Monday, 3 November 2008

Innovation

"Innovative - featuring new methods or original ideas - creative in thinking" - Oxford English Dictionary of English, 11th edition.

There have been some interesting comments over the weekend, specifically from EMC in regard to this post which I wrote on Benchmarketing started by Barry Burke and followed by Barry Whyte.

"Mark" from EMC points me to this link regarding EMC's pedigree on innovation. Now that's like a red rag to a bull to me and I couldn't help myself going through every entry and summarising them.

There are 114 entries, out of which, I've classified 44 as marketing - for example appointing Joe Tucci (twice) and Mike Ruettgers (twice) and being inducted into the IT Hall of Fame hardly count as innovation! Some 18 entries relate directly Symmetrix, another 18 to acquisition (nnot really innovation if you use the definition above) and another 7 to Clariion (also an acquisition).

From the list, I've picked out a handful I'd classify as innovating.

  • 1987 - EMC introduce solid state disks - yes, but hang on, haven't they just claimed to have "invented" Enterprise Flash Drives?
  • SRDF & Timefinder - yes I'd agree these are innovative. SRDF still beats the competition today.
  • First cached disk array - yes innovation.

Here's the full list taken from the link above. Decide for yourself whether you think these things are innovative or not. Acquisitions in RED, Marketing in GREEN. Oh and if anyone thinks I'm being biased, I'm happy to do the same analysis for IBM, HP, HDS etc. Just point me at their timelines.

  • Clariion CX4 - latest drives, thin provisioning?
  • Mozy - acquisition
  • Flash Drives - 1980's technology.
  • DMX4 - SATA II drives and 4Gb/s
  • Berkeley Systems etc - acquisition
  • EMC Documentum - acquisition
  • EMC study on storage growth - not innovation
  • EMC floats VMware - acquisition
  • EMC & RSA - acquisition
  • EMC R&D in China
  • EMC Clariion -Ultrascale
  • EMC Smarts - acquisition
  • Symmetrix DMX3
  • Smarts, Rainfinity, Captiva - acquisitions
  • EMC - CDP - acquisition
  • EMC Clariion - Ultrapoint
  • EMC DMX3 - 1PB
  • EMC Invista - where is it now?
  • EMC Documentum - acquisition
  • Clariion AX100 - innovative? incremental product
  • Clariion Disk Library (2004) - was anyone already doing this?
  • DMX-2 Improvements - incremental change
  • EMC VMware - acquisition
  • EMC R&D India - not innovative to open an office
  • EMC Centera - acquisition - FilePool
  • EMC Legato & Documentum - acquisitions
  • Clariion ATA and FC drives
  • EMC DMX (again)
  • EMC ILM - dead
  • EMC Imaging System? Never heard of it
  • IT Hall of Fame - hardly innovation
  • Clariion CX
  • Information Solutions Consulting Group - where are they now?
  • EMC Centera - acquisition
  • Replication Manager & StorageScope - still don't work today.
  • Dell/EMC Alliance - marketing not innovation
  • ECC/OE - still doesn't work right today.
  • Symmetrix Product of the Year - same product again
  • Joe Tucci becomes president - marketing
  • SAN & NAS into single network - what is this?
  • EMC Berkeley study -marketing
  • EMC E-lab
  • Symmetrix 8000 & Clariion FC4700 - same products again
  • EMC/Microsoft alliance - marketing
  • EMC stock of the decade - marketing
  • Joe Tucci - president and COO - marketing
  • EMC & Data General - acquisition
  • ControlCenter SRM
  • EMC Connectrix - from acquisition
  • Software sales rise - how much can be attributed to Symmetrix licences
  • Oracle Global Alliance Partner - marketing
  • EMC PowerPath
  • Symmetrix capacity record
  • EMC in 50 highest performing companies - marketing
  • EMC multiplatform FC systems
  • Timefinder software introduced
  • Company named to business week 50 - marketing
  • EMC - 3TB in an array!!
  • Celerra NAS Gateway
  • Oracle selects Symmetrix - marketing
  • SAP selects Symmetrix - marketing
  • EMC Customer Support Centre Ireland - marketing
  • Symmetrix 1 Quadrillion bytes served - McDonalds of the storage world?
  • EMC acquires McDATA - acquisition
  • EMC tops IBM mainframe storage (Symmetrix)
  • Symmetrix 5100 array
  • EMC 3000 array
  • EMC BusinessWeek top score - marketing
  • Egan named Master Entrepreneur - marketing
  • EMC 5500 - 1TB array
  • EMC joins Fortune 500 - marketing
  • SRDF - innovation - yes.
  • Customer Council - marketing
  • EMC expands Symmetrix
  • EMC acquires Epoch Systems - basis for ECC?
  • EMC acquires Magna Computer Corporation (AS/400)
  • EMC R&D Israel opens - marketing
  • Symmetrix 5500 announced
  • Harmonix for AS/400?
  • EMC ISO9001 certification - marketing
  • Mike Ruettgers named president and CEO - marketing
  • Symmetrix arrays for Unisys
  • Cache tape system for AS/400
  • EMC implements product design and simulation system - marketing
  • Product lineup for Unisys statement - marketing
  • DASD subsystem for AS/400
  • EMC MOSAIC:2000 architecture
  • EMC introduces Symmetrix
  • First storage system warranty protection - marketing
  • EMC Falcon with rapid cache
  • First solid state disk system for Prime (1989)
  • Reuttgers improvement program - marketing
  • First DASD alternative to IBM system
  • Allegro Orion disk subsystems - both solid state (1988)
  • EMC in top 1000 business - marketing
  • EMC joins NYSE - marketing
  • First cached disk controller - innovation - yes
  • Manufacturing expands to Europe - marketing
  • EMC increases presence in Europe and APAC - marketing
  • Archeion introduced for data archiving to optical (1987)
  • More people working on DASD than IBM - marketing
  • EMC introduces solid state disks (1987)
  • Storage capacity increases - marketing
  • EMC doubles in size - marketing
  • Product introductions advance computing power - marketing
  • HP memory upgrades
  • EMC goes public - marketing
  • EMC announces 16MB array for VAX
  • Memory, storage products boost minicomputer performance
  • EMC offers 24 hour support
  • Testing improves quality - marketing
  • Onsite spares program - marketing
  • EMC delivers first product - marketing
  • EMC founded - marketing

