back to article Oh, 3PAR. One moment you're gliding along. The next, you're in the rain as HPE woos Nimble

This week, HPE offered to acquire Nimble Storage for around $1.09bn, plus another $200m in share options. Nimble sells all-flash and hybrid storage solutions, with a lot of intellectual property focused around storage analytics in the form of their “InfoSight” SaaS platform. Commentators are seeing this as a good deal for both …

  1. Missing Semicolon Silver badge


    .. isn't 3PAR inclined to croak and throw all your data away?

  2. thomn8r

    Meg is running out of people to lay off, so she has to acquire another company in order to have enough warm bodies to keep the wood chipper going.

  3. Anonymous Coward
    Anonymous Coward

    Oh goodness here come the meatheads...

    3PAR has seen nothing but continued growth and that's because the product was designed completely different than any of these other systems. The reality is that evolution was built into the product which is why features have continued to be added over the years.

    That aside, the fact is the 3 most successful arrays TODAY are VMAX (30+ years), NetApp (20+ years), and 3PAR (15+ years). That just kills the whole "20 year" premise, doesn't it? (In All Flash NetApp WAFL for Flash & 3PAR are the top 2 individual products. EMC, NA, HPE, IBM are the top 4 vendors.)

    I will concede there's some truth to the power of processors BUT there's plenty of evidence that shows their limitations as well:

    - EMC VMAX AF uses latest/greatest procs. Will never dedupe because it overruns them.

    - Pure claims to use latest/greatst procs. Hasn't added a single feature in 4 years.

    - HDS uses multiple procs--more than most. Best Practices still require Cache Partitioning and other measures to minimize the CPU impact.

    - NetApp has numerous limitations with Dedupe, Compression, Thin, and so on.

    - IBM V series--their #1 seller--no dedupe same reasons as EMC

    That's just to name few. These are all good systems with limitations. 3PAR, like anything else, has hit speed bumps but ultimately has delivered. I would say this: 3PAR is the fastest growing array in the industry and has continued to grow even in the world of spinning disk. More features are still coming actually. The CPU+ASIC parallelization is really at the core of that success. It just works. That's how you keep adding features, scale to 8 nodes, federate to 32 nodes, in-line everything, and so on and so on and so on...

    Instead of speculating and writing these types of blogs why not wait 5+ weeks to get the actual data? Or you can keep writing these types of pieces and reinforcing the whole "Reg = Tabloid Caca" thing...

    1. diodesign (Written by Reg staff) Silver badge

      Re: "Oh goodness here come the meatheads..."

      Chris is welcome to his opinion.


    2. Anonymous Coward
      Anonymous Coward

      Disclosure: Nimble (soon HPE) employee

      I couldn't disagree more. The reason why those old systems have problems with technologies like dedup, compression, thin provisioning etc. has nothing to do with CPU. It's the old architecture, a design for disks, that limits them.

      Nimble is a CPU driven architecture. We implemented all those features without ASICs and we're doing it much more efficient than any of those legacy systems. It also enables much better scaling capabilities. Every year the systems get faster because Intel produces faster CPUs with more cores. And we don't have to change a single line of code to make use of it, because it multithreaded anyway.

      Another thing on Chris's comment: InfoSight is a good product which can easily be extended to the 3PAR platform.

      Absolutely not! If you believe that, you didn't get it at all!

      At Nimble we've programmed the product to collect and log millions of data points and sensors from day 1. There's absolutely no way that can back ported to ANY product.

      All in all, we at Nimble are excited too! This opens a lot of doors for us, will accelerate the product development. We now have 10.000+ customers. I bet my ass in two years from now it's more than double!

      1. sal II

        >Nimble (soon HPE) employee

        My sincere condolences

        Signed: former HPE employee

        1. Stand001

          Former HPE

          Hi former HPE Employee, sorry to here that you did´nt had the same experience as I have. For me HPE is a great company to work for, been here for 9 years

          1. thomn8r

            Re: Former HPE

            For me HPE is a great company to work for, been here for 9 years

            Don't get too comfy.

            If you've only been there 9 years, then you missed out when it really was a great company to work for.

