customer base or also the opportunity to gain as its customers?

Joe, this is Hat.

So we're super excited about AI in general.

You can hear the enthusiasm, hopefully, in some of the successes that we outlined in the prepared comments with AI and PAIGE.

AI.

But we're seeing the expansion of it beyond self-driving cars and social media.

We're seeing it across retail and health care and finance as well.

And so the momentum that NVIDIA clearly has in their business -- I think their data center business grew from $800 million to $1.9 billion over the last fiscal year.

We see that as a great opportunity to work together in the field, and I think a lot of activities that we've had from a sales-to-sales perspective continue.

But I think we augmented that with our launch at their user conference.

I think there were 6,000 people where we jointly launch the AIRI solution, and Charlie was on the road with their GM as well.

Maybe you can comment on that.

Yes.

I was on the road with Jim McHugh from NVIDIA, who's the GM for the DGX business as well as Jayshree Ullal, CEO of Arista, and we went across the country.

I have to say, we -- the number of Fortune 500 logos and the titles within those logos was really almost unprecedented for us so very, very high level interest and a very high level of interest in AI.

I think we sported a statistic -- or rather, a study by Gartner that indicated that 80% of large -- of their customers are going to have AI projects by 2020, live AI projects.

So it clearly is garnering a lot of attention.

I don't know that we have fully accurate attach rates to DGX.

What I can tell you is that the NVIDIA team believes that they get the best out of their product when it's attached to Pure Storage, and they are -- and when we go into accounts, whether it's together or separately, that's generally what our customers see as well.

So we're confident in that partnership both with respect to the quality of the high-level touch point but more importantly, the level of cooperation in the field.

And then just a follow-up.

I was kind of curious thinking these past few weeks, you've seen a couple different NVMe-based product launches most recently from NetApp.

I know you guys put out kind of your thoughts on EMC.

I was just curious, what are your thoughts on NetApp's NVMe solution?

Yes.

This is Kix.

I'll take that one.

Look, so many ways this kind of reminds me of all-flash arrays in 2014.

Pure was out early innovating the concept, but once the big guys kind of finally came onboard with their 1.0 offerings, it really told everyday mainstream customers it was time to jump into the new technology.

And so it feels like that same movie all over again with NVMe where we were out years ahead innovating with NVMe in our products since FlashArray//M and on.

And now we've had the big guys come in and bring out their first-gen retrofit solutions, and we think it's going to do a great service to us to just tell the whole market you'd be crazy to buy an FA in 2018 that doesn't have NVMe built in.

Your next question comes from the line of Katy Huberty with Morgan Stanley.

Do you have any metrics around what percentage of customers or new shipments are going into AI type of workloads and where you would expect that metric to trend over time?

Katy, this is Hat.

So we don't have any specific metrics on attached.

I would say that, over the last several quarters, we've gotten really crisp in the use cases that we have for FlashBlade.

AI is clearly one next-generation analytics as well as the rapid restore.

We saw really great momentum in that use case, particularly in the cloud segment, where we're just placing direct attached infrastructures as well.

So no specific attach rates, but we think we have 3 or 4 really repeatable use cases for FlashBlade and starting to see some great traction on [each].

Yes.

And clearly, the other thing I would say, Katy, is, clearly, AI is one of the bigger chunks of our FlashBlade business, and so you think about it in terms of the FlashBlade business being very successful.

A lot of that's come from the AI side.

Okay.

And you made more comments today around the opportunity in the DAS market.

Is that new addressable market for you?

Or was that included in your long-term guidance for 30% growth a year ago?

This is Kix.

I'll take that one.

We included a small amount of DAS when we -- in kind of our TAM, I guess, 2 years ago now.

But as we start to see X become more mainstream and NVMe take off, we think more and more of that DAS market is becoming in play.

And so it's an opportunity we're increasingly excited about.

Okay.

And just finally, what about penetration of NVMe into FlashArray?

I think last quarter, you talked about 20%?

Where was that in the April quarter?

Right.

So we were giving some -- as we will do occasionally, just some color on a statistic.

We did mention 20%, which -- as of the end of the last fiscal year.

We did say at that time that we were -- it was going to be a majority of our product shipments by the end of the year, and we are well on track to accomplish that.

