AWS Snowball Edge: What It Is & How It Works

AWS Snowball Edge is a rugged, portable storage and compute device that AWS ships to you so you can transfer large amounts of data to the cloud without using your internet connection, or run compute workloads in places with poor connectivity. You fill it with data or run applications on it, then send it back, and AWS imports the data into Amazon S3. It is part of the AWS Snow Family, which also includes the smaller Snowcone and the larger Snowmobile.

Why AWS Built the Snowball Edge

Moving terabytes or petabytes over the internet is slow, expensive, and often impractical. A 100 Gbps connection sounds fast, but real-world throughput is far lower, and many environments—remote research stations, oil rigs, ships, or disaster zones—have no reliable broadband at all. AWS created the Snow family to solve this “data gravity” problem: instead of streaming data to the cloud, you ship the storage to the data.

The Snowball Edge is the middle option. It offers more capacity and compute power than the Snowcone (a briefcase-sized device) but is far more portable than the Snowmobile (a shipping container pulled by a truck). For many organizations, the Edge is the sweet spot between capacity and practicality.

Two Flavors: Storage Optimized vs. Compute Optimized

The Snowball Edge comes in two main configurations, and choosing the right one is the first decision you’ll make.

• Storage Optimized – This model offers 80 TB of usable HDD storage (up to 210 TB in the newer generation) and 40 vCPUs with 80 GB of memory. It is designed for bulk data transfer and for running lightweight compute jobs like data validation or compression while the device is on-site.
• Compute Optimized – This model provides 42 TB of usable SSD storage (up to 28 TB in the older generation) but packs 52 vCPUs and 208 GB of memory. It also includes an optional NVIDIA Tesla T4 GPU, making it suitable for machine learning inference, video processing, or running EC2 instances in the field.

Both models run Amazon EC2 instances and AWS Lambda functions locally, so you can process data before shipping it back. This is useful for pre-processing, filtering, or compressing data to reduce the amount you need to transfer.

How the Process Works

Using a Snowball Edge is a straightforward, but multi-step, procedure.

1. Create a job in the AWS Management Console. You specify whether you want to import data to S3 or export data from S3, choose the device type, and provide your shipping address.
2. Receive the device. AWS ships the Snowball Edge in a rugged case with a built-in E-ink display. The display shows a shipping label and a QR code.
3. Connect and unlock. You connect the device to your local network, power it on, and use the AWS OpsHub app (or the older CLI) to unlock it with a credentials file that AWS provides.
4. Transfer or compute. You copy data to the device using S3-compatible endpoints or NFS, or you launch EC2 instances and run your applications directly on the device.
5. Ship it back. When done, you zero out the device (if you want), disconnect it, and ship it back using the prepaid label. AWS erases the device according to NIST 800-88 standards.
6. Data is imported. AWS copies your data into the S3 bucket you specified, and you can verify it in the console.

Costs and Pricing Model

You don’t buy a Snowball Edge—you rent it. AWS charges a per-day fee for each device, plus a one-time job fee. The per-day fee starts when the device ships and ends when it is received back at AWS. There is no charge for data transfer into S3, but you pay standard S3 storage rates once the data is in the cloud.

For example, a 10 PB migration would require roughly 125 Snowball Edge devices (80 TB each). At the current per-day rate, this could cost tens of thousands of dollars, but it is often far cheaper than upgrading your internet connection or paying for dedicated network circuits over many months. AWS also offers a data transfer acceleration option that uses a dedicated network path for the final leg, but the device itself is the primary transfer method.

Pricing is region-specific, so check the official AWS Snowball Edge pricing page for up-to-date numbers. Also note that you pay for the device even while it sits in your office waiting for the return shipment, so plan your logistics carefully.

Common Use Cases

The Snowball Edge is used in a wide range of industries and scenarios.

