“ We’re building a massive amount of infrastructure. At the end of this year, we’ll have ~350k Nvidia H100s – and overall ~600k H100s H100 equivalents of compute if you include other GPUs” - Zuckerberg
Maybe Pure will also get some new orders this year to support Meta’s build up in their AI infrastructure. At least that’s what the market is hoping for I guess… (+5.3% today) - TSMC earnings results also help.
If their performance enhancement + lower energy cost vs other storage solutions like hdd/ssd is as great as they say, it should show up in their revenues over the coming year.
With Pure Storage ($PSTG) running laps around everything else in my portfolio today, I went looking for some news. I don’t know why I didn’t get these releases in my email, but this is from yesterday. The headline cuts off at the amount of reduction in the Data Center footprint, which is 85%.
Making a similar point was this from March 26:
Apart from being good news, I think it’s also showing where the pain points are now that all eyes are on Data Centers.
** “Embracing Pure Storage’s all-flash portfolio over legacy tape was a strategic slam dunk for our organization. It goes beyond a mere transaction; it’s a genuine partnership that has not only elevated our performance but also ensured the longevity of our critical IT systems. Pure Storage has become an integral part of our success story, providing a robust foundation that aligns with our goals for the long haul” – Andy Marosi, Sr. Manager, Infrastructure & Operations, MPI**
I didn’t even realize there were still people using tape storage. That is crazy.
I believe long term archive storage (infrequently or rarely accessed) still use tape because it has a longer shelf life than other media, up to 50 years. That staying power comes from its high capacity, low cost per megabit, and durability. BTW My first job out of college was to program a device driver for a storage cassette tape…sigh…that was a long time ago. Whereas, HDD and Flash memory shelf life is ~5 years. CD/DVD is ~10 years. So this kind of explains why old microfiche storage is still around for property and legal records. For all the people who store their precious family pics in the cloud, will they still be there in 50 or 100 years from now? You need to think about this.
Total cash, cash equivalents, and marketable securities $1.7 billion
Of particular note:
Revenue growth rose from -2.5% in Q4 and -5% in Q1 FY24 to +17.7% (highest growth in over a year)
Subscription Services came in at their highest ever contribution level and ~50% of total revenues
Subscription ARR growth at 25% has levelled out (with 25% at last quarter arresting the ever shallowing growth rate down trend). This strips out billed work not yet commenced - including that it would have been 27% growth - in line with RPO growth
Q1 beat the guide and Q2 is above consensus. Pure has left full year guide unchanged but will have to raise revenue and profit guides next quarter if they have another significant beat
Accompanying remarks pointed to new AI use cases with Nvidia and extended self serve capabilities in the subscription space
Earnings Call:
Opening remarks… "The recent advances in AI have opened up multiple opportunities for Pure in several market segments. Of greatest interest to the media and financial analysts has been the high-performance data storage market for large public or private GPU farms. A second opportunity is providing specialized storage for enterprise inference engine or RAG environments. The third opportunity, which we believe to be the largest in the long term, is upgrading all enterprise storage to perform as a storage cloud, simplifying data access and management, and eliminating data silos, enabling easier data access for AI.
Pure is seeing early success in all three of these AI-based opportunities, and we can address them all with our unified Purity platform."
“The ease of use of our platform, and a notable interest in saving power and reduced environmental impact, led to a notable win with a managed service provider specializing in high-performance computing. Their accelerated environment for both large language models and inferencing delivers top-tier AI infrastructure and training solutions for their financial services, energy and life sciences customers.”
“Enterprise and International market segments were strong this last quarter. Our //E family continues its strong growth and was also a key enabler in our discussions with hyperscalers. FlashBlade had a record Q1, including in AI workloads.”
“Also, while storage spending for AI is still in its early stages, we believe that we are well positioned as demand for data storage accelerates.”
Q on Meta:-
“So when we – you mentioned Meta, and, yes, we’ve sold into their GPU environment. By the way, they have multiple AI environments, and we are in the majority of the environments that we’re aware of in Meta AI”
Q on NAND pricing increases:-
“So, net-net of all that is NAND prices generally don’t affect our gross margin as much as they do with the market – the top-line of the market as a whole. And so, with prices rising, I think that’s somewhat of a tail – that should be something of a tailwind in general”
Q on Pure’s tech advantage:-
“And as a reminder, QLC flash represents the majority of the flash we consume today, really providing a meaningful sustained cost advantage both against TLC flash and commercial SSDs using QLC flash.”
Q on hyperscaler opportunities:-
“When we talk about hyperscalers, and this maybe go back some quarters, we’re really speaking about the top 10 hyperscalers. So, it goes beyond just the three public clouds, if you will, to the other, the FAANGs, et cetera, in this. So, we – though – the quantity, meaning the number of those players that we speak to and the quality of those conversations have improved. I mentioned as well that we’re now not just experiencing testing in some of those environments, but some commercial discussions as well.”