10 comments:

Mark said...

This is Storagezilla it's your login system which has me classified as "Mark".

This idea you have that incremental innovation doesn't matter is ludicrous there was more work done on the UltraPoint DAE than some competitors do on entire products. It took HP a number of years to reverse engineer it with their Cut Through Switch and they still got it wrong.

Yes the scientist in all of us thinks that it's only innovative if you're creating black holes but that's nonsense.

You also mistake previous Symm generations to be like each other. The DMX is nothing like the bus based systems which came before it so it isn't a case of introducing the same product just a bit faster. It's a new product with the Symmetrix name.

What were people using before PowerPath and Control Center? Mainframes? Before Replication Manager? Scripts?

What about tiered storage in the same box? Switch based virtualisation? VirtualLUNs? NDU NAS upgrades? Protected rollback. None of those made it into that timeline as they're features of products but they're all innovative.

There was no VTL appliance market before EMC created it, you did all the integration work and optimisation yourself. That's why EMC scooped up 60% of that market in less than four years because it pioneered the integration and manufacturing processes.

There was no CAS market or CAS product before or during the Filepool era, only after.

EMC did not invent Enterprise Flash Drives it was the first to test at scale, work with the partner to get it's manufacturing and quality processes up to speed and then ship products with full software support. Obviously so easy everyone else has an offering available by now right? Wrong.

And does anyone in this business have anything even close to the size of eLab?

To say EMC is not innovate is disingenuous and to write off something because it was an acquisition is a mistake since the people who had the idea in the first place tend to stick around and more stuff gets created after an acquisition than before as you're not just buying unfinished code which needs a good few months of work to make solid. You're buying the people who came up with it in the first place.

Chris M Evans said...

I should have guessed! Who else in EMC is usually awake 24/7 and online, woking over the weekend.

Chris M Evans said...

...that's WORKING over the weekend.

Vicente said...

Just as a disclaimer, I also work for EMC, but I wanted to chime in with a definition of innovation that I have used in the past:

To me, there is only one type of innovation that a company should pursue, and that is one that fulfills a customer need that was unmet and does it in a way that provides value to the customer. Whether it is via acquisition, process improvement or the creation of a new product altogether is a matter of existing constraints, available technology, budget available and expected ROI.

Vicente

hollis_chuck said...

Hi -- interesting discussion, to be sure.

So, this begs the question -- what really is innovation in this (or any other) technology industry?

Is it inventing something entirely new? Or being able to spot promising areas of technology, and investing in them wisely?

One of your counter arguments is "that doesn't count, that was an acquisition!". Well, given the high failure rate of most tech acquisitions, I tend to give EMC points for spotting the hot areas, paying big bucks, and making the results work in the bigger picture.

If one imagines that EMC started with a very small employee count (3 or 4 as I remember), then -- by definition -- just about every idea was "acquired" after that point, either by buying a company, or hiring an individual.

I won't recite the industry litany to you -- Apple didn't invent the MP3 player, but boy they really ran with the concept. Ditto with Microsoft and MS-DOS and Windows. Amazon was not the first online retailer. Google was not the first internet search engine. IBM didn't invent virtual memory, nor the mainframe, but they did more with the concept than anyone else.

On the other side, Xerox PARC came up with ethernet and the WIMP interface, but did nothing with it. Bell Labs begat UNIX, but couldn't figure out how to commercialize it.

Lots of woulda-coulda-shoulda stories in this industry as well, no?

OK, I'll quit now, but you get the point -- what really matters in the big scheme of things?

Indeed, what matters more? Being the very first on the planet to come up with an idea? Or being able to make it work for a very large audience?

I would argue that the latter is far more important than the former.

Cheers!

Martin said...

Storagebod here! I must defend Replication Manager, it works great in our environment! Wouldn't say it was entirely innovative but it's pretty good! I just wonder if EMC could do a version which supported NetApp's replication!!

Now Control Center! Yeah, that's innovative! How many products have made it to version 6 and still not worked properly!

Stephen Foskett said...

Chris,

Although I agree that lots of those marketing-type things don't count as particularly innovative, the following are definitely truly completely innovative, and EMC deserves a whole pile of credit for them. I came on the storage scene in the mid-1990s and saw many of them come, and I was impressed!

* Flash in DMX - Flash != Orion - this was probably the most important innovation of 2008!
* DMX - this was huge, a complete reinvention of the Symmetrix and a really important product
* PowerPath - incredibly important and truly innovative (but it came from an acquisition)
* Multi-platform FC - I remember this at the time and it was a serious change
* TimeFinder - no one would doubt the importance of TimeFinder and SRDF
* Symmetrix 3000 - enterprise storage for open systems was huge, and I watched it happen as a customer
* SRDF - see above
* Symmetrix - no need to say more
* Orion/Allegro SSD - I had one of these way back when, and they're impressive even today - served as the architecture for the Symm!
* IBM DASD - EMC's most important innovation, and the birth of the enterprise storage market

I say these things as an independent - I've never worked for EMC. The company has a serious history of innovation!

Even some acquisitions were innovative - VMware, RSA, and DG all looked questionable at the time but worked out well.

Stephen

Chris M Evans said...

Chuck

The original point of this thread seems to have been lost and that was to question whether EMC (BarryB) was right to discount IBM's Quicksilver project as "benchmarketing". http://thestorageanarchist.typepad.com/weblog/2008/10/1028-benchmarketing-badly.html being the original post.

Now we've gone full circle with your comments and those from Stephen and by the definition there's no reason why Quicksilver shouldn't count as innovation. Whether it leads to a new product or an improvement of a product is irrelevant. To quote Barry on Quicksilver:

"Hardly revolutionary, much less innovative or awe-inspiring.

It is, however, Benchmarketing."

Does that apply to EMC's announcements last year - Hulk and Maui? Were they purely "Tuccimarketing"?

If EMC/Barry is prepared to sling the mud then expect some of it to come back. I'm not defending IBM, far from it, they have bloggers who can do that themselves, however once in a while I feel obliged to comment - as do Storagezilla, you, BarryB, BarryW, Tony Pearson and others when they feel strongly on a subject.

I like the debate - keep it coming!

the storage anarchist said...

Wow - I honestly didn't realize that it was *I* that started up this whole innovation debate, until I read Chris' last comment.

At the risk of digging a deeper hole, what i was trying to say was that lashing together a bunch of unprotected flash devices behind a jury-rigged combination of servers-acting-as-storage, all fronted by a non-standard configuration of SVC nodes just to demostrate a "magical number" of 1M IOPS was nothing more than a marketing game...

There was nothing involving "new methods or original ideas" in this demonstration. It was truly nothing more than simple aggregation of I/O hardware, that in fact was unable to leverage more than a fraction of the performance of the fastest components. Heck - you could have connected a couple of TMS boxes directly to a p5 server and done the same thing.

That, and the demo proved nothing, opened no new doors. We have no example of a real-world application that would generate that many I/Os from such a small amount of capacity, nor that could tolerate a storage solution that offered no data integrity validation or availability protection. Nor do we have a reference of cost for putting such a kludge together if one actually could define a need for it.

QuickSilver was a best a clever marketing ploy - admittedly better than at least some of the Green Items in your list, even.

But innovative?

Hardly in the same class as any of the black items in your list. Heck - I'd say that both IBM StorageTank and IceCube were more innovative than QuickSilver...and BOTH of them have failed the test of time.

Ah well - 'tis truly a spirited debate...glad to have been part of it.

By the way: what color do you think Atmos is today, 3 days after announcement?

Chris M Evans said...

Barry, actually I've been so busy this week (feeble excuse I know) that I haven't yet looked at the specs online for Atmos, however rest assured that I'm on the case...