instance type (f1.2xlarge) is not supported in your requested Availability Zone

The issue you're experiencing with F1 instances not being supported in your requested Availability Zones, despite the describe-instance-type-offerings command showing availability, could be due to several factors:

1. Account-specific limitations: Even if F1 instances are generally available in a region, there might be account-level restrictions preventing you from launching them. This is particularly common for specialized instance types like F1.

2. Capacity constraints: AWS may not have sufficient capacity for F1 instances in the specific Availability Zones at the time of your request, even though they are generally supported there.

3. AMI compatibility: Ensure you're using the most up-to-date FPGA Developer AMI that's compatible with F1 instances in your chosen region.

4. F1 instance approval: If you've never used F1 instances on this particular AWS account before, you may need additional approval to use them, even if you've used them on other accounts.

5. F2 recommendation: For new F instance customers, AWS recommends using F2 instances instead of F1. If you haven't used F1 instances before, consider trying F2 instances.

To resolve this issue, you could try the following:

1. Contact AWS Support to check for any account-specific limitations and request approval for F1 instances if needed.

2. Ask AWS Support to increase your F instance on-demand quota and specifically enable F1 instances for your account.

3. Try launching the instance at a different time, as capacity issues may be temporary.

4. If you're new to F instances, consider using F2 instances instead, as recommended by AWS.

5. Verify that you're using the latest version of the FPGA Developer AMI.

If the issue persists after trying these steps, it's best to reach out to AWS Support for more detailed assistance, as they can provide specific information about your account's status and any regional availability issues for F1 instances.
Sources
Fail to create EC2 F1 instance | AWS re:Post
Amazon EMR cluster error: Instance type not supported - Amazon EMR

Hi Mark,

Are you currently an F1 customer? IE do you have active F1 instances? If not, we recommend that you check out F2, which is the successor to F1!

Here's some additional information about F2, as well as our new developer AMI to go with it.

The F1 platform is going end-of-life on December 31st, 2025 and we are actively guiding customers in their transition to the F2 platform.

Hope this helps!

Best, Max

All of the industry on-boarding materials are still dedicated to F1:

Chipyard

• https://chipyard.readthedocs.io/en/latest/Simulation/FPGA-Accelerated-Simulation.html#example-firesim-workflow

FireSim

• https://docs.fires.im/en/latest/Getting-Started-Guides/AWS-EC2-F1-Getting-Started/Running-Simulations/Running-a-Single-Node-Simulation.html
• https://docs.fires.im/en/latest/Getting-Started-Guides/AWS-EC2-F1-Getting-Started/Running-Simulations/Running-a-Cluster-Simulation.html

Before you start forcing new users to move onto the bleeding edge: I don't believe it is unreasonable to ask that you address why the industry wide introductory materials are not moving on to F2. Especially when the AWS materials are so appalling in comparison.

Also... when I subscribed to the FPGA developer image the only type it put forward is the F1, hence I've written all my scripts and gathered all my tutorials on that basis... not happy!: