ÜBERTECH | A ZDNet Multiplexer Blog What's this?

The Biggest, Ever

This month SAP partnered with BMMSoft, HP, Intel, NetApp, Red Hat and MD&Profy to build the biggest data warehouse ever.

I often betray my age when referring to storage / file size—getting bemused looks when I, say, can’t believe that a file is so huge (2kB!), or ask for a flash drive with 16 MB of storage (when of course what I mean to say is 2 MB and 16 GB). When it comes to storage size, my mind is often stuck twenty years ago. In fact, last year at the Edinburgh Fringe, I entered a near Rain Man-state of audience participation when, during Dan Willis’ “PC, Mac And Me – The Funny Side Of Computers”, he asked the audience for the memory size of a variety of computers, I shouted out all the relevant numbers from the back row. I couldn’t resist, as Willis asked about my first computer—the Sinclair ZX-81. (The answer is of course a massive 1 kB, or if you were lucky enough to have the very wobbly RAM pack, then an enormous 16 kB).

Today, it’s hard to believe I was so envious of the BBC Model B. With so much memory—32kB—you could model an entire universe in Elite!

But tell kids today, and they won’t believe you.

Memory just keeps inflating. On the BBC you could run Elite in 32kB, whilst a blank word doc I’ve just created is 20.8kB.

Or how about this: the Nintendo NES —the combined RAM of all the 61.9 million NES units sold equates to about 118 GB (See 0:25 in this video, “5 Awesome NEW Facts!”) —or the equivalent of 118 iPhone 5s! If you work at a largish office, between everyone’s mobile devices you’ll have more RAM than the all the NESs ever produced.

Having more storage and more devices creates more data that needs to be stored and, for many industries, analyzed.

This month SAP, with the help of our partners BMMSoft, HP, Intel, NetApp, Red Hat, and MD&Profy, set the Guinness World’s record for largest data warehouse: 12.1 petabytes (PB).

View the press release here.

How big is that? Well, 12+ PBs could store the entire printed collection of the US Library of Congress (10 Terabytes) 12,000+ times over. Or, it could store the entire printed content of all academic research libraries (2 Petabytes) 6 times over. Or to go back to the NES scale, then as the largest NES game was 1 megabyte that would be 13 billion (different) games!

From a business big data and analytical perspective, this means you have the capacity and performance to never archive critical business data. Imagine storing and analyzing 10 years of social media data in real-time. The potential impacts on business and decision-making are startling. The end result is enabling a whole new set of applications and capabilities as a result of this unprecedented scale of analysis.