          2. Anonymous Coward
            Anonymous Coward

            Re: soon to be Former HPE

            Look around future ex-HPE employee - when you reach 40 into 50 they start writing your bad reviews and lay you off because you are no longer a "culture" fit for the new business model of young hipsters who will work endless hours without complaining ; Look around and count how many middle age employees are around ... you will come up short .

      2. spinning risk

        I hope you keep your job. The fact that all the unvested options were paid out is not a good sign for Nimble Team; that sucks and I am sorry.

        When HP bought 3PAR we did not get paid our unvested options; we received a check for all vested in our W-2 (OUCH) and all unvested converted to HP stock at some crazy Nazi math calculation. They wanted the 3PAR team to stay on because the legacy HP storage team was lost and could not spell 3PAR if you spotted them the 3 and the P.

        Many of us split because HP was too big and life turned into spreadsheet management; not to mention that HP screwed up the support model and pricing went through the roof. I knew it was the beginning of the end when they started charging for Thin Provisioning; something that was always free because it was by-product of the OS not something that was actively created.

        I hope HPe does not break Nimble but why would they change...

        1. Anonymous Coward
          Anonymous Coward

          "I knew it was the beginning of the end" sorry bit HPE sold more 3PAR during the first 3 month (7000) than 3PAR manage to do during 10 years. 3PAR is the most sold midrange array in the world, we have diffrent view on "I knew it was the beginning of the end"

          1. Anonymous Coward
            Anonymous Coward

            HA HA.Sorry. I was ambiguous. I meant the beginning of the end for me at HP. I was in early at 3PAR and could not stay and watch the magic seep away.

            Of course HP sold more than 3PAR, DUH??

            That was the point of buying us (3PAR) and getting global reach, supply chain management efficiencies, big margin (at the point of acquisition we averaged 52 pts of margin and still ate the incumbents lunch on price) and re invigorating a stale storage line.

        2. Anonymous Coward
          Anonymous Coward

          vested options are not being paid out...they are setting aside that money to turn them into HPE unvested shares

      3. Anonymous Coward
        Anonymous Coward


        Hello soon to be colleague, Nimble is a great products but I´m not sure if you understand how 3PAR works and when it comes to Infosight I´m sure that we will be able to use it also for 3PAR, already today we have a product called Storefrontremote that is simular but not as advanced as Infosight, Welcome to HPE

      4. returnofthemus

        So HPE, what is your long-term storage strategy?

        I would've thought the anwser to that obvious, there isn't one!

      5. returnofthemus

        Another thing on Chris's comment:

        "InfoSight is a good product which can easily be extended to the 3PAR platform.

        Absolutely not! If you believe that, you didn't get it at all!"


        So esentailly HP have just bought a bunch of boxes filled with SSD's and Intel CPUs, who does the due dilligence over there?

        1. Anonymous Coward
          Anonymous Coward

          Re: Another thing on Chris's comment:

          You really think HPe bought it without knowing if it could port over to 3PAR? That was the only reason they bought this $$ hemorrhaging product..

          1. Anonymous Coward
            Anonymous Coward

            Re: Another thing on Chris's comment:

            Disclosure: The Nimble (soon HPE) employee again

            Allow me to stay AC until the deal closes. Comment #3

            >>You really think HPe bought it without knowing if it could port over to 3PAR? That was the only reason they bought this $$ hemorrhaging product..

            Yes, absolutely. The long term plan has to be to replace 3PAR with Nimble.

            This is the only way this whole thing makes sense.

            1. spinning risk

              Re: Another thing on Chris's comment:

              You have been drinking too much Nimble cool aide. There is no way Nimble could ever replace 3PAR with a scale up only architecture. How does Nimble scale out like a 3PAR or a Kaminario for that matter? (Scale out cluster does not count btw) Active active controllers that scale up and out; true performance....

        2. Anonymous Coward
          Anonymous Coward

          Re: Another thing on Chris's comment:

          Disclosure: The Nimble (soon HPE) employee again

          Allow me to stay AC until the deal closes

          >>So esentailly HP have just bought a bunch of boxes filled with SSD's and Intel CPUs, who does the due dilligence over there?