And it certainly was up quarter-on-quarter, Katy, so good momentum in the business.

Your next question comes from the line of Rod Hall with Goldman Sachs.

This is RK on behalf of Rod.

I just wanted to follow up on a previous question in terms of the new NVMe products released by your competitors.

What do you see as your competitive advantage?

Is it just performance?

Or should we be thinking about something else?

Yes.

This is Kix.

I'll take that one.

I thought the messaging that EMC came out with Power Macs, in particular, was very indicative of the mindset.

They called it a Tier 0 array, and you saw NetApp followed suit and come out with a very high-end 800 array.

Both those vendors seem to be making the same mistake as flash the first go round, where it was viewed as this exotic, niche, high-end technology.

And I think we said all along, it's all about mainstreaming.

For us, NVMe is just a wire.

We've done a lot more with our direct flash architecture to be in -- interfacing our software directly with raw NAND, so we can get dramatically higher performance but also dramatically higher levels of efficiency.

And I think that's going to be very interesting in this market over this year because there's still a pretty big premium to dual-port NVMe drives out there.

And so we obviously don't have to take advantage of those.

We build all our own modules from scratch, and so that puts us in a position to really help drive the mainstream adoption of this technology.

Okay, great.

Could you also comment on what kind of trends you're seeing in the flash -- for your FlashBlade business this quarter?

Yes.

This is Hat.

So I'll take that.

Continued progress, like I said, on repeatable use cases, so we love to see the momentum there.

I think more broadly, just the portfolio and platform selling motion that we've been working on over the last couple years is really bearing fruit.

We're seeing larger deals and more of them come through, and we're seeing great insertion points into enterprises where you don't have to go necessarily directly to IP.

You can go to data sciences and DevOps, get a new use case and expand from there.

So we're seeing nice traction and pull-through on both the FlashArray business and the FlashBlade business.

We're really training our sales folks to go find the opportunity that they get on the floor the quickest and then expand from there.

And we know when we look at things holistically, the repeat purchase rates really kick in.

That's helpful.

One final question from me.

Could you give us any color on why you renewed the $500 million in convertible debt?

Yes.

I'll take that.

First of all, an old friend once told me when money's on sale, take it.

And it was a good time to raise money.

Convertible debt is a popular item, and it was a good time to raise.

And so of course, you always want to raise money when you don't need it rather than when you need it.

But secondly, we continue to invest in this business.

So last quarter, the growth rate was 40%.

We still -- we foresee good growth ahead, and for that, we need to continue to invest in the business both organically and inorganically.

And our strategy is to build, partner or buy.

And so we wanted to be prepared for all of those different things, and it seemed to be a good time to raise money.

Your next question comes from the line of Alex Kurtz with KeyBanc Capital Markets.

Just 2 questions here.

So on the U.S. federal side of the business, I know that you guys made some investments last year to get that organization up and running.

We're going to be in full swing in that vertical's spending patterns in a few months.

So just how do you think about that vertical as far as contributing to growth this year?

Is it really baked into the full year number?

Do you see it as more of an upside?

And then, Tim, on Evergreen, just maybe an update on how we should be thinking about how to model this as it becomes, obviously, a very popular part of your overall program.

You guys -- I think services was down sequentially, so I just want to better understand how we should be modeling that going forward.

Thanks, Alex.

This is Hat.

I'll take the first one.

The investments that we made in the public sector and Fed specifically are really starting to produce some really nice wins.

In fact, we announced the Department of Energy win as part of our press release today, and we'll see continued benefits and traction, I think, in that industry vertical.

It is baked into our assumptions.

We have a very predictable model as sales reps ramp as pipelines build and what our close ratios are for those.

So we see great momentum there and continued momentum, but it is baked into the guide.

And then, Alex, on the question on Evergreen, you'd mentioned that support was down sequentially.

Are you talking about revenue?

Or are you talking about margin?

Maybe I didn't update it for the quarter yet, but I just -- I mean, maybe it looked flattish to me sequentially from January to April on revenues.

Yes.

On a revenue basis, it's up.

You may want to take a look -- you might be looking at historical 605 versions because, remember, they all reset from your models.

So 606 basis, it's still gone up quarter-on-quarter very nicely.

Business is performing very well as we continue to amortize those Evergreen businesses into our P&L.

