You are here: Home / Featured / IBM buys Blade Networks – Cisco keeps its buddy ?

IBM Buys Blade Networks – Cisco Keeps Its Buddy ?

28th September 2010 by 4 Comments

IBM buys Blade Networks

IBM has ìrun for the stackî as customers ask for end to end guarantees for servers and networking in the data center. With Cisco and HP playing the FUD card and telling customers that only a single vendor solution can guarantee the next generation of blade servers and networks, IBM has been forced into delivering their own switch.

Why note Brocade ?

Hard to tell, but some guesses:

  • Brocade bought Foundry two years ago, and still no new products of note. Ergo, their ethernet products aren’t up to much.
  • Brocade’s share price is falling most of the time. Thus everyone else thinks they haven’t got much
  • Brocade continue to think they are market titans and didn’t want a reasonable deal.

IBM isnít returning to Networking, yet

What seems clear is that IBM does NOT want to build a networking division. By purchasing the smallest switch vendor that is focussed on just one tiny market segment for a rumoured $400 million, IBM arenít taking on HP and Cisco. So Uncle John getís his wish to keep selling Cisco gear to IBM customers, but with a little bit of fear that IBM could change their mind anytime.

Thatís a good way to keep the business relationship level, donít you think ?

IBM Partnerships

Blade Networks makes switches that live inside the blade server backplane, and some top of rack, but not so much the backbone. Although, it wouldnít take much to make a simple Layer2 switch with lots of ports, itís all the rage with the VMware kids right now ( until they find out it doesnít work quite the way they expect ). Oh, and HP has been kicking Blade Networks out of bed since buying 3Com and thus half (?) their revenue was going away. So, IBM will continue itís ìlove inî with Cisco, selling Cisco WAN and Security gear, and packaging up juicy profitable services on all of Ciscoís satellite product divisions such as IP Video Surveillance, Smart Grid, Teleconference etc.

What does it mean ?

I think Ciscoís Nexus 7000 switches become further niched into the VCE stack and have less relevance in the open market. Without HP and IBM using them to connect 10GbE ports from servers, sales volumes are going to be low. And the Nexus 7K is not ready for 40GB / 100GB due to silicon problems (possibly both manufacture and design issues due to slow corporate process). This means that innovation on Nexus7K will slow down as product investment will match sales volumes. Impact ? The NX7K will get even further behind the innovation curve.

Since they haven’t upset Cisco, IBM will continue to sell professional services as their primary focus and hardware sales remains as loss leader. Hur, what else is new.

And we will wait a few more weeks to see the next consolidation move in Networking Industry. HP needs a firewall / security product of some sort, and maybe will buy a wireless company to own the technology. DELL needs an entire networking stack – at least a data centre partner such as Force10 Networks.

The EtherealMind View

Cisco has been vulnerable to this sort of attack for some time. Making large profit margins meant that other companies were going to attack them and their customers eventually. What isnít clear yet is what the new Cisco will look like ? Will Cisco look like IBM with a full stack of sales and professional services (and limited resellers partnerships) in lots of vertical markets ? Or more like HP with a ìweíll sell you anything at a good priceî ( with lots of reseller partners ) in any market ?

And what happens to resellers and customers who are caught in the market transition ?

Cynical Note

Interesting times perhaps but still not as interesting as the last Networking consolidation in 1998. This is just the same thing over again, with bigger numbers.

Filed Under: Featured, Musing Tagged With:
About Greg Ferro

Greg Ferro is a Network Engineer/Architect, mostly focussed on Data Centre, Security Infrastructure, and recently Virtualization. He has over 20 years in IT, in wide range of employers working as a freelance consultant including Finance, Service Providers and Online Companies. He is CCIE#6920 and has a few ideas about the world, but not enough to really count.

He is a host on the Packet Pushers Podcast, blogger at EtherealMind.com and on Twitter @etherealmind and Google Plus

  • http://wikibon.org/blog Stuart Miniman

    Greg,
    I agree with most of what you said about why the BNT acquisition makes sense for IBM, but not the impact on Cisco. I don’t see how what a primarily embedded switch has to do with core switches like the Nexus 7000. Over the last year, almost every end-user that I’ve talked to is deploying 10GbE at the core and Nexus 7000 seems to be the default product. I can see how the Nexus 2000, 4000 and maybe 5000 are affected, but not the 7000. IBM is not looking to go to war with Cisco (unlike HP). Do you really think that the server solutions have such a large impact on networking sales? I see the integrators, consultants and channel as a much bigger piece of the networking ecosystem. It seems like a big stretch to say that Cisco is going to abandon investment in this product line – do you have some details to back up “even further behind the innovation curve”? I think any of us covering the networking space are looking to see possible weakness in the market leader, but I’m not seeing much supporting information on that in your post.

    • http://etherealmind.com Greg Ferro

      Two strands of thought here. One, internal and top of rack switches mean less ports on core switch. Second & more specifically what I was alluding to, IF Blade Networks ships a big switch for collapsing the server connections (say, bigger than top of rack) then that means less ports on the core switch such as the nexus 7000. Both scenarios means less sales for Cisco at inflated profit margins and undermines their overall profitability.

      In Cisco, once a product loses profitability it stops being funded and starts to whither. That’s a risk right now, and the Nexus7K is already struggling to gain traction and meet the expected sales targets.

  • http://www.staticnat.com Josh O’Brien

    What I find interesting about this acquisition is that they have chose to ignore the Nexus 4000. IBM was the only vendor to adopt the 4000 to begin with and now with this acquisition they have no reason to continue down that path. This takes a big piece out of Cisco’s end to end data center plan and will now force Cisco to rely on its UCS sales to push their end to end DC plans forward.

    As for biting into TOR or DC aggregation switching, I really don’t buy into that. HP, Arista, Juniper and maybe Force10 but people don’t rely on IBM for any type of networking outside of the Blade systems now and it would take a long time for them to do so. What is left open in my mind though is how will this enable IBM to be a true network player in the emerging network world of virtual switches, firewalls and routers.

    We have yet to see anyone toss the old hardware model out the door and focus strictly on servers with hardware assisted networking such as the Cisco PALO cards. But as we have seen with the Nexus 1000v and the new firewalls from Cisco and VMWare that day is coming. IBM if they decide to could use this purchase to not only enter the networking market in a major way over time but more importantly change the paradigm of the networking world by offering some sort of virtual unified networking platform.

    • http://etherealmind.com Greg Ferro

      HP also chose to ignore the Nexus 4000. At the time when they bought 3Com, Cisco had the NX4000 ready to manufacture it just needed signoff from HP. That never happened.

      It becoming clear that Cisco will not be the ONLY networking vendor around, and integration of various vendors is back on the agenda for Network Professionals. And that is a good thing.