SDNchat

the customer - increased choice, leads to more competition, leads to *hopefully* better products/services

Good answer Jake!

The dominant incumbent will not gain share through SDN, so SDN represents only downside risk for Cisco.

For all the incumbents, SDN increases competition, which will make differentiation more important. Expect downward price pressure.

Stuart Miniman

@mbushong will SDN decrease the % of spend on networking overall for IT? server virt pushed $ to storage+networking...

Should add that Cisco can mitigate risk, so that's not to say they necessarily lose.

speaking of incumbents, this to jake - the latest Gartner DC MQ placed vmware in for the first time, and over to the right. what are the implications of this? first SW-only vendor in the MQ.

@stu I think IT spend tracks with GDP more than anything. Over time, it might shift, but most budgets are zero-based from previous year.

@stu So I expect that there is not major movement in overall spend. For an increase, you would have to have corresponding decrease elsewhere.

I think it will enable more spending and investment to be shifted to innovative applications, especially analytics

Like the Web mode, but making Web infrastructure open and cheap, an explosion of apps

agreed, but there are many vectors here - opex vs capes, HW vs SW, managed vs unmanaged, lightweight apps, etc.

That money comes from somewhere. So we have to flag where the spend moves from. My guess is you see pricing drop and $$ diverted to software.

@daveginsburg OpEx is rarely saved (except hard costs like power/cooling). Effort gets shifted elsewhere, no?

Agree with Michael. Commoditized HW enables software innovation

I see handful of startups winning but incumbents will rise fast when they see the "straight and narrow" path

so back to the original question - i believe both will survive - it is a large market. maybe some incremental changes in share, but 5 years from now the market will not be fundamentally different wrt the incumbents.

The history of paradigm shifts is not kind to incumbents. See: DEC, Nokia, Microsoft, Novell

i may be overly optimistic, but i believe that the valley networking companies are 'smarter'.... can learn from the past and adapt. they also have better records wrt M&A to ingest new technologies and expertise.

Rayno26

I'll repeat my Q1 as well: What's the top goal of #SDN? 1) Automation 2) Flexibility and Openness 3) Reduced costs (opex & capes)

5 Votes Vote

[+] Show Hidden Comments

@mbushong on Q1. Makes sense. But service agility part of the picture for SPs, yes?

depends on who is using it :) IMO, its automation

if i group analytics into automation, then yes. but customers already expect C&O reductions as a given.

Thanks Jake. Anybody want to bet on something else?

I think you see Capex reductions on new technologies because new SDN vendors don't need to recoup costs as much as incubment vendors, hence lower intiial CAPEX from new SDN players...

scott - i remember you saying something about current VC SDN investments and exits?

So, 40GE and 100GE SDN platforms are WAY cheaper in SDN versus incubments - not so much on 1 GE and 10 GE

@daveginsburg I have a report coming out on #SDN startups: More than $800M already invested

Already $3B in venture capital exists in #SDN startups. More to come. I predict at least $20B in value creation

jake - how are they cheaper? the base HW costs from BRCM or INTC are the same. the CPUs are the same. it is just a matter of corporate margins, open optics, and protecting existing lines of business.

Exactly @daveginsburg you're reasons are why the SDN new entrants like Pluribus are cheaper than HW incubments

Yes, @JakeHowering I have the same question

But you make a point that is a premise in my report: Startups are driving this market.

RT @mbushong @daveginsburg @rayno The cost of a device is not just the hardware components. The SW costs matter too.

@JakeHowering is open an issue if automation, flexibility and cost reductions happening; if complexity is abstracted away what the issue; hence white box purpose built works

agreed on the SW front, and you get what you pay for. admittedly the level of sophistication from the likes of cisco or vmware and the $$$ put into this are very different from a cumulus or pica8.

@furrier I think they are all part of the same thing. Open technologies reduce costs, drive flexibility

Rayno20

Q3: Where are we in the rollout of #SDN today, in terms of innings 1-9?

[+] Show Hidden Comments

if i look at a 20 year tech lifecycle, and we've been discussing it for 5 years, then 25% in

It feels like a "rain delay" to me at the moment. Nothing jumping out at me but see lots of movement. Great question

LOL @furrier, that is funny. Rolling out the tarp or rolling it up?

sharks could have used an 'ice' delay :(

A Valley VC point out to me that server virtualization took 11 years to achieve 60% marketshare.. SDN may be similar

I think we are in 3rd inning. Changing pitchers though. Pitcher got ejected for "substance" on his arm

Arista IPO could draw more attention to this space. Any thoughts on that?

Arista will pull money in, and will likely make fundraising for all those companies founded in last 3 years easier.

also the #box IPO got pulled so people think cloud is stalled

SDN will be faster than server virt largely because incumbents are backing it now (having learned from prev exp)

@furrier Pulled because of cloud or because of lawsuit?

@furrier Disregard. Thought you were talking Arista.

Anurag Agrawal

Our research finds that 11% of mid-market businesses and 3% of small businesses globally have heard about SDN. In the US the awareness jumps to 19% among mid-market businesses. SDN still has a long way to go.

@mbushong I don't know maybe both; the S1 was very telling bc their economics are not good; they spend 1.40 to get 1.00 in sales that is not sustainable