Dell claims an A+ on blade strategy report card

Dell claims an A+ on blade strategy report card

Summary: The director of Dell's PowerEdge Server Product Group, Tim Golden, cites marketshare numbers from IDC and claims that it didn't take long for Dell's new blade strategy to devastate the competition.

SHARE:
TOPICS: Dell
17

Download this PodcastBy all accounts, even those of Dell's, when blades became all the rage in terms of a form factor for servers, Dell misfired and ceded the market to the other two tier one vendors HP and IBM, who had a bit of a field day while Dell regrouped.  Regroup it did. 

About this time last year, Dell began to drop hints that a new blade strategy was in the works -- one that by virtue of Dell's highly optimized direct model (what Dell insiders call the Dell Effect),  would turn the blade market upside down (so the company promised).  Because of the Dell effect, the company said it could do what no other blade vendor was doing -- save buyers money over other server form factors (eg: towers, 1Us, etc.)  on the cost of acquisition versus what other vendors were doing: charging a premium to get all that technology squeezed into a tiny package that eventually produced a savings over time based on other cost-crunching techniques such as better management and power consumption  

Then, in November 2004, Dell delivered the first tangible evidence of that strategy -- its PowerEdge 1855 Blade Server.  Now, a little more than six months later, the director of Dell's PowerEdge Server Product Group Tim Golden cites marketshare numbers from IDC and claims in my interview with him that  it didn't take long for Dell's new blade strategy to devastate the competition.   Since launching the new blade strategy, at least one leading blade seller -- blade pioneer RLX -- has exited the blade hardware market, focusing on blade management instead.  The audio version of the interview is available as an MP3 that can be downloaded or, if you’re already subscribed to ZDNet’s IT Matters series of audio podcasts, it will show up on your system or MP3 player automatically. See ZDNet’s podcasts: How to tune in.

In the interview, Golden does not mince words.  "There were a number of promises that we perceived that other vendors had made about blades that they weren't necessarily living up to. ... Blades were supposed to save you money.  That has not historically been the case."   Golden also argues that with other vendors, buyers might have to face as much as a 30 percent premium to get the space savings that blades offer over other rack mountable servers (1Us).   Meanwhile, he promises that buyers of Dell's blade servers will start to experience a savings versus similarly configured 1U servers from Dell once a blade chassis is populated with its sixth server.  By the time a chassis is fully loaded with 10 servers, the savings (on acquisition cost) is supposed to be 25 percent.  Whether or not Dell's blades can save as much as those of HP or IBM on the management and power consumption fronts was not discussed.  But, in response to my previous write-ups on the issue, readers have come forward to say that it's a marketshare bloodbath behind close doors and that IBM and HP are offering steep discounts off their street prices -- discounts that may bring their costs of acquisition more in-line with what Dell has to offer.  Nevertheless, the IDC data, which indicates that Dell grew in 2005Q1 at IBM and HP's expense (both retreated) could be validating Dell's strategy.

In the interview, Golden also had a bit of news and may have shown Dell's hand in some as-of-yet undisclosed announcements.   On the news front, where Dell previously offered a Brocade-based FibreChannel Storage Area Network (SAN) switch for its blade chassis, the company now has a similar offering that supports McData-based SANs.  Golden says the SAN market is in a near 50/50 split between the two manufacturers and that the new offering means that Dell's blades can now plug into just about any legacy SAN (whereas before, they could only plug into half: Brocade-based ones).   In discussing how Dell's chassis has a backplane that's been designed to accommodate faster versions of Infiniband and Ethernet (4 GB), I asked Golden if we could expect to see a Cisco-based offering any time soon.   Said Golden, "I wouldn't be surprised to see a Cisco switch show up there as well."  When I asked when might that be, he said "You won't have to wait too long."

Topic: Dell

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

17 comments
Log in or register to join the discussion
  • Did you look at their management software?

    Dell may have cheap hardware, but they are significantly behind their competitors in management software. They seem to have figured out they need to make strides in the 'Enterprise' arena, but they have a lot of ground to make up before they arrive at where Compaq was a decade earlier with CIM.

    In regard to SAN connectivity, can it really be a 50/50 split? Brocade has far more ports installed and running than anyone else. The competition is pretty heated in the director class, but not everyone is in the director market space.

    Our experience has showed that you are more likely to run out of electrical and thermal capacity in a rack space before you run out of physical space. We gain more flexibility and the systems have been easier to manage in a 1U rack mount form factor.

    ...but not Dell's 1U form factor.
    Uber Dweeb
  • 1 server blade and <10% market share = A+ ???

    That's why schools today email the report cards directly to the parents.
    TechBoom
    • Amen

      Amen to that, brother. Dave ol' buddy, stop drinking the Dell Koolaid...
      dm94133@...
      • 4 to 12 percent?