          I usually don't comment much here, but this keeps me up at night... so here is my comment #2

          No, they bought a our IP. They bought InfoSight and our filesystem. But the two things come together.

          CASL is hardware agnostic. So it could run on MSA hardware and it could theoretically run on 3PAR hardware.

          MSA hardware makes sense. It will be an entry system with lots of features, a massive gain in performance, easier to use, VVol integration, superb snapshots, predictive analytics... the whole lot.

          And HP can make it cheap!!!

          IMHO running CASL on 3PAR hardware doesn't make sense at all. I think our "off-the-shelf" hardware is much cheaper as we don't rely on ASICs.

    3. Mr.Nobody

      "EMC VMAX AF uses latest/greatest procs. Will never dedupe because it overruns them." - I have no idea what you mean here, unless you are saying the VMAX is so fast, that the processors can't keep up with dedupe? EMC runs them too hot? I think the really answer they will never dedupe is that would mean fewer sales dollars for EMC.

      "HDS uses multiple procs--more than most. Best Practices still require Cache Partitioning and other measures to minimize the CPU impact." Can you explain why this is bad?

      "Pure claims to use latest/greatst procs. Hasn't added a single feature in 4 years." I think their filer offering is certainly a new feature. Since they are new product, they don't really have anything else to add, as would be the case with all the legacy systems.

      "- IBM V series--their #1 seller--no dedupe same reasons as EMC" - I don't understand the point you are trying to make about EMC, ergo I don't get the point with IBM (unless its about IBM making more money since their customers have to buy more disks)

      "NetApp has numerous limitations with Dedupe, Compression, Thin, and so on." Huh? ONTAP 9 has a lot of excellent new features, and dedupe, compression and thin provisioning have been options with them for a long time, and are more impressive with AFF (I can say from experience). I think they continue to make great strides in this business.

      Don't get me wrong, I have always liked what I have read about 3PAR (except the data loss part, which frankly has happened to every storage vendor at some point), but i don't get any of your points against 3par's competition here, they all seem to be explanation free.

      3Par also seemed like the perfect aquisition for HP in 2010. I was an EVA customer then and got burned by the EVA team picking up and leaving after HP aquired compaq back in the day. HP was totally screwed by that. Not only could they not add new features to it, they couldn't fix bugs.

      We had a DR event with one of our primary HP EVAs, and we quickly learned that without the Primary EVA being online, you couldn't make the destination EVA volumes writeable (or anything-able). We actually got HP people on site, and they couldn't figure out how to fix it, they just wanted to fix the primary array (which they did). It only made sense at that point to buy something to replace the EVA, and 3PAR was the perfect fit.

    4. Anonymous Coward
      Anonymous Coward

      Just a bit of advice - it's obvious you're closely affiliated with 3Par, and none of the other vendors you listed, so stick to talking about what you know.

      VMAX-F isn't skipping dedup due to processor overruns. It's an architectural issue, a massive architectural issue, due to the way they handle I/O. Processing power is the least of their concerns.

      Pure hasn't added a single feature? Seriously? Ignoring the fact they started from day1 with 90% of the features the other vendors you listed needed to tag on (deduplication, compression, flash management - which most still don't do to the same degree). They've since added things like replication (several generations with significant changes), VLAN tagging, NVMe support, containers for hosting virtual machines, etc. What exactly is the breakthrough feature you think is missing?

      HDS - uses more than most? Cache partitioning isn't about CPU impact... I'm not even sure this is worth touching. Cache partitioning is about... memory utilization. I'm guessing you had to go this route because you *REALLY* want to believe that an ASIC is always superior despite the market showing it's not. There's a reason HDS - the KING of ASIC has moved to x86.

      NetApp - what limitations? There is not and has not ever been a "limitation around thin" - that sounds like something coming from a 3PAR shill. Compression in it's first iteration impacted performance but that's been fixed for several years now. Deduplication - they were one of the first to market and keep making improvements, again, with vague "limitations" and no actual stated limitations that's tough to refute.