Your next question comes from the line of Andrew Nowinski with Piper Jaffray.

So it's sort of a follow-up, but I guess, as it relates to your ability to support private and hybrid cloud, I was hoping you could elaborate on that, and then, really, your intentions for separating software from hardware, enabling customers to manage their data no matter where it resides, whether it's on premise or in the public cloud, kind of similar to what NetApp did with their ONTAP solution.

Yes.

This is Kix.

I'll take that one.

So first off, goal #1 for us is to make sure we have deep support for every cloud operating platform out there, cloud delivery platform out there.

And so we have deep integrations across all the classic VMware, Microsoft, Red Hat, et cetera, platforms so that we can easily tie in and automate within people's hybrid cloud environments.

And then last year at our Accelerate conference, we announced our CloudSnap integrations, where we're working on integrating directly with the public cloud.

You can expect an update on that and a wide range of other stuff at Accelerate this week.

And then I -- just another follow-up, I guess, on the convertible debt offering and your comments you made there.

Are there any areas within your portfolio that you feel like you need to own, areas like backup or data management where you're currently partnering with?

Or is it something that you would look into a completely new area that you don't currently have either a partnership established with?

Yes.

I -- let me answer that, but I'll do so a bit indirectly.

If you write to me, I'll ship you a list of our intended acquisition targets.

I'm teasing of course.

We have a variety of new opportunities that we're looking at.

All of them are adjacencies, so close to what we do today, I believe very strongly in making sure that anything we acquire fits well with both our internal expertise as well as our external selling capability and the customers that we interface to.

And so that's what you -- you can expect us to focus on those areas that make sense in that environment.

And I'd just add, again, back to Charlie's motto of build, partner and buy.

This has really been an MO that we've been following for quite some time now.

So nothing that was unusual now versus any time in our history.

We've always been taking a look at and scanning the environment.

Your next question comes from the line of Steve Milunovich with UBS.

I wanted to return to the DAS comments you made early on, Charlie.

Why do you think there's an opportunity there?

Folks like Nutanix would argue the opposite that hyper-converged is much simpler to implement and is now able to take on enterprise workloads.

And at least on the enterprise side, isn't the network load much bigger than DAS?

So kind of where exactly do you think the DAS is used?

And why do you think you can replace that?

Right.

I think there are a number questions embedded in the question that you're asking.

First of all, I would say that the hyper-converged world works well at a certain level of scale, but at very large scale, very large data center environments, we believe that a much more structured environment for -- with a data-centric environment on the storage and the ability to instantiate spin-off apps to scale them, to reduce them all around the petabytes of data that customers are trying to analyze is the more efficient and effective way to go.

And we'll be able to prove that at our Accelerate conference that's coming up.

DAS is still the way that a lot of application environments work, and we're going to be able to show that, in fact, having shared accelerated storage is just a stronger, higher performance, lower-cost, higher density solution set.

And we already have customers going in that direction today.

And the final thing I'll mention is that the major cloud companies have already gone to shared accelerated storage.

They've already separated storage from compute, specializing in both, creating more efficient and effective environments.

And so we are enabling the same for our customers.

Okay.

And I think you suggested that, obviously, you want to make the company profitable, but once you've shown that you can do that, you're going to continue to invest in a pretty aggressive rate.

And the market seems very forgiving of that right now.

Investors seem to want growth over profit if you will.

If we go into a different environment, where profits are something much more important, which is something that kind of tripped Nimble up 2 years ago, would you change your view of things?

Or would you say, no, we've still got this huge market opportunity, and damn the torpedoes, we're going to really go for investing?

I think the important thing, which was something that we were not given credit for last year, is to know that we can be profitable if we want to be, right?

And I think we've shown that.

I think we've already shown that, but this year, we're forecasting profitability.

We always believe on delivering on our promises.

That being said, we do think -- I think for a company of our scale that growth is the most important thing.

And we'll follow the Rule of 40, but on a -- we're not going to be buffeted by short-term swings and mood.

We're going to really focus on growing this company to the scale that we believe we can take it to.

Your next question comes from the line of Jayson Noland with Baird.

I wanted to ask on net new hires first.

Plus 200 is the strongest we've seen for 6 or 7 quarters.

