AWS cluster error message not enough vcpu despite within quota

I'm trying to setup a cluster through AWS. My current standard spot instance (A C D H I M R T Z) request quota is 256. However, when I start the cluster on cloud center, I keep recieving

"Instance Launch Error You have requested more vCPU capacity than your current vCPU limit of 32 allows for the instance bucket that the specified instance type belongs to. Please visit http://aws.amazon.com/contact-us/ec2-request to request an adjustment to this limit. Launching EC2 instance failed. "

These are the spcifications I used

MATLAB Version: R2024a

Shared State: Personal

Auto-Manage Cluster Access: Yes

Worker Machine Type: General Purpose (m5.16xlarge, 32 core)

Workers per Machine: 32

Use a dedicated headnode: Yes

Headnode Machine Type: Standard (c5d.4xlarge, 8 core, 1x400 NVMe)

Machines Requested: 5

Machines Available: 0

Allow cluster to auto-resize: No

Access to MATLAB Drive: Yes

Initial Worker Count: 128

Available Worker Count: 0

Persisted Storage: 100GB

Cluster Log Level: Low

Answer 1

[Jaswanth]

Hi,

In response to the issue you have encountered, specifically 'Instance launch Error' when trying to setup a cluster through AWS, I would like to direct your attention to a potentially helpful solution answered in MATLAB Answers Community:

https://in.mathworks.com/matlabcentral/answers/2112911-matlab-cluster-on-aws-ec2-vcpu-utilization#answer_1451516

As per the reference provided above, Cloud Center limits the number of workers to one per physical core for optimal performance, despite Amazon Web Services machines having multiple virtual cores.

Since each physical core is associated with two virtual cores sharing a Floating-Point Unit (FPU), and most MATLAB computations are double-precision floating point which rely on the FPU, this restriction ensures that each worker has exclusive access to an FPU, thereby optimizing computational efficiency.

For a more detailed explanation on EC2 Instance Machine Types, please refer to following MathWorks documentation link:

https://www.mathworks.com/help/cloudcenter/ug/choose-supported-ec2-instance-machine-types.html#:~:text=For%20example%2C%20a%20m5.8xlarge%20has%2016%20physical%20CPU%20cores%2C%20which%20corresponds%20to%2032%20vCPUs

I hope the information provided above is helpful.