      3PAR has had significant issues at their largest customers - see Australia articles. The fact they acquired a different storage product shortly after that snafu with no explanation as to why, and the general public can't help but wonder what is going on.

  4. Nate Amsden

    more than 8 nodes

    just FYI HPE 3PAR has had storage federation abilities for several years now, as of last check currently goes to 4 arrays, so the doc I see here says max say 32 controllers 60PB of usable capacity and 10 million IOPS(at the high end). Individual volumes I would assume can only reside on a single system at a time, but with that much data and I/O you are likely to have many many volumes that can be distributed/load balanced over a federation if the customer desires to do so.

    I have not used federation myself.

    The true hybrid(as opposed to tiering) SSD/HDD tech that Nimble and a couple of the other startups offer seems like a great solution for many types of apps. I am not sure how much of their sales is their original hybrid vs their all SSD and where that is going in the future. Other than my past history with 3PAR one of the main things that keeps me on the platform is support for 4+ controllers, I really like having a true active-active 4 controller array as a base, especially when my workloads are often 90% write.

    Though Nimble seems very light on data services relative to 3PAR. Myself I don't use many data services on 3PAR but they are important to many customers.

    Back in about 2011 or 2012 I really questioned whether or not the 3PAR architecture had the ability to grow into the flash era(and at the time I didn't think it would), but the stuff they have done since really blew my mind (with exception of compression taking too long).

    The company I am at has some Nimble(another group manages it, the other group deployed Tegile in the past year or so because it was cheaper than Nimble), relatively speaking a lot of 3PAR, and about to get some Isilon as well (if only HP could make a decent NAS..I miss my Exanet from eons ago). I wanted to deploy Isilon SD Edge but like some other NAS products I have tried imploded pretty quick, their hardware platform addresses shortcomings not possible in the SD Edge product).

    1. Anonymous Coward
      Anonymous Coward

      Re: more than 8 nodes

      With a 90% write workload you're going to get shite performance out of a 3PAR. Every optimization, improvement, and tweak they make is to improve read, not write. AFC-read cache only, 3D X-Point/Intel Optane-used for read, Adaptive optimization-does not preferentially target writes to the higher tier, snapshot model is COW, so even worse for write.

      My 3PAR is among the architecturally worst arrays for writes I've ever seen. 7-mode ONTAP is better, Compellent is WORLDS better, EMC VNX/Clariion/Unity is better.

      I too am running 90% write, and it consistently gives us performance not nearly in line with the hardware we've got. With an equivalent amount of flash in any of the above mentioned arrays we'd be screaming fast. As is, we've just had to move 100% of our sensitive workloads to flash to deal with the performance issues.

      Also, and this is totally separate, 3PAR Dedupe is terrible. If it was any good, would they have trashed it and started over with 3.3.1? I think not.

      1. Anonymous Coward
        Anonymous Coward

        Re: more than 8 nodes

        With that kind of write persona you should be in purpose built Flash array. Pure or Kaminario will outperform any of those and get 4:1 DR except on SQL or Oracle; they get 2.3:1 on those apps plus a whole lot less $$..

        1. Nick Dyer

          Re: more than 8 nodes

          ...funnily enough, one of Nimble's strengths is that it can perform at 100% writes with small block random or large block sequential IO without performance issues. Regardless of all flash or hybrid thanks to the file system.

          1. Zerolab

            Re: more than 8 nodes

            Any system can perform at just depens whats the absolute number translates into. And I can't see that number ever being better on 2 controllers vs 8. Also if Nimble performance were so awesome there would be some public evidence of this in the benchmarking or analyst community....

            1. JRW

              Re: more than 8 nodes

              JohnW from Nimble here

              To Zerolab, you are right, every system can run at 100%, one nice thing about Nimble is it won't choke if it happens to be 100% random writes. What the 10,000+ Nimble customers experience is a platform that performs consistently well with mixed workloads and backs that up with the industry's best analytics to learn from the install base and get the best availability and support experience. Is Nimble going to win a synthetic benchmark? Not against the like of Violin and DSSD which are built just for speed. Not against systems configured RAID 10 and data services turned off. A Nimble cluster can do a million IOPS at under 1ms latency, with features turned on and all the other benefits. Nimble makes a great platform for most applications and that is how we have got to >10,000 customers. Look at sites like reddit for how Nimble customers feel about using our arrays in the real world. IMO HPE have made a very smart move.

