"AWS Snowmobile" to efficiently transfer exabyte-class data to AWS

In Amazon, as a data transfer service from the data center to AWS "AWS Import / Export"And announced" AWS Snowball "as its option in 2015. In 2016, as an option further upgraded"AWS Snowball Edge"Evolved in a different direction"AWS Snowmobile"Was announced.

AWS Snowball Edge - more storage, local endpoint, Lambda function | Amazon Web Services blog
https://aws.amazon.com/jp/blogs/news/aws-snowball-edge-more-storage-local-endpoints-lambda-functions/

AWS Snowmobile - move exabyte data to the cloud in a few weeks | Amazon Web Services Blog
https://aws.amazon.com/jp/blogs/news/aws-snowmobile-move-exabytes-of-data-to-the-cloud-in-weeks/

"AWS Snowball" is a hardware that stores 50 TB storage in a waterproof / shock-resistant case and can connect with 10 Gb Ethernet.


Amazon Web Services event "AWS re: Invent 2016", Capacity increased to 100 TB"AWS Snowball Edge"Was announced.


However, if the data capacity has reached the unit of exabyte (above petabyte) in total, it is necessary to collect 100 thousand storages for 10 thousand units. Therefore, what was thought to move data on the exabyte scale was "AWS Snowmobile"is. It carries 100 PB storage in a 45 ft container, moves directly to the data center of the data transfer source with a semi-trailer track, takes in data and transfers it to AWS.


If you use 10 Gb Ethernet like AWS Snowball, it will take 26 years to transfer if you have exabyte class data, but if you use 10 AWS Snowmobile you can shorten it to about six months is. Price is not announced but it seems to be "faster and cheaper than using a network based data transfer model".