• Data center migration – Moving hundreds of terabytes of backups, archives, or legacy data to S3.
• Remote site data collection – Gathering data from oil rigs, ships, mines, or research stations where internet is absent or unreliable.
• Edge compute – Running real-time analytics, video processing, or ML inference on-site, then shipping the results to AWS.
• Disaster recovery – Sending critical data to AWS as an off-site backup when you can’t use a VPN.
• Content distribution – Shipping large media files (movies, games, datasets) to AWS for processing or distribution.

Security Features

Security is a top concern when shipping physical hardware. AWS builds multiple layers of protection into the Snowball Edge.

• Tamper-evident enclosure – The device has a tamper-resistant case and a tamper-evident seal. If anyone opens it, the seal is visibly broken.
• Encryption – All data is encrypted with AES-256 encryption keys that you manage in the AWS Key Management Service (KMS). The keys are never stored on the device.
• Secure erasure – After the data is imported, AWS performs a cryptographic erasure of the device, meeting NIST 800-88 standards.
• Trusted Platform Module (TPM) – The device uses a TPM to ensure that only authorized AWS software runs on it.

You can also use the Snowball Edge Client to verify the integrity of your data before shipping. The client calculates checksums, so you can confirm that the data on the device matches your source.

Limitations and Things to Watch Out For

While the Snowball Edge is powerful, it has some limitations that can trip up first-time users.

• Transfer speed – The device uses 10 GbE or 25 GbE network interfaces. To get the full speed, you need a compatible network card and switch. Many users see real-world speeds of 300-500 MB/s, not the theoretical maximum.
• File system support – The device presents an S3-compatible endpoint, so it works with S3 tools like the AWS CLI, but not with NFS unless you use the file interface option (which is available on some models).
• Storage capacity – The advertised 80 TB is usable capacity, but you lose some space to metadata and the compute instances. Plan for about 10% overhead.
• Shipping time – The entire process can take 1-2 weeks, depending on your location and the shipping carrier. If you need data in the cloud quickly, this may not be the right tool.

Comparing to Other AWS Transfer Options

The Snowball Edge is not the only way to get data into AWS. Here’s how it stacks up against the alternatives.

• Snowcone – Smaller (8 TB), lighter, and cheaper. Good for smaller transfers or edge compute in tight spaces.
• Snowmobile – A 100 PB shipping container. Best for extremely large migrations (over 10 PB) where you can’t ship dozens of individual devices.
• AWS DataSync – A software agent that transfers data over the internet or AWS Direct Connect. It’s faster for smaller datasets (under a few TB) and gives you real-time replication.
• Direct Connect – A dedicated network connection from your data center to AWS. Ideal for ongoing, high-volume data transfer, but requires a physical connection and can take months to set up.

In general, use Snowball Edge for one-time, large (10+ TB) transfers, or for edge compute in disconnected environments. Use DataSync for ongoing transfers or for smaller datasets.

Frequently Asked Questions

How long does it take to ship a Snowball Edge?

Shipping time depends on your location. In the US, it typically takes 2-3 days each way. International shipments can take a week or more. AWS provides a tracking number, and you can see the expected delivery date in the console.

Can I use the Snowball Edge without an internet connection?

Yes, that’s the whole point. You can transfer data to the device and run compute workloads entirely offline. You only need internet to create the job and to download the credentials file before the device arrives.

What happens if the device is lost or damaged in shipping?

AWS insures the device and the data. If the device is lost or damaged, AWS will replace it. However, you are responsible for the data until it is imported into S3, so it’s a good practice to keep a copy of critical data elsewhere.

Can I use the Snowball Edge to export data from S3?

Yes. You can create an export job, and AWS will load the device with the data from your S3 bucket. You then ship it to your location and copy the data off. This is useful for disaster recovery or for moving data to a different cloud provider.

Is the Snowball Edge compatible with the AWS CLI?

Yes. The device exposes an S3-compatible endpoint, so you can use the AWS CLI with the –endpoint-url flag. You can also use the AWS OpsHub GUI, which is simpler for most users.

Related

• Arc Raiders Extraction & Quest Progression Guide
• Space Editor: Role, Workflow & How Space Journalism Works
• GST 2.0: Unlocking Working Capital for Indian Builders?