              1. spinning risk

                Re: more than 8 nodes

                Stop selling man. HPe owns you now. Noone cares if it works or not; its HPe

          2. Anonymous Coward
            Anonymous Coward

            Re: more than 8 nodes

            hmmm. Define performance issue? Just because a dump truck can go wide open @ 20 MPH empty or full it does not mean its high performance. Nimble is not fast; its consistent with many write or read personas but its not known for speed.

      2. luis river

        Re: more than 8 nodes

        I dont Know, but here many incongruencies on the mentions of "re: more than 8 nodes reply" , I believe on the virtues of 3PAR storage, dedicated ASIC is one formidable Tech

  5. StorageBuddhist

    Storage Spaghetti...

    Ahem, I did call it back in 2014...

    1. spinning risk
      IT Angle

      Re: Storage Spaghetti...

      You did call it.

      Old OS also means it was written for HDDs not SSD. That write algorithm was mean to account for rotational latency; problem #1.

      #2 Metadata- ouch

      #3 Garbage collection-Problematic

      #4 Write cliff- easy to fall off

  6. GrumpyOF

    Is it my imagination

    or has HPE lost the INVENT that used to characterise the old, old HP?

    Certainly on the storage subsystem level it would seems so. Just about every product/solution they bring to market is not HP developed (wasn't the MSA a DG product??)

    So, in my thinking, they are basically going the same way as IBM in losing any capability of developing leading technologies (I am convinced the SVC was based on FalconStor ). DS8000 is hardly an industry leading technology.

    So it is confused your sales force with too many options, confuse your existing clients with too many options and no-one to unravel and develop a best fit solution (extremely poor pre-sales support).

    regardless of what/ who they buy they are in trouble.

    (Not a 3PAR fan either)

    1. ManMountain1

      Re: Is it my imagination

      How many major vendors invent? There is a whole ecosystem out there that relies on the fact that startups innovate, some wither and die and the best ones get bought (which was their intention from the start). What was the last thing that EMC, NetApp, IBM, whoever 'invented'? And as if SVC was Falconstor ... inbound vs out of bound virtualisation for starters. HP used to rebadge Falconstor SDS for a short while. They were very much competing technologies. And DS8000 is clearly IBM's IP as it used to be bloody great big RS/6000 arrays under the hood! You just throw random criticisms around in the hope someone believes your drivel!

      1. GrumpyOF

        Re: Is it my imagination


        Thank you for pointing out some errors in my comments (or drivel if you prefer):

        of course SVC was not based on Falconstor , it was based on DataCore SsnSymphony.

        However the inband versus out of band comment is a bit off since Falconstor NSS runs inband (Ability to integrate with both FalconStor CDP for out of-band data protection and FalconStor NSS for in-band storage virtualization.)

        The big issue I have is that although the IBM, HPE and EMC's of the world throw billions of dollars at research (not a lot of it on storage, storage software etc), they still go and buy their technology and with perhaps the exception of EMC fail dismally to incorporated whet they have bought (or OEM-ed) into a cohesive stratgey with all of the requisite ecostructures to support their clients adequately (reliance on partners!!!).

        Dell EMC has gotten itself into something of a tangle with xtremio and DSSD, now Nutanix.

        I did say the DS8000 was IBM through and through and I have only seen it in a mainframe environment, which unfortunately is a dwindling market place..

    2. Fortycoats

      Re: Is it my imagination

      The MSA was a Compaq product. I think it was based on their Smart RAID controller for Proliant servers, just in an external box. And the EVA was based on the HSx Controller originally developed by Digital (DEC), who were acquired by Compaq, and then by HP.

      Clariion was a DG product. Even today, an EMC Unity LUN identifies itself as DGC

      1. GrumpyOF

        Re: Is it my imagination


        thanks for sorting me out with my "D"yslexia. Always seem to mix up Data General and Digital Equipment Corporation, although both were very good and competitive companies when they existed.

