As you may have heard, the new iPad was announced Wednesday. While most of the new capabilities were in line with expectations, what may be unexpected is the impact that it, and devices like it, will have on worldwide storage demand. Essentially, cloud storage infrastructure providers just saw their storage forecast for the next year go up by a factor of 10. Addressing these demands in a cost effective manner is going to be critical for these organizations and will be a precursor to storage design for mainstream data centers.
Devices like the new iPad have increased the resolution of the photos and movies they take. The movies it can play are now at a higher resolution. All of this means greater storage demands. Also, you can assume that the higher quality of this content is going to mean that the owners of the devices are going to want to use it more than ever and they will create more of this higher resolution content than they did on previous models.
The problem is the new iPad did not upgrade storage capacities within the different models. We still have the 16-GB, 32-GB and 64-GB capacity levels. That means that people are going to have to get this content off of their device faster than ever to make room for the next round of photos and movies that the owner wants to take or view. The good news for the owner of the device is that the new iPad and its similar competitors have access to the 4G broadband network, and Wi-Fi hotspots seem to be everywhere now.
The problem for cloud storage providers is that this means a higher number of larger files coming at them almost the moment they are created. To summarize: More data in larger files being sent more often at faster speeds. Cloud storage providers are going to need to increase their ability to ingest data and store it long term. This new architecture is something we discussed in our recent article "What is Data Monetization". To solve this problem is going to require all the tools in the storage tool belt.
The ingest problem is going to need to be solved by flash memory. Flash memory for a cloud storage provider seems like an odd coupling, but at the speeds and frequency that this data will be transferred as well as the likely number of simultaneous inbound transfers, it is doubtful that a hard disk-based system will be able to keep up.
The second layer will likely be disk based. Videos and photos will likely have their highest playback rate soon after they were taken. This stored data, however, may not benefit from deduplication and compression. Most of this data is entirely unique and already compressed.
The final layer is were life gets interesting. The default solution today is more disk, but we have to begin to wonder if the cloud storage providers can keep all of this data on disk and afford to offer their services at current price levels. Tape really deserves a fresh look here. It is proving itself more reliable for longer term data retention and it is certainly going to win a cost comparison, especially on data that can't be further optimized through compression or deduplication.
Cloud storage providers are going to need to embrace two technologies that almost seem at odds with each other and that they have not considered thus far--flash and tape. This does not mean they throw out their disk investment--that will be big enough on its own--but they will need to augment with the other two technologies.
Follow Storage Switzerland on Twitter