Amazon Elastic Compute Cloud
User Guide (API Version 2012-04-01)
Print this pageEmail this pageGo to the ForumsView the PDFShare this page on TwitterShare this page on FacebookBookmark this page on DeliciousSubmit this page to RedditSubmit this page to DiggDid this page help you?  Yes  No   Tell us about it...

Instance Types and Architectures FAQ

What happened to the original instance type?

The original instance type is still available. It is called the small instance (m1.small) and it has the same technical specifications.

Will the original instance type be retired soon?

There are no plans to retire the original instance type.

If I do not specify an instance type at launch, what type of instance will I get?

You will get the smallest instance type available for the type of AMI you selected.

Does my instance limit apply to all instance types or is there a separate limit for each type?

The instance limit applies to the sum of all instances, regardless of type. There is no separate instance limit per type.

Can I mix instance types, or do I have to use the same type for all of my instances?

You can launch any combination of instance types. Choose the instance types that have the most appropriate memory, CPU, and storage for each function within your application.

How do I select the right instance type?

Amazon EC2 instances are grouped into several families. For more information, refer to Instance Families and Types.

[Tip]Tip

One of the advantages of Amazon EC2 is that you pay by the instance hour, which makes it convenient and inexpensive to test the performance of your application on different instance families and types. A good way to determine the most appropriate instance family and instance type is to launch test instances and benchmark your application.

When should I use High-CPU instance types (c1.medium and c1.xlarge)?

High-CPU instance types have a proportionately higher ratio of CPU to memory and are well suited for compute-intensive applications. To determine whether they are appropriate for you, launch an instance and benchmark your own application on different instance types and calculate which is most appropriate.

Which instance types are 32-bit and which are 64-bit?

All instance types can run 64-bit AMIs. The micro (t1.micro), small (m1.small), medium (m1.medium), and High-CPU medium (c1.medium) instances can also run 32-bit AMIs.

Can I launch any AMI on any type of instance?

No. Cluster compute (cc1.4xlarge and cc2.8xlarge) and cluster graphics (cg1.4xlage) instances can only run HVM AMIs. All other instances can run Windows HVM or Linux/UNIX paravirtualized AMIs.

Can I use my own kernel?

Yes. For more information, see Enabling Your Own Linux Kernels.

Do I have to do anything special to bundle the large or extra large Amazon EC2 instance store-backed instances?

Make sure to use the latest AMI Tools.

Can I build an AMI that works on both 32-bit and 64-bit instances?

No, an AMI is either a 32-bit AMI or a 64-bit.

Can I run 32- bit applications on 64-bit AMIs?

You can run a 32-bit application on a 64-bit host if the Linux/UNIX kernel is compiled with IA32 emulation and the correct 32-bit libraries are available.

By default, the Amazon DomU Kernel has IA32 emulation enabled and there are many public AMIs that include preinstalled 32-bit libraries. If the library you require is not included with the AMI, you can install it using standard tools (e.g., yum).

How does I/O performance of an instance store compare to the I/O performance of an EBS volume?

The instance stores on large and extra large instances have higher and more consistent I/O performance than the original (small) instance. Amazon EBS volumes perform consistently for all instance types.

[Note]Note

The first write to any given block of the disk will be slower than subsequent writes. For more information, see Disk Performance Optimization

Can I RAID the spindles exposed on large and extra large instances?

Yes, you can use software RAID on top of the exposed spindles.

[Note]Note

The initial RAID setup might take a long time. For more information, see Disk Performance Optimization