    3. StorageBuddhist

      Re: Is it my imagination

      MSA was Dothill I think. XP is Hitachi. EVA was great in it's day, but Carly's layoffs broke it. HP does not seem to have invented since the Compaq acquisition in 2002. 3PAR was solid and got them out of a hole, but it was already ageing when they bought it. At least this time Nimble is a much newer cleaner technology.

  7. Anonymous Coward
    Anonymous Coward

    I obviously work for HPE. And I'm excited about this acquisition ... not because 3PAR needs replacing but because Nimble are successful in areas we're not and have some great complementary features that can be shared across the products,

    Age – as someone else said, 3PAR is comfortably the 'youngest' of the most successful arrays around, and needed no rework to move to flash. This is not an acquisition to replace 3PAR. Other arrays are catching up, that's fair, but there isn't a competitive array with a standout advantage.

    Market – 3PAR is flying in the midrange and enterprise space. Nimble are flying in the midrange and entry space. We actually rarely lose to Nimble as we rarely come across them. When one of us does lose to the other, it's tended to be because we're out of our relative comfort zone.

    Customers – Nimble has been gaining customers - but it's not been at the expense of HPE (not that I've seen, genuinely). Nimble's Enterprise team (couple of guys) have been struggling on all accounts, and HPE have struggled to get into the low to mid-range space so this is a great fit.

    Technology – Infosight is meant to be a real standout product! Can't wait to get it in the portfolio.

    Price – Not seen much commentary that HPE overpaid. Seems like a good deal to me.

    1. returnofthemus

      This is not an acquisition to replace 3PAR


      Yes it is, you just don't know it yet ;-)

      1. Anonymous Coward
        Anonymous Coward

        Re: This is not an acquisition to replace 3PAR

        It's not a replacement. It fills the gap for sub 100K deals. 3PAR margin is terrible in those scenarios. Nimble is 50 points or higher.

        1. returnofthemus

          Re: This is not an acquisition to replace 3PAR

          "Nimble is 50 points or higher".


      2. Stand001

        Re: This is not an acquisition to replace 3PAR

        And you do know it, interesting, can you share your background?

        1. returnofthemus

          Re: This is not an acquisition to replace 3PAR

          No, not something I'm prepared to do in a public forum, though what I will say is that as part of the channel when we go to market with our preferred storage partner, we've had very little difficulty in displacing 3PAR as part of our overall solution.

          That being said storage probably represents less than 2% of our business, so it would be fair to say we have a very limited view. However, I'm sure no-one would dispute that it's a congested market, in which quite frankly the margins are severely stressed, making it ripe for consolidation.

          The fundamental point being cost has become the main criteria and assume that this pending acquisition acknowledges that. In turn this should help restore those margins and reinvigorate HP's channel, so while 3PAR may co-exist in the short-term, in the medium-to-long term it would be foolhardy for HP not to take full advantage of the economics and Nimble's sticky IP, which is exactly what the channel will be doing, regardless of any pending product positioning strategy.

          Of course HP may well attempt to apply Nimble IP to existing 3PAR kit, but then the phrase, 'putting lipstick on a pig' comes immediately to mind.

          1. Anonymous Coward
            Anonymous Coward

            Re: This is not an acquisition to replace 3PAR

            "what I will say is that as part of the channel when we go to market with our preferred storage partner, we've had very little difficulty in displacing 3PAR as part of our overall solution."

            Based on your posts elsewhere that would be IBM then ? which makes the axe you're looking to grind understandable, but your assertions of displacing 3PAR much less plausible.

    2. returnofthemus

      Nimble's Enterprise team (couple of guys) have been struggling on all accounts

      Oh the arrogance, reminds me of IBM back in the 80's

      Can't imagine those 'couple of guys' being too entusiastic about being presented with the 3PAR portfolio though, LOL!