Was that by design or just a function of timing?

Yes.

Jayson, by design, obviously, we're a much larger company now.

So on a percentage basis, that's still a nice growth rate for us.

And remember that Q1 is the time that we really do like to bring in a lot of our go-to-market professionals.

They are coming off of good quarters at other companies, and we want to have them have great quarters here -- or years, and we want to have them have great years here at Pure.

And that's just been a natural cadence in the business.

So nothing unusual; as planned.

And then a follow-up on international.

28% of revenue, that's really good.

Clearly, incremental investment there has helped.

Any additional color you would add?

Yes.

I think our margins, they continue to have a higher and higher contribution from outside of North America.

We were pleased with the year-over-year compares.

It'll be up and down quarter in, quarter out, but on a full year basis, we feel like the investments we made are really starting to bear fruit.

Are there specific countries that have been strong or just broad based?

We've been really focused on the largest GDP.

I mean, that correlates pretty closely to annual storage spend.

So rather than really diluting ourselves across too many countries, we focus on the largest.

And so we see nice contributions across EMEA, across APJ as well.

So I think that we're going to continue to stay focused versus really distributed.

Your next question comes from the line of Ittai Kidron with Oppenheimer.

A couple from me.

First -- and maybe I'm over analyzing here, but when I'm looking at your U.S. growth, which was in the mid-20s, Tim, correct me if I'm wrong.

Assuming that your SaaS business is growing faster than that, it kind of implies that the rest of your U.S. business is growing somewhere in the low 20s, maybe upper teens.

And I'm just wondering, with the challenges that Dell EMC is going through and with the technology that you bring to the table, should we not be expecting more out of that U.S. business?

Yes.

So we feel great about the progress in our international business but also feel good about the Americas business as well.

It's got a big TAM.

Our focused upmarket is bearing fruit as well, and so we're going to keep investing here.

One of the things that we look at very focused -- in a very focused manner is the cohort analysis for our sales productivity.

And as we continue to hire U.S.-based sales professionals, they continue to ramp as fast or faster than the most productive cohorts out there.

And so provided those folks cross over the magic line, we're going to keep investing there.

So we feel very good about the Americas business, and the overall top line growth of 40%, we think, is something that we're very pleased with.

Yes.

But let me dig into that a little bit.

I understand the productivity seems to be going very well for you according to your matrix, but maybe it's just the reflection of the fact that -- do you feel like you have enough people in the field?

Or you feel like you're short relative to what you really want to be here in the U.S.?

Well, it's a $35 billion market, so what we've been trying to do is maximize for the top line growth while delivering the profitability.

And we think we have a nice balance there.

And so as Charlie mentioned earlier, we're going to keep investing and optimizing for that, and we think that we've got -- in a very good position to do so.

And we're always -- Ittai, we're always balancing that dynamic, and so every quarter, we kind of look at it and get a feel of where the business is headed and how that productivity is shaping up to step on the gas a little bit more or kind of stay steady.

And so we monitor that all the time as you'd expect.

Well, you had a chance to step on the gas this quarter.

You had a very nice beat on the bottom line.

Is it just a reflection of bandwidth?

You just don't have enough bandwidth to hire even more than you've done this quarter, which is a good number.

Don't get me wrong.

But it feels like maybe another 10, 20 people in there could have been pushed in with still good -- delivering very good numbers.

Yes -- no, I guess, I would say a couple of things on the bottom line.

Obviously, we need to make sure that who we bring in are top A+ talent, right?

So we're not going to sort of lower our bar there.

But I think on the OpEx side, the OpEx beat this quarter was more about some timing things in other parts of the business than necessarily sort of a lower sales and marketing headcount add that we would like.

So I wouldn't read too much into that on the S&M line.

I'll just say that we agree with you.

Yes, okay, very good.

Charlie, just you've been in the company for 2 quarters now.

I guess, help me think about what...

Three quarters.

Who is counting?

Three quarters.

Time flies when you're having fun, I guess.

Help me understand what is it that you've changed since you've come here.

I mean, they've brought you to clearly scale the business.

What are some the changes that you have implemented already in the organization that came from your initiative?

All right.

Well, we've changed the structure of the organization internally, gone through a business unit orientation.

