Media Storage Report

It’s that time of the year: Tom Coughlin has published his annual report on digital storage for the media and entertainment industries. We thankfully read StorageNewsletter‘s summary and liked what Coughlin had to say. If he is correct, the media and entertainment segment for the storage industry will grow from a $3.8 billion business to $6.4 billion, and from 11EB to 62EB in capacity. Here is our attempt to add some comments to the comments. Two returning elements in (the summary of) the paper are efficiency and archival: there seems to a growing awareness of the need to store data more efficiently and most of the capacity growth is for data archives.

“Digital storage requirements are exploding due to use of higher resolution and stereoscopic content in the media and entertainment industry. 4k (4000 horizontal pixels) seems to be the next thing and 8k is the thing after”. Well this already explains a lot of the growth: assuming that compression ratios remain similar, 4k movies will require four times as much storage as today’s movies. The open question is of course how fast 4k will become a standard. 3D Cinema will continue to boost storage needs as well. Avatar, for example, required hundreds of petabytes (according to our sources) for production and post-production. The final version of the movie was about a petabyte. The figures are confirmed in Coughlin’s report report: “Several petabytes of storage may be required for a complete stereoscopic digital movie production at 4K resolution.”

“The single biggest application (by storage capacity) for digital storage in the next several years as well as one of the most challenging is the digital conversion of film, video tape and other analog formats.” This explains another big chunk of the digital storage growth. Everything that can be digital will be digital (I believe someone else said that in the past). For many of those digitization projects, companies have no idea how large the storage infrastructure will have to be eventually so scalability is very important. To further illustrate the importance of archival, the report states “Over 61 Exabytes of digital storage will be used for digital archiving and content conversion and preservation by 2015.”

“Active archiving will drive increased use of HDD storage for archiving applications supplementing tape for long term archives.” This is really a new market and will have a visible effect on HDD sales; while it will only moderately affect the tape industry. What will happen here is that media producers will postpone the moment when they offload an asset to tape. Rather than doing that immediately after the production is finished, a movie will be kept on tape for a certain amount of time. This saves them the hassle of offloading movies from tape when needed for a new release of the same movie. An important question is how long an asset will be kept on the Active Archive.

“Between 2011 and 2016 we expect the media and entertainment industry will see about a 7.7X increase in the required digital storage capacity and about a 5.6X growth in storage capacity shipments per year (from 11,248 PB to 62,736 PB).” I should probably read the report to be sure, but there seems to be a serious gap between required storage growth and the increase in storage capacity. This calls for a more efficient approach of storage management.

“Total revenue for media and entertainment storage systems will increase about 1.7X from 2011 through 2016 ($3.8 B to $6.4 B); for storage media and devices about 1.6X” This either points to a decrease in margin for storage providers or improved efficiency. This is not different than the trends we’ve seen. As margins cannot decrease eternally, efficiency will be key.

“About 57% of the total storage capacity will be used for content archiving and preservation in 2011. We believe that this will increase to 60% of total capacity by 2016 due to more efficient and cost effective conversion services, lower overall storage costs and a greater ROI on long tail content.” It makes a lot of sense that most of the storage efficiency improvements happen in the archive layer. This is the fastest growing storage tier, this is also the tier that is least affected by performance changes.