      1. JRW

        Re: Nimble's Enterprise team (couple of guys) have been struggling on all accounts

        JohnW from Nimble here

        OK, I'll take the bait on this one. I am one of the Nimble Enterprise guys. Nimble is a great solution for Enterprise customers as, depending on requirements, 3PAR and XP can be. I should know, I helped write 2 XP exams and the first HP 3PAR exam. (Apologies to anyone who had to sit them, they were not easy.) I was excited about 2017 before the announcement. When the deal closes I'll post on LinkedIn my views but since I digested the news I've been going round with a big smile on my face, as have the other Enterprise folk here.

        1. Anonymous Coward
          Anonymous Coward

          Re: Nimble's Enterprise team (couple of guys) have been struggling on all accounts

          Was this the K34 exam by any chance? If so, then what a bleeding mess that was.

          1. JRW

            Re: Nimble's Enterprise team (couple of guys) have been struggling on all accounts

            JohnW from Nimble here

            Not K34, it was J49 (I think) on 3PAR and J37 and J47 for the XPs. J49 suffered a little from everyone still trying to get their heads around 3PAR. Not a bad exam as such but probably too much simple/dull material in the syllabus, so lacked focus. J37 & J47 were tough for different reasons. They were aimed at admins, service staff and pre-sales, making them tough for any of the target groups. J37 was better then the exam that went before and J47 improved again, but still unfriendly exams.


    HPE - Nimble Acquisition

    I like the article, and broadly agree with the author on the storage architecture life cycle, however, I think 20 years is a historical reality and not a rule of thumb for the future. Major changes in storage technologies have been very few over the last 30 years (probably SCSI to FC being the biggest change). I think the new cycle will not be beyond 10 years.

    This raises the question on strategy and whether storage vendors are innovators or acquirers of technology and IPR.

    I do not believe that any large established vendor in the storage market has executed a strategy based on a long-term vision other than Dell EMC. The use of ASICS rather than CPU's was something EMC engineering did two or three generation ago and returned to CPU based systems based on what the author identified correctly as being a limitation on ASIC development cycle time.

    My big issue with many storage vendors is around vision & strategy and HPE and NetApp don't have one that lasts more than 12 months.

    1. returnofthemus

      Re: HPE - Nimble Acquisition

      "I do not believe that any large established vendor in the storage market has executed a strategy based on a long-term vision other than Dell EMC".

      That' a joke, right?

      I know they've got the most bloated storage portfolio in the industry, but that's by default, not by design!

      The acquisition in question is much easier to summarise, Nimble needed bailing out and HP needed to refresh it's storage product line.

  9. Androgynous Cow Herd

    This was a smart deal for HP

    And maybe the best that Nimble could hope for....given that Pure and Nutanix got spanked by the street after beating the numbers and providing realistic guidance for next quarter, Nimble was staring down the barrel of a stock back down to $5 or maybe lower. HP got a ton of very valuable IP for a bargain price. Dell paid more EqualLogic and that certainly went well for them, and Nimble is a greatly superior product.

    I believe there is a perception that 3Par is a more "Enterprise ready" platform, and Nimble folks probably should learn to embrace that least for a while. The reality is that the feature gap between the two is pretty small and as the article points out, Nimble does not have to develop ASICs to introduce new features, whereas 3PAR and Nimble teams will have to figure out how and what to poll off of the ASICs for Infosight telemetry.

    More interesting (and a whole lot easier) will be integration of Openview with Infosight, and telemetry instrumentation added to SGI and Proliant servers. Add in the extensibility that already exists to monitor an entire virtualization stack from VMs discrete ports down to block storage with the ability to leverage predictive analytics or automation and you are on your way to the self-driving car in the data center.

    THAT will be easier and have more value to the customers than trying to insert infosight sensors into 3Pars code base and Asics.

  10. Anonymous Coward
    Anonymous Coward

    I'll bite as well. Current Nimble employee - I guess soon to be HPE.

    I've had a little time to digest this all. I think this is a positive for both Nimble and HPE. Innovation can be very challenging for large companies in a fast moving and highly competitive market. I think the article does strike a cord in that the larger the code base and more complex (feature rich?) the architecture, the harder it is to implement significant changes to said architecture (saw this at NetApp ... there is only so much you can do ... which will get you by for a while, but every time you are forced to "innovate" it gets harder). More so than an architecture - more important is technical thought leadership at a company. I saw a large part of the brain trust at NetApp walk out the door in recent years - to be replaced by nothing (sad, because I liked working there at one time). It doesn't matter how new or old your architecture is if you don't have anyone with the smarts to innovate in a real way.