We are going to be rolling out some new channel programs starting tomorrow at Accelerate as well as some new product programs associated, which I don't want to ruin the announcements so I won't give it up, but some new product programs that will be going out tomorrow.

All of which, I think, we've started since I've come onboard.

And I feel that, from an organizational and communication structure across the company, we've become a bit simpler to navigate.

I'd probably highlight those 3 things.

Your next question comes from the line of Tim Long with BMO Capital Markets.

Just 2 quick ones if I could.

You talked a little bit about visibility into the second half of the year.

Obviously, a lot of things moving in the right direction here.

If you could just give us a sense if you feel visibility has improved over the last 3 or 6 months.

And then back to the gross margin, just I hear the comments on flash; how it's kind of a positive either way on the top line, and the business and market share.

Just talk a little bit about the gross margin impact of rationalization of prices there.

And any other levers that there are to turn on that -- the gross margin line?

Yes.

So this is Tim, Tim.

I'll take both of those questions.

The first 1 from a pipeline perspective, I would say that the visibility that we have, and remember, we look at 3 or 4, 5 different measures in terms of how we instrument that business, that visibility is -- continues to be roughly the same in terms of that we've seen in the past.

And as you heard us guiding up for the year, that gives you some confidence in terms of what we're seeing building both in Q2 and obviously beyond.

So that's what I would say as it relates to pipeline.

As it relates to gross margin, we've been very pleased with operating within our long-term guidance for quite some time now.

From my vantage point, we were actually stable from a gross margin perspective quarter-on-quarter.

We're -- it's down about 2/10, but really, a lot of that was -- or all of that was really the fixed cost on a seasonally lower revenue number than in Q4.

So if you sort of normalize that, our gross margins are stable, if not increasing, which obviously is a good thing.

It speaks to the differentiation of our product.

Your next question comes from the line of Simon Leopold with Raymond James.

This is Victor Chiu in for Simon Leopold.

I wanted to circle back on your AI partnership with NVIDIA.

I guess, I understand the role of the GPU and that they're ideal for performing the matrix calculations required for deep learning.

But can you maybe help us understand what specifically about Pure's platform makes it optimized for AI applications and workloads?

Or is it just flash in general that benefits and Pure kind of indirectly benefits because of your strong market position there?

This is Kix.

I'll take that one.

Look, it turns out that AI is probably one of the most massively parallelizable applications out there, which is why GPUs in the thousands of cores they can bring to bear have been so dramatic in that market.

And based on the name of the game, when you go buy expensive GPU compute, is you've got to keep it fed with data.

And so FlashBlade turns out to be a perfect fit for that with its kind of perfectly scale-out nature.

It can scale to a massive amount of bandwidth to be able to feed data to those GPUs.

And the question was asked earlier a little bit about the synergy with NVIDIA's DGX business.

When you start to scale beyond one DGX to multiple DGX systems, then you just can't use the storage inside the DGX anymore.

You have to break it out externally into a shared pool that all the GPUs can start to share, and that's when we see the synergy really happening with FlashBlade.

Once the customer has gotten their kind of basics going on with AI and they're really ready to scale their project, shared storage with FlashBlade can come in and allow them to start to scale linearly as they add more and more DGXs.

So it's a win-win from the customer point of view, and it's a win-win and synergistic between ourselves and NVIDIA's business.

I think the only thing that I would add is that it's beyond just AI, too.

It's the overall data pipeline and explosion of unstructured data that's really feeding this.

And so I think AI is a great technique and a great tool that's going to be -- we're seeing terrific traction in the -- it's much bigger than that.

You're really literally replacing 50 racks of legacy technologies with less than half a rack of DGX-1s and Pure FlashBlades.

And so the ability to be able to quickly stamp those out and scale them is orders of magnitude better than anything else on the marketplace.

So we think there's a number of trends that are feeding this but AI being the lead.

That make -- yes, that makes perfect sense.

I guess, maybe is there something that differentiates FlashBlade from other unstructured (inaudible)?

Two things, I'll add.

This is Charlie.

I'll add 2 things.

One is the DGXs don't do a lot when they're sitting idle.

And if you're not feeding the data fast enough, you waste a lot of money on a lot of cores that are just sitting there waiting for data.

And the second thing is that, as Hat said, we fit all of this together between them and us in half a rack.

