Introducing ePAT: The Solution To Your Storage Problems
The new technology that's sure to revolutionize the industry.
I recently returned from the Fujifilm Global IT Executive Summit in Hollywood, followed by a brief sojourn to Boulder, Colo. for Spectra Logic's analyst day. Following those outings, I have to say that I am impressed by the strides that a newly introduced storage technology has made in just a short time since its first appearance. I am talking, of course, about ePAT, and I'm convinced that it will be a key part of your storage infrastructure in the near future, if it isn't already.
Some folks may be wondering if storage really needs a new media type. With all the up-and-coming flash storage products appearing in the market -- soon to replace disk altogether with the advent of 3D chips, their excited vendors say -- why ePAT?
Well, truth be told, Flash RAM does still have its wear problems. Despite best efforts to mask multilayer cell wear by using on-controller algorithms to do brute force swaps of unlisted "virgin" cell groups for those with heavier usage and wear patterns, the problem of wear persists. Rates are hard to predict with accuracy. Your workload will ultimately determine whether your flash storage delivers three to five years of use … or three to five months, weeks or days. Plus, flash memories really aren't optimized for writing data. For that, you still need to use more volatile DRAM, which kind of defeats the low cost argument that the Flashies are trying to make.
Alternatively, if you are a disk dweeb, you may be confused by the need for ePAT, given the huge advances in your preferred technology. Seems like only a few years ago, folks were worrying about the Superparamagnetic Effect and its effect on parallel magnetic recording. There was a lot of crepe hanging around disk technology's roadmap. Then came Perpendicular Magnetic Recording (PMR) and a whole new tick-tock in drive capacity improvements.
Now that PMR has just about reached its limits, about the only way to add capacity to a disk seems to be by adding more platters in a helium-filled case. Once again, disk doubters are predicting end times for HDD technology.
There is shingled magnetic recording (SMR), of course. That's a way to write data to disk that gets you more space on the same size platter. But when you look into it, SMR disk is really like tape, but with random access capability. I could do a whole column on this, but suffice it to say that Spectra Logic is already doing some interesting work with SMR in their just-announced ArcticBlue kit. Worth a look.
And, of course, work is proceeding at Seagate and elsewhere on heat assisted magnetic recording (HAMR), and maybe bit patterned media (BPM), to further boost the storage capacity of spinning rust. Until those technologies are realized, however, the roadmap for increasing disk storage capacity is in flux. By contrast, ePAT is available now.
Rife With Strife
You might be thinking about optical disc, especially after the shot in the arm the technology got recently from Facebook. But that technology is still mired down in intramural strife. In the last decade, the vendors of optical storage seemed to exhaust themselves in a fratricidal conflict over HD DVD and BluRay.
By the time that the issue was decided, it seemed like even consumers had moved on, by streaming their audio and video media from online sources instead of buying and/or recording it to optical. And, of course, enterprise storage seems to have passed the optical storage medium by. Even the industry's latest promises that a 1 TB optical disc will appear over the next five to 10 years has been met with a collective yawn.
Collectively speaking, the reality is that none of the above storage media, separately or collectively, can possibly handle the tsunami of data that analysts are predicting will crash and smash corporate storage levees by 2020. The size of the data deluge, with estimates ranging from 15-20 Zettabytes in one case to upwards of 60 Zettabytes in another, is simply mind numbing. Even with optimistic levels of media manufacture and new technology improvements, demand for space to store all the bits will simply outstrip the capacity that's deployed or available for purchase.
No Silver Lining in the Cloud
Even the cloudies won't be able to handle such a deluge. Though they are often (and incorrectly) treated like another storage medium, cloud storage peddlers are service providers using the same flash, disk, and optical media as everyone else. They are confronting the same challenges and fielding the same technologies as the rest of us. So, even with all of the combined capacity of all storage technologies, including clouds, our capacity for storing data will not be able to keep pace with all the bits. Period. End of story.
Without ePAT, our options will be simple. On the one hand, you could do what the labs conducting particle collider experiments do and simply discard 90% to 95% of the new data they generate with each experiment. Is discarding a lot of data for lack of storage capacity really the best option? The alternative is simple: look for an entirely new storage media.
Suddenly, that ePAT stuff seems like just what we need, and just in time for 2016. From where I'm standing, ePAT is truly a technology whose time has come.
ePAT to the Rescue
With ePAT storage, we will be able to store Zettabytes of data at a pretty low cost for up to 30. The latest tests have demonstrated that a single ePAT storage unit will hold up to 220 TB of data on a single instance of media about the size of pack of ramen noodles, or small box of Godiva chocolates; though it may be a few years before you start seeing ePATs with that capacity. Right now, you can readily get ePAT with 15TB of capacity per unit if you compress the data on the ePAT drive.
Moreover, ePAT sports the fastest write speed of any media; so fast in fact that you may not be able to port data to it fast enough to fully realize its speed. On playback, ePAT offers the fastest and most steady and jitter-free read/stream rate of any media on the planet. Durability is 30 years or more, and energy consumption is extraordinarily low. And placed behind your existing storage, especially NAS, ePAT can make your disk storage appear to be limitless.
Low Cost, High Capacity
The really good news is that you can get into an ePAT storage solution with a Petabyte of capacity right now for under $40K. While it may not be the best option for randomly accessed, frequently updated data, ePAT is perfect for either tier 0/1 high speed data capture, or tier 2/3 long term data retention/preservation storage for that 70 percent - 80 percent of your data that is neither re-referenced nor changed very often.
You can buy ePAT from trusted vendors like IBM, Oracle, Spectra Logic and a few others, and the media is available en masse from one of the most trusted names in the business, FUJIFILM. Some ePAT comes with special ingredients like aromatic polyamides (aka Aramid) or polyethylene terhthalate (aka PET), but all of the products make use of Barium Ferrite (BaFe) as a recording layer. It's a unique material providing extremely fine and very reliable data recording.
You've Known It All Along
For those who continue to be confused, ePAT is "tape" spelled backwards. I got the idea a few years back from the anti-tape deduplicating disk array maker, SEPATON ("NO TAPES" spelled backwards). My objective was to help you to consider the merits of tape technology by framing it as a shiny new thing. Too often, we simply switch off any consideration of the technology upon hearing its name, either because of old data, incorrect perceptions, or prior bad experiences. In this case, you owe it to yourself to learn more about BaFe on ePAT; that is, tape storage. A lot has changed and improved in the technology; and frankly, it is the only choice for storing Zettabytes of data in the future.
Be sure to tell them that oGIOT sent you.
Jon Toigo is a 30-year veteran of IT, and the Managing Partner of Toigo Partners International, an IT industry watchdog and consumer advocacy. He is also the chairman of the Data Management Institute, which focuses on the development of data management as a professional discipline. Toigo has written 15 books on business and IT and published more than 3,000 articles in the technology trade press. He is currently working on several book projects, including The Infrastruggle (for which this blog is named) which he is developing as a blook.