    Mind you, 20-30 years essentially encapsulates the most significant part of the modern computer age. We don't have much history and lessons that have been learned based on the previous 20-30 years before that (punch cards and mainframes anyone?) --- at least as it relates to the exponential age of invention we are entering into now. But I digress.

    I think short to mid term HPE + Nimble means that HPE fills some holes in it's existing storage product portfolio. Longer term, between 3PAR, Nimble and even Simplivity - HPE gains a lot of very interesting IP that can be leveraged into the future - if done intelligently (that's the real risk). Nimble has a lot of runway left and putting the HPE investment and sales force behind it is going to be an interesting thing to watch.

    InfoSight is truly a great product. Many at Nimble might say you can't extend it into other storage systems --- but realistically **you can** extend InfoSight to anything that is willing to produce data. Now, that being said - the point that Nimbler's will make around this - and why Nimble arrays are different - is that the number and type of data points being produced by the Nimble architecture was designed into the product from day one. This isn't just "your PSU is running hot" or "X number of IOPS have occured" ... it is much more than that. Every storage system has stats on many things --- shipping that information efficiently to a tool like InfoSight can be challenging (e.g. NetApp was truncating Autosupports simply because they wanted to include a few basic performance stats to help with initial troubleshooting). Nimble arrays provide a **complete** history of all information from the array. That isn't a trivial task. Combine that with the back end data science and machine learning across the install base and you have what is the real value of InfoSight. Not to mention the awesome tools that are available not only to customers/partners - but internally as well. VM Vision and more shows how you can extend the capabilities of InfoSight up the stack. I look forward to seeing new and expanded capabilities in InfoSight and for it to become even more than it already is today.

    As for the other storage startups. We are in a unique time. The list of those willing to buy a storage startup is tiny. The likes of Tegile and Tintri I can't see ever "making it" ... and aren't innovative enough to be worth buying IMO. Pure is a one trick pony with a seriously good marketing engine (perhaps kill the storage products and just go be a marketing firm) - someone might buy them but I can't imagine who (it isn't going to be Cisco, I can almost guarantee that). If anything, this HPE purchase of Nimble is a solid kick in the kidneys to Pure (admit it or not). Pure has the same problem Nimble had ... to make the money you are making you have to spend more. Yeah, the gap might shrink a tiny bit each quarter ... but not at the pace necessary to make for a profitable company before a - through the roof burn rate - caves the roof in ... but, we live in crazy times, so you never know.

    I've blubbered on long enough. Call me cautiously optimistic.

  11. Anonymous Coward
    Anonymous Coward

    David Scott

    HP Storage Strategy died when David Scott left HP, the fact the current GM was previously responsible for HP HCI and Entry storage products and the company ended up acquiring Simplivity and Nimble says it all.

    1. Anonymous Coward
      Anonymous Coward

      Re: David Scott

      I think HP abandoned there own Smart Array Raid ASICS in favor of LSI devices on their blades and Dl380s and cut the entire staff years ago

  12. Anonymous Coward
    Anonymous Coward

    CPU vs ASIC

    Disclaimer, HPE 3PAR guy here.

    Article should point out that besides the ASICs, the 3PARs make plenty use of the 4/6/8core and even dual 8core Intel chips (per node), depending on the array model. The ASICs are still vital and very relevant to the architecture, but I do think now that AMD is back in the game, things are about to get interesting and more competitive, both price and performance wise.

  13. Anonymous Coward
    Anonymous Coward

    ETA: Nimble availability @HPE?


    when will Nimble be available with HPE part number? When will we be able to buy HPE/Nimble storage


    1. Androgynous Cow Herd

      Re: ETA: Nimble availability @HPE?

      Right after the acquisition is final. But you can buy Nimble storage RIGHT NOW!!!

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like