So on Monday 27th March 2011 Hitachi Ltd announced that it was effectively selling its disk manufacturing business (Hitachi Global Storage Technologies) to Western Digital (WD).
Could be very significant for enterprise disk architectures. Read on to see why…
On the one hand I’m surprised. When you consider the colossal amounts of data we’re storing these days, surely there is money to be made out of their disk drive business going forward. After all, disk drives are key in the storing of this data, and there are no signs of this trend slowing.
On the other hand, IBM offloaded the business to Hitachi not that long ago, so IBM clearly didn’t see it as an easy profit making business either.
Keeping up with the Joneses
There is no doubt that the R&D required to keep pace with the competition (Seagate, WD and Toshiba) is immense, and I can only think that Hitachi Ltd don’t feel up for the fight.
Interestingly, as we all know, Hitachi is a huge conglomerate, manufacturing things from TV sets to Bullet Trains, to Nuclear Power Plants, whereas Seagate and Western Digital are both specialists in Disk Drive technology.
On the topic of specialising in disk drive technology, I’m a bit of a fan of WD and like what they have pioneered in 4K Sector Advanced Format drives. If you don’t know about it, go check it out.
Enterprise Drives – the Big Question
In the enterprise disk space, Seagate and Hitachi were the major players, with Seagate having significantly more market share than Hitachi. However, enterprise disk drives account for a tiny fraction of drives shipped, so must be an expensive side of the business to run.
For me, the big question is, will WD be up for the fight with Seagate?
IBM and Hitachi clearly weren’t.
So, what happens if WD decide not to compete in this space?
If Seagate became the sole supplier in the enterprise space, they could effectively charge whatever they want. Competition drives cost down.
Competition also drives innovation, so where would the drive for innovation come from if this were to happen?
On the other hand (again, if WD decide not to scrap with Seagate in the enterprise space) we could see the end of the enterprise drive.
After all, we’re already seeing commodity type drives making their way into the enterprise space. 7200rpm SATAII drives are becoming very common place now (I appreciate these are not exactly the same as the drives we have in our laptops etc). It’s no secret that SSD drives sucking up much of the oxygen that 10K and 15K drives used to take for granted. So with slow SATA drives are squeezing at the bottom and SSD squeezing at the top, life is not so cosy these days for our trusted traditional 10 and 15K drives.
I personally think that SSD is already killing the value prop for 15K drives. I also think that NL-SAS will up the pressure on 10K drives. Add my opinions to the hype we get from array vendors, it may not be that long before our enterprise storage deployments are comprised of only two tiers – Flash and Trash.
SSD Will Not Take Over the World. At Least Not Any Time Soon
It’s my opinion that that SSD is not going to replace our beloved rotating rust any time soon, if ever.
My opinion is nothing to do with $/GB, Watts/GB, performance or any of the usual arguments. It’s far more basic than that, I’m lead to believe that there just aren’t enough raw materials or fabs in the world to do so. It’s my opinion that while HDD’s will of course one day become extinct, they aint going nowhere in the foreseeable future.
Despite the fact that Hitachi Ltd owned both a drive manufacturing business and an enterprise array business, they never leveraged any potential advantage this could have brought. Not surprising I suppose, but from a technologist perspective I’d have liked to have seen this. If they couldn’t leverage advantage, what is the advantage of having the business?
In the same way that drives are commodity, I wonder how long before arrays become commodity (providing only hardware RAID type functions) with so-called intelligence such as replication, copy services, thin provisioning, dedupe etc residing further up in the application/hypervisor…? Just speculation of course.
Good luck to Western Digital.
If WD decide not to compete with Seagate in the enterprise space, I think we could see the death of our trusted enterprise drives accelerated.
NOTE: Hitachi will maintain 10% stakehold and two seats on the WD Board of Directors.
Thoughts and comments welcome.
You can follow me on Twitter, I’m @nigelpoulton. I had a great conversation with several folks last night relative to this acquisition.