You can't fit in half a rack if it takes a rack and a half of equipment of our competitors' products to do it.

So we're differentiated both in physical size as well as in performance.

Your next question comes from the line of Mehdi Hosseini with Susquehanna Financial Group.

This is David Ryzhik for Mehdi Hosseini.

So in the past, you've talked about AI as a key application for FlashBlade.

But what about backup and rapid restore?

You touched on it today, and it seems like a pretty sizable opportunity.

Would love to get any metrics around that or maybe dive into the example of a deployment for that.

And I have a follow-up.

Yes.

David, this is Hat.

So I'll give you a quote that was shared with me earlier, "In a world of moving stagnant data to the cloud, why wouldn't I buy a solution that can be repurposed for analytics and AI?" So this is a very large enterprise customer that had a number of back-up options that they were evaluating in the quarter and chose FlashBlade to do it.

So there's this phenomenon where there's a lot of the stagnant data that they're just trying to push off into the cloud.

And so they want to have something that they can use to rapidly restore data and solve that immediate pain point problem but then also have that same platform we leverage for analytics and AI as well.

So we're seeing great traction there.

I think the other thing that I would say is that the back-up use case is something that our traditional IT buyer has money for.

And so if we can deliver something that is much more efficient and much faster for those mission-critical workloads that they need to be able to restore, leveraging the same budget, we can use that as a bridge to be able to introduce capabilities into data sciences and DevOps, which are different personas.

And so our sales team and our channel partners and our existing customers all really know and understand the back-up use case, so it's an easy thing from a selling motion for us to attach to.

Got it.

And relative to your fiscal '19 guide, what type of NAND flash pricing trends are embedded into that guidance for the balance of the year?

Yes.

This is Tim.

So we've said for several quarters now that we anticipate the latter half of this year that we're in, the NAND prices to start sort of coming down a bit.

And we're already starting to see signs of that, and that's sort of how we thought about modeling the year.

And it's really playing out kind of as we thought it would, like I say, 3, 4 quarters ago.

And lastly, so Micron announced that they're already shipping QLC SSDs.

Do you see this incorporated in both FlashArray and FlashBlade?

And as a follow-up to that, you've talked about one of your core competencies is optimizing consumer-grade NAND.

Can you optimize consumer-grade QLC for enterprise performance use cases?

Yes.

Look, that's a great question.

I love where you're going with it.

We've said all along that we've been building this kind of unique capability compared to our competitors to really understand how to interface our software directly with raw NAND.

And that gets only more exciting as you get into QLC because the challenges of making QLC reliable and perform and at scale are even harder.

And so yes, we're excited about QLC and its ability to further drive down cost and thus get to different and unique new use cases for solid state within the data center.

So not at the point right now we're ready to announce times for shipping or anything like that, but it's definitely something that we're excited about and working on.

Your last question comes from the line of Jason Ader with William Blair.

Yes.

I'd like to drill down on the DAS replacement opportunity.

Is this the top of rack/architecture that you referenced a few quarters ago?

And can you provide any more details on the bake-off that you had with the SaaS customer and why you guys won?

I'll take that one on.

When we look at this opportunity, it's really about going after tomorrow's applications.

And so if you look at the wide range of "web scale" applications ranging from scale-out databases to modern analytics on through, frankly, AI, we increasingly see them be scale-out, container-centric and their 1.0 architecture ran on DAS.

But we're finding customers that as that DAS environment scaled, see all the inefficiency of DAS.

And they also start to realize they don't just run one application.

They run 15.

And so consolidating an architecture that can allow them to truly run their classic apps and their next-gen apps on one architecture is a great opportunity.

And so that's why we're out talking so much about it.

Within that particular SaaS customer environment we mentioned, one of the really exciting things was it was for their DevOps use case.

And so folks think that you have to go to the public cloud to get the agility for DevOps, but in this case, we're able to actually show that by investing in an architecture for shared flash, they can get the flexibility of that on demand and really power a modern DevOps CI/CD pipeline.

Thank you, Kix.

And with that, ladies and gentlemen, I want to thank you all very much for joining us today, and we look very much forward to seeing some of you tomorrow and Wednesday at our Accelerate User Conference.

This concludes today's conference call.

You may now disconnect.