Got here a bit later due to giving feedback to a product manager on MyService360…when I came in, the room was doing introductions and topics were interested in. General topics of interest are around Dell/EMC/VMware, Software Defined Cloud/Storage, NFV – lots of familiar faces in the room.
I’ve sat in several of these sessions (invite only) and John’s perspective is incredible – he’s very candid and often goes off the record when giving background and stories (so I’m leaving out some details below to respect that as I really appreciate his candor).
Starting with NFV discussion – telecom companies can be very “historical”. Tier 1 telecom operators run around $10B/year in CapEx spend and around $40-50B/year in operational spend. Economic models changing due to people pruning off the high revenue sections – total loss of text message revenue which was $10’s of billions of revenue. Couldn’t compete with web scale companies, etc. that can deploy new services far faster than the telco’s…so telco’s said had to adopt agile/hybrid methods to roll out services. Interestingly enough, this looks like the Enterprise shift…but on a much more compressed timeframe.
EMC had to figure out how to engage with people on the revenue-producing side of the business when they’d been working with people on the “cost” side of the business.
Verizon was ready to kick out the whole federation (EMC, VMware, etc.) out at one point. EMC proved to them that they could be relevant – years back Verizon was looking at Cassandra, CoreOS, etc. for a 100 PB infrastructure that was projected to grow at 100 PB per month. EMC showed they could go in, focus on their problems first (using an inside skunkworks team for some custom development), map products to it with some custom adjustment. Turned from being kicked out to being critical in the platform 3 application infrastructure.
Moved EMC Smarts under John Roese from CJ Desai – focusing it on telco’s. EMC is doubling down on Smarts and John really wants input around the product.
Interesting competitive discussion about Huawei – I probably shouldn’t post much of this discussion. 😉 Huawei does well in markets that are more commoditized – not as well in markets where there’s a faster pace of innovation. John Roese interviewed about that here – http://www.zdnet.com/article/emc-unfazed-by-compeition-from-huaweis-young-enterprise-biz/
SDS/SDDC/etc. – have to split out the data storage products to follow the way that data is created/consumed….this is a huge shift from 5-10 years ago. In 5-10 years only about 6% of data will land on transactional storage, etc. – but that will be the most valuable data so worth focusing on. If the majority of data in the future will be landing on software defined storage, EMC doesn’t have enough software defined storage stacks….sounds crazy but true.
Discussing VMware & EMC conflicts – back to the “better overlap than gaps” strategy from Tucci which is still correct. John wants to see more strategic alignment – the term “Federation” has disappeared and it’s now “a collection of strategically aligned business”.
Total addressable market for ScaleIO (all spots where ScaleIO could go) is $8B – that’s a big TAM but it’s stratified across 5 very different segments.
Big shift in EMC messaging around cloud this year – moving away from messaging that you approach cloud by buying/selling a bunch of small components/collection of technologies. Rather you buy/sell a whole platform – packaged up and delivered to you ready to provide a cloud experience. Also explicitly excluded one use case – not going to build an at scale public cloud and compete with Google/MS Azure/Amazon. That means long term vCloud Air will move towards well… 😉 Original aspiration of vCloud Air was to be the “Amazon for VMware customers” but that has changed. Long term strategy is to disintermediate cloud providers who are racing to the bottom around selling abstracted hardware resources.
Amazon spends around $2B for CapEx per year – Dell+EMC will have a $40-50B CapEx supply chain per year. Will be the biggest purchaser of Intel processors, Intel chipsets, hard drives, SSD’s, etc. – could even have strong enough “most favored nation” pricing that might sell to other cloud providers.
Will likely sell a lot of DSSD to at scale cloud providers.
Microsoft is not interested in Windows – they want to win Azure and cloud applications (Office365, etc.). That does NOT compete with EMC…strongly complementary to what EMC wants to do.
Question from Ed Walsh – what will see in the next year specifically? Answer = 3rd platform converged infrastructure (see Project Neutrino announcement from keynote today). Will see customers that panic when realize they don’t have the right hardware/software stack to run 3rd platform apps…so if EMC can wheel in a rack of infrastructure to handle that will be a huge win.
Cisco & EMC relationship? Gotten a lot better in the last year – several off the record examples there. 😉
Awesome perspectives from John – always a valuable session for EMC Elect at conferences and really appreciate Mark Brown coordinating it.
That’s a wrap!