        IDC's report shows Dell quarter over quarter growth as having gone up from 4 to 12 percent while HP and IBM's q/q numbers are under zero (negative growth). Maybe "devastating" wasn't the right word because HP and IBM already have volumes that, even with negative growth, still exceed Dell's raw volumes by a lot. But if the trend continues, then I'd say Dell definitely has game.

        db

        ps: I've cut back on sugary drinks as of late. Especially Kool-Aid.
        dberlind
  • Can I get a fact check on aisle 12 please?

    Or how about a reality check.

    Dell may be migrating some of their own installed base who have or were threatening to go to IBM or HP blades - but devistating the competition? Together, HP and IBM represent 74 percent revenue share of the blade market, while Dell is a distant third with 9.7 percent.

    Dell taking credit for the fall of RLX is laughable, unless you believe the 3% of the market share Dell held in 2004 really hurt RLX that bad. Was that really the '04 strategy? The only vendor selling fewer blades than Dell last year was RLX. Did Dell also have a hand in the fall of the Berlin wall and the invention of the Internet?

    And is the "Dell Effect" like the Twilight Zone? Where Dell can enter and exit markets at will until they manage to sell something and failures like their four/eight-ways, Itanium servers, blades, management software, Linux - are simply glossed over? Attention, do not attempt to adjust the picture on your 42" Dell plasma TV screen . . .

    Dell does some good things for some consumers and mom-and-pop-shops, but datacenter-ready, they are not.

    Anyone want a cheap PC, err, server? Call Tim Golden at Dell. Want a blade, call HP or IBM.
    TechBoom
  • TRUE GRADE ON DELL = F----

    They get an F----- at all they do, computers, blades, servers, printers, etc.

    Smart businesses don't buy Dell!!!!
    ITGuy04
    • yeah right

      And no business buys Apple.
      jmcguire@...
      • So? Not like businesses do smart things...

        After all, NO INDEPENT STUDY HAS WINDOWS WITH THE LOWEST TCO. Yet they all blindly buy Windows?

        All I know is I used to spend less time babysitting Novell than I do Windows.

        I know my Macs work perfectly 99.99% of the time and have never let me down.

        I know I never reccomend Windows and have yet to hear anyone disappointed.
        ITGuy04
    • small minds

      The mom and pop moron has shared his wisdom with us.
      11bravo
  • Dell cures Cancer!

    Will that be the next ZD story?



    Buy a Mac and be done with it!
    Reverend MacFellow
    • Heh... have you seen the server Mac performance numbers....

      ...don't try to laugh too hard.

      http://www.macnn.com/articles/05/06/15/os.x.server.review/
      shadar
      • Read the comments on the article!

        You will find that the writers had no idea of what they were doing and results were highly flawed and have been discredited!
        Read a little more:
        http://www.macnn.com/articles/05/06/15/os.x.server.review/#comments
        An_Axe_to_Grind
      • Yeah, and you have no clue

        One badly written application with no testing at all. You'd get just as bad numbers with FreeBSD on X86.....

        Buy a clue.
        ITGuy04
  • We're exclusively Dell, and this is news to us

    I'm reading this and am a bit dubious of the claims in here. We contacted Dell awhile back about their blades - say 1/05 - and they were the most clueless of the IBM/HP/Dell triumvirate when it came to blades. So much, in fact, that we're considering ditching our "Dell only" strategy and going with HP stuff.

    The big advantage that IBM and HP have over Dell is their willingness to offer AMD blades. Dell is certainly correct that blade offerings don't necessarily offer a cheaper alternative until a certain threshold is reached, but I believe HP also claims it takes about 6 blades before the lines cross on the costs. HP currently offers switches from Brocade, McData, and Cisco, so Dell needs to mirror that in order to compete. Dell is still behind in that respect.

    Dell may very well catch up at some point, but I get the distinct feeling they are desperately trying to generate some major FUD here in order to make up a lot of the lost ground they gave away in the early days.
    ejhonda
  • Lies, Damn Lies, and Statistics

    The "blade server market" is fairly small, so any increase in sales looks like a HUGE jump in market share. The power, cooling, cabling and management issues have yet to be solved - so this market stays small. IMHO blade servers will never catch on - over 1U and large virtualized servers. So if I want a 50 server blade rack, I need 155 network drops (100 for dual connections, 50 for "RSA"-like management modules, and 1 for each of the 5 blade "racks"), 50 power drops, 100 SAN fiber connects (for dual-pathing), plus a few keyboards, mouses and displays (I'm ignoring the INTENSE heat that needs to be tamed here). So, now I'm standing behind this ONE rack that's blowing PURE HELL heat on me, trying to locate the one SAN connection that is failing - out of that bundle of 300+ cables. I guess I can console myself thinking about that 25% I was saving . . .
    Roger Ramjet
    • You've obviously never set up a blade rack...

      Blades are here to stay. Soon, virtualization will catch on, but it will only enhance the future bladed chassis... the 1U market will die before blades.
      shadar
  • nm

    nm
    TechBoom