EMC Admits NetApp is the Leader

Chris Mellor of The Register recently published the article EMC races to catch up with NetApp, in which he covers a recent EMC webcast where EMC’s President and COO, Pat Gelsinger admitted EMC is hoping to catch up to NetApp in delivering storage efficiencies for virtual server and desktop environments.

I agree with Chris when he summarizes the current challenges facing EMC as…

“If EMC has misread the situation and NetApp’s lead lies in its current feature superiority, such as block-level primary data deduplication, as well as in the general smoothness and integrated management and feature sets across the FAS array range, then NetApp will continue to enjoy its long run of above-average results; leaving EMC scratching its head and wondering just what the hell it has to do to rein in the Sunnyvale upstarts.”

UPDATE 2/11/11: I obtained the transcript and below is likely some of the core content that Chris is referring to in his article…

“Obviously you can look here and you can see there is one feature that NetApp, in this case, has that we don’t have. Block or primary block; a DDUP. And in that one feature today, we don’t support it in our product line. We will deliver that in the second half of this year, so we’ll close that last piece of competitive gap that we have. But collectively, it’s just another compression or another data-reduction technology” – Pat Gelsinger, VP & COO EMC

(If you are interested obtaining a copy of the transcript from the webcast go to www.callstreet.com or call 1-877-FACTSET)

Mr. Geslinger states NetApp has a technological lead, which EMC will try to close, yet I’m not sure he gets it. When discussing NetApp, Mr. Gelsinger states that the only advantage NetAp offers is the ability to run production workloads with the storage savings of data deduplication. The reality is NetApp is delivering much more than storage savings to virtual infrastructures.

I found the comments Mr. Gelsinger shared following his ‘EMC will close the gap’ statement rather interesting as he suggested compression and deduplication were merely alternative means to achieve storage efficiencies on par with that of deduce. This statement is only about 50% accurate. While both technologies provide capacity savings, dedupe is more ideal for all production use cases. It appears Mr. Geslinger is either misinformed as to the similarities and differences in these capabilities, or is attempting to obscure said differences in order to not raise questions around the ‘EMC 20% Storage Savings Guarantee.’

I truly believe Mr. Gelsinger simply misspoke, and never intended to close with a message that was in stark contrast to his previous statement regarding the direction of EMC’s engineering.

Mixing Messages Around Efficiencies

I’ve written a number of posts intended to provide some clarity and increased understanding to the unique storage capabilities NetApp provides around storage efficiencies and performance results with these technologies. Suffice to say these posts receive a significant amount of negative feedback from the like of EMC, HP, and others primarily claiming that all storage vendors offer similar technologies and capabilities.

Based on Mr. Gelsinger’s comments, I think its fair to say that EMC senior leadership has validated the information I’ve shared in posts such as these.

There’s Much More Than Dedupe For Virtual Infrastructures

The conclusion Chris came to wasn’t solely based on storage efficiencies, as a storage analyst Chris is well aware of the core similarities and differences between storage arrays from NetApp and other vendors. His summary reflects this view, and suggests EMC has areas to catch up beyond storage savings in order to increase the value provided to virtual and cloud architectures.

The comparison of dedupe and compression is analogous comparisons that can be made regarding NetApp’s unified storage platform and EMC’s unified storage array.

NetApp offer’s unified management and functionality across all of our storage arrays, while EMC offers an array platform (well actually three with NS, VNX, and VNXe) which provides multiprotocol access with a very different features set and management tools from their other numerous platforms. The advantages provided by a unified storage platform span a wide spectrum ranging from simplified technical training, to standardized architecture designs & operational practices, to consistent features sets, points of integration, etc…

Over the past year I’ve authored a number of posts on NetApp leading the storage industry with capabilities and solutions specifically tailored for shared, virtual, cloud-based infrastructures.

I’ve attempted to highlight where ‘the rubber meets the road’ in these posts covering a number of topics such as unified storage architecture, performance, VDI, and joint solutions from Cisco, NetApp, and VMware. I would suggest to you that these articles are as accurate in their content as the one’s I’ve written on storage efficiencies… even if NetApp’s competitors disagree.

After the dust settlers on this discussion, I doubt we will hear much more about NetApp from EMC. From a sales & marketing strategy, it is to the advantage of the incumbent to ignore the challenger they fear to battle. Doing so calls out a comparison of the tape, and frankly I’m with Chris on this one… they don’t measure up.

UPDATE 2/13/11: It appears my agreement with the conclusion in Chris’ article has raised the ire of Chad Sakac, VP at EMC and some others. Much like this genesis of this post, i’d suggest you check out the opinion of another storage analyst, Chris Evans (aka the Storage Architect) for his thoughts on this topic and the responses from EMC in, “You can’t trust what you read on the Internet – Unless it’s EMC“

5 Comments

Dave Lawrence February 10, 2011 At 11:13 am

I knew you would have to mention this after our discussion. Great to see you this week.

Dimitris Krekoukias February 10, 2011 At 12:20 pm

What the article fails to address are some of the other major shortcomings of EMC (and actually most other) gear:
1. High impact snapshots
2. High impact RAID6
3. Not a truly unified storage (regardless of marketing)
Dedupe is but one part of it.
NetApp systems perform well WITH dedupe and snaps and RAID-DP all at the same time.
D

Mike Shea February 10, 2011 At 2:21 pm

As usual, nice post. If we look back over time, we see *over and over* again how EMC has found itself behind NetApp, only to work to cob together some half-baked scheme to claim parity. Truth be told, if that is all that they have – ‘parity’, then they cannot claim leadership, the two don’t fit together.
If nothing else, the scoreboard shows a NetApp growth rate twice that of EMC. There is a reason for that, and you highlighted it perfectly.
Take out VMware from the EMC equation, and what is left over? Were I still and EMC serf, the very thought would make my blood curdle.
(BTW – we are hiring leaders who want to join us in the journey to the top!)

Michael Colby February 11, 2011 At 12:03 am

Vaughn, you failed to quote the first of two paragraphs of which the one you quoted was the second. Quoting the 2nd without also quoting the 1st misleads the readers of your blog by misrepresenting the context. You failed to include the quote which included that EMC “will bite NetApp in the ass.”
Your blog makes for good entertainment for the home crowd (and I read it for the sheer entertainment value, like I read good fiction) but your citation standards are suboptimal. I’m as much a fan of “bread and circuses” as the next guy, but I thought it would only be honest to provide the complete quote instead of just the portion you provided out of context.
For the record, the 2 paragraphs from the article in The Standard for the complete context:
If this really is EMC’s sole remain deficiency compared to NetApp for virtualised server storage then Viper’s product will bite NetApp in the ass. Netapp’s recent strong run of results in the virtual server and desktop areas, compared to EMC, will come to an end later this year, other things, such as VNX array functionality and pricing, being equal.
If EMC has misread the situation and NetApp’s lead lies in its current feature superiority, such as block-level primary data deduplication, as well as in the general smoothness and integrated management and feature sets across the FAS array range, then NetApp will continue to enjoy its long run of above-average results; leaving EMC scratching its head and wondering just what the hell it has to do to rein in the Sunnyvale upstarts.
be well my friend.

Vaughn Stewart February 11, 2011 At 1:15 pm

@Michael – Thanks for the feedback. I authored this post from the floor of PEX and suffice to say it reads as if I might have attempted to do so between numerous discussion with partners. 🙂
This morning I made some edits including providing a quote from Mr. Gelsinger and elaborating on a point or two.