Q on customer growth:-
Similar to Q1 last year
Q on Storage as a Service:-
Slower TCV start but built through the quarter - good pipeline
Q on customer concentration:-
No 10% customers
Q on subscription service revenue inflection point:-
“That is approaching 50%. It’s in the 40%s current – we expect in the 40%s this year. That’s approaching 50%. And as we go into next year and beyond, we do expect to see it increase above 50% as a whole.”
On AI related revenues:-
“the dollar attached to an AI training environment after you’ve paid for the GPUs and the networking and the racks and so forth, is in the 5% to 10% range”…
“That being said, we think that, probably in the past 12 months, about $1 billion was spent on storage specifically tied to AI training environments. Again, that might be a bit less than people expect, but I think our data on that is very, very good”
Thanks for the effort on your detailed writeup. PSTG is my wife’s previous employer. Now I have to run off and see how shares she still carries. I started buying PSTG last year as I researched the Evergreen service and saw the initial uptake. The drop in 2023 Q4 revenue startled many and hurt the stock, but it was really an endorsement of storage as a service offering poaching hardware sales. (a different SAAS). So I bought some more. This kind of transition is similar to AXON’s business model transition to cloud services. Kind of a sleeper move beneath the covers of hardware sales. PSTG has really an awful PE and PEG until you recognized the growing RPO.
Playing “If you rather” I think PSTG has a long runway like AXON and am surprised it is not more widely followed and held on this board.
13 PS on AXON is much higher than 7 PS on PSTG
41 forward PE on PSTG is lower than 63 PE on AXON
I need to run some valuation numbers versus growth rate but I think PSTG looks much better by the numbers and growth. And some AI tailwind. But PSTG has a bunch of established competition whereas AXON kind of has a monopoly. I own both.
@Andy
Not sure what the comparison point is here, I hold SMCI also (and used to hold Nvidia) but just comparing PSTG with the highest appreciation stocks of the year doesn’t necessarily show it is a good or poor investment choice for this board. For many of us who have bought PSTG more than 1 year ago are looking at between 200-300% appreciation from taking a position in a company with a compelling investment case. PSTG’s investment thesis and share price appreciation wasn’t based on a once in a lifetime AI investment opportunity arriving out of nowhere. Of course if we want just to compare gains then nothing comes close to Nvida and SMCI but that’s not so helpful as we think about pursing an investment strategy. Of course Nvidia and SMCI also have a case to be made on their go forwards growth potential if that can be sustained.
@Zane
Dell and NetApp results out today so should give some useful head to head comparison data but I agree with you, Pure’s business model transition has been misunderstood, it has been undervalued for a long time and Flash storage has a long runway of market penetration and demand growth ahead.
Maybe I am missing the point, but how is Pure Storage a growth company? What metrics make it a growth company?
The year over year revenue growth rate this quarter is 18%, and last quarter it was -2% year over year. The last 20%+ growth revenue quarter was seven quarters ago.
@anthonyms
I think you should have held SMCI and NVDA instead of selling them. That was the whole point Ant. PSTG isn’t even growing 20 percent. How could you be holding them when SMCI and NVDA are growing so fast? Let me show you a few charts.
Now looking at the fundamentals on those charts (lower left hand quarter) you can see exactly which ones are the best. With NVDA and SMCI it is impossible to project the future because they are growing so fast.
I do still hold SMCI - I sold Nvidia many years ago just like Tesla which are the reasons why I am not retired.
My issue with Nvidia and SMCI is that just as this revenue explosion came out of nowhere with zero visibility, the potential to hit the wall might also come without warning.
There is still room in the portfolio to enjoy other growth stocks on their own merit which have more visibility and an outstanding track record (like Pure - although I admit it is one of the lower growth holdings but it is growing at scale ($3bn) that many others of our portfolios are yet to reach).
Anyhow don’t want to get dragged into portfolio discussions, and there are other threads for Nvidia/SMCI, this one is for Pure.
Their headline growth rate is a combination of growth rates of their legacy business plus their newer subscription business. Subscription business is growing faster and has better margins.
Part of what we do here is identify companies with good numbers, and then investigate to see if there is a good thesis-narrative behind the numbers.
But another part of what we do here is identify a good thesis-narrative and then investigate to see if the numbers back it. $PSTG is an example of this latter approach.
It’s important to assess $PSTG numbers not as a single point-in-time take on the headline numbers. Instead, we need to break out two stories (legacy and subscription) and assess DIRECTION, not just point-in-time.
When we do that, we can see that the most important numbers are not the headline numbers but are in fact subscription revenue, ARR and RPO, all of which are trending nicely.
Consider further that storage for AI deployments will likely happen AFTER GPU is deployed, and a narrative thesis emerges that we can track this AI tailwind via subscription revenue, RPO and ARR.
So what the headline numbers are disguising is that growth is pretty good right now, and likely will accelerate in tandem with large GPU deployments. My take is $PSTG will be much more expensive by the time the headline numbers are reflective of the actual thesis-narrative…and of course I could be wrong about that.
Bert says it better than I can; I definitely recommend Bert’s article linked above by @anthonyms
Their headline growth rate is a combination of growth rates of their legacy business plus their newer subscription business. Subscription business is growing faster and has better margins.
Part of what we do here is identify companies with good numbers, and then investigate to see if there is a good thesis-narrative behind the numbers.
But another part of what we do here is identify a good thesis-narrative and then investigate to see if the numbers back it. $PSTG is an example of this latter approach.
How do the numbers back that this is a growth company though? The subscriptions growth you mentioned is growing 23%
These are the numbers they just reported,
Revenue $693.5 million, an increase of 18% year-over-year
Subscription services revenue $346.1 million, up 23% year-over-year
Subscription annual recurring revenue (ARR) $1.4 billion, up 25% year-over-year
Remaining performance obligations (RPO) $2.3 billion, up 27% year-over-year
There’s not a single metric growing over 30% here. Pure Storage is simply not a growth company no matter which way you look at it. As Saul has mentioned many times he usually looks for companies growing 40%+ or at least 30%+ for revenue.
“There’s not a single metric growing over 30% here. Pure Storage is simply not a growth company no matter which way you look at it. As Saul has mentioned many times he usually looks for companies growing 40%+ or at least 30%+ for revenue.”
A very good and fair challenge that in the spirit of Saul investing principles and given the number of Pure investors on this board that I see amongst portfolios shared, it deserves a response.
What I would say to you is that for this Q1 2025 and in the prior period you have referred to over the past 2 years, leaving aside the fact that this is a $3bn revenue company growing at scale, is…
Storage is cyclical and the last year has been a down year for storage, Q1 is the return to growth stage of that cycle. (Pure in previous cycles has previously grown at 40%+ to fall back to 10-20% growth only to return to 30% growth rates before.)
On a like for like basis RPO grew at 30% or greater every quarter in FY 2024, (minus the impact of a lumpy contract).
In the 2 years preceding that FY22 and FY23, Pure grew subscription revenues at 30% or more in all but 1 of the 8 quarters.
Storage and Pure storage are expecting growth acceleration and with the explosion of the AI demands of CPU/GPU processors, anticipate that storage will be the next beneficiary, then software.
Whilst the fact that Pure Storage has returned far more than most Saul investments I carry or have carried doesn’t make it a Saul like investment, this is still a growth at scale story of a great company and now a great stock. It might be at the slower end of growth, but I warrant deserves consideration.
(BTW When I discovered Saul’s board, 20% growth was the threshold).
Ok so Pure Storage clearly continues to take market share from Dell and NetApp judging by their results overnight…
NetApp (Q4 FY24):-
Revenue of $1.67B (+5.7% Y/Y) beats by $20M.
Billings: $1.81 billion, compared to $1.67 billion in the fourth quarter of fiscal year 2023; a year-over-year increase of 8%.
All-flash array ARR: $3.6 billion, compared to $3.1 billion in the fourth quarter of fiscal year 2023; a year-over-year increase of 17%.
BTW their “ARR” isn’t Annual Recurring Revenues it is Annualised Revenue Run Rate so their projection of $3.6bn extrapolating from their current quarter (Q4 which is their bumper quarter) would be equivalent to Pure’s projected total revenue of $3.1bn FY25 which will be raised, (or the using equivalent bumper Q4 $790m x 4 for Pure). So Pure has almost caught up to NetApp’s flash business which is astonishing given the legacy storage sell-in advantage NetApp has over Pure, (kinda like Palo Alto’s advantage in cyber security). https://www.netapp.com/media/107262-earnings-slides-q4-fy24.pdf
DELL (Q1 FY25):-
Storage revenue was flat at $3.76 billion (down 16% QoQ)
NetApp got an upgrade but more to the point, here is what the JP Morgan analyst review said about the potential impending AI driven demands for flash storage which equally applies to Pure as NetApp - maybe even more so…
“We are upgrading shares of NetApp to Neutral from Underweight prior as our concerns around gross margin headwinds from a backdrop of rising memory prices has not transpired with the company executing robustly to manage the headwinds through pre-buy of memory while driving a richer portfolio mix in relation to higher margin all-flash drives,” said J.P. Morgan analysts Samik Chatterjee, Joseph Cardoso and Priyanka Thapa, in an investors note.
The investment firm also increased its price target to $125 from $95.
“Post the early phase of the AI investment cycle focused on Cloud and Compute, Storage is expected to benefit more in the Enterprise investment cycle toward inferencing workloads,” Chatterjee noted. “Should the AI investment cycle accelerate and Storage becomes a beneficiary in AI in a later stage, that would lead to customer spending increase and a rebound in Storage earlier than we expect.”
The other point that I take away from seeing Dell and NetApp revenue numbers is the $ billions per quarter total addressable market for storage which has been made up predominantly of spinning disks and hybrid storage.
If one believes in Pure’s prediction that there will be no new spinning disks sold within a few years and with the TCO of Flashblade E comparable with spinning disks now, all of that TAM becomes accessible with All Flash. That’s a lot of market potential and revenue growth right there irrespective of end market demand growth.
