AWS Lambda vCPU/memory ratio table
1
When Lambda maximum memory size was 3GB, we had official information that below 1.8GB it has 1 vCPU and above 1.8GB it has 2 vCPUs. So we were able to share some multi-threading optimization tips with the customers.
Now with support of up to 10GB of RAM and up to 6 vCPUs, do we have any official information on the number of vCPUs based on the memory size?
asked 3 years ago8607 views
1 Answer
- Newest
- Most votes
- Most comments
0
The details can be found here.
Also more details was shared in the What's New in Serverless re:Invent session.
Relevant content
- asked 2 years ago
- asked 6 months ago
- asked 2 years ago
AWS OFFICIALUpdated a year ago- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated a year ago
the link to re:Invent session is broken.
Mentioned above details can be found here: https://youtu.be/aW5EtKHTMuQ?t=339
What's interesting is that for the
nodejs16.xinternally process detects different CPU configurations, we get:128-3008 MB - 2 Allocated cores 3009-5307 - 3 Allocated Cores
And rest of the numbers are as in the presentation
I wonder whether it's a quirk of Node.js CPU cores detection, or AWS does something different for Node.js specifically