Final Mission Or Just A New Beginning?

Last weekend I watched a video on YouTube entitled “Ascent – Commemorating [the Space] Shuttle”. The video is a compilation of amazing slow-motion footage of mostly engineering shots of shuttle launches. During a launch there are hundreds of cameras of various types that record just about every aspect of the machine from every angle.

The footage reminded me why I love engineering: seeing all that stuff gives me the chills. The shuttle reaches orbit in 8.5 minutes, travelling at 5 miles per second. Clearly these guys didn’t take lessons from the people I see struggling to hit 40 mph on a freeway onramp. For cyrin’ out loud, what the thinkin’, it’s safer to go 40 as opposed to say, mergin’ in at the speed of traffic? Good lord.

There’s a lot of stuff like this in our lives, people travelling 40 mph to merge into 65 mph traffic, and space shuttle launches. At my age I try to focus on the positives. When I was younger I could watch a NOVA program about poachers killing elephants or gorillas, but no longer. Nope, now, I get 3 minutes into one of those documentaries on mankind wiping out a species or habitat and I want to puke. My disgust level gets too high so I turn it off. I’ll donate money to curb the insanity, but I don’t want to see another elephant foot wastebasket or set of tusks in some jackass’ office.

Anyway, back to the positives. As you can tell from prior blogs, I get a little wound up over the Space Shuttle. I find the Shuttle and projects like it enthralling, and certainly inspiring. Now I know the Axiom is no space shuttle, but hey, we all have our stuff to work on. Besides, the Axiom is easier to use than the shuttle and a whole lot less expensive. And the Axiom gets way better gas mileage too.

Sure the shuttle is more visible and exciting than a storage array, but any good engineering project excites me. A friend and I were chatting about high frequency semiconductor components the other day. You know, like who doesn’t chat about RF chips? Communication equipment, satellites, cell phones and the like all contain critical high frequency components. Parts such as these use fairly exotic process technologies like Silicon on Sapphire or Gallium-Arsenide to enable their astounding multi GHz capabilities.

As exotic as they may be, some of these parts sell for only thirty cents — or $900 if you sell them to the government. Some people hear the thirty cent price tag they are underwhelmed. In other words, the price tag is not really a good indicator of the sophistication of the underlying technology. I always thought disk drives were that way you know, like $150 for a Terrabyte drive?

I mean, it’s pretty obvious that the space shuttle costs more than it would to buy Greenland, and it probably has more moving parts in the winter. But the engineering and technology that goes into making a 30 cent part or a $150 disk drive with 10 to the 12th power bytes storage capability is no less astounding. It is harder to take something as grand as the shuttle for granted though, as compared to relatively invisible technologies in your cell phone or computer.

Storage arrays are like that, at first they look like a pile of disk drives hooked to a computer, with a few fans and power supplies. Actually, even at last look that’s how they appear because in fact that’s what they are. The magic is in how they are connected, made serviceable, the exact constituents and ratios of each, and all the software that glues them together. Over 5 million lines of code run in the Axiom, and some vendors have more. Much of the value of an Enterprise array is in the software.

Open source aside, software isn’t free. I think at times people underestimate how much engineering goes into designing, releasing and maintaining software like this. Some people start wondering why a disk drive from an array vendor costs more than it does from the local electronics outlet. Well, because the value isn’t the disk drive per se, it’s the software wrapped around it to deliver the features and functions the array offers through a network protocol. It may look easy if we’ve done our jobs right, but it is nowhere near easy.

Science and Engineering of nearly every discipline advances with big projects like the space shuttle, and it always starts out relatively expensive. The same parts that sell for 30 cents in a cell phone used to be thousand dollar subassemblies in a satellite. So we figure it out, and then we miniaturize it and make it cheaper.

I will be sad when the space shuttle takes its final voyage next year, but it marks the beginning of new projects which will make the shuttle pale in comparison. That’s kinda how it goes. In the context of the time they were built in, these engineering projects are truly marvels.

Ya gotta love this stuff. I do.

SHARETweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Pin on PinterestDigg thisShare on RedditShare on TumblrShare on StumbleUponEmail this to someone

Mike Workman is Chairman & CEO of Pillar Data Systems. Mike has spent his career breaking new technical ground in the storage industry. In his 25+ years in the storage business, Mike's appointments have included vice president of worldwide development for IBM's storage technology division, senior vice president and CTO of Conner Peripherals, and vice president of OEM storage subsystems for IBM. He has a PhD and Masters from Stanford, a Bachelors degree from Berkeley and holds over fifteen technology patents.