OCI Data Science now offers E3 Flex Shapes with AMD’s 64-core processor

June 24, 2021 | 2 minute read
Wendy Yip
Data Scientist
Text Size 100%:

If you’re dealing with increasingly compute-intensive workloads, you want more cores and memory — but you also want flexibility. That’s why it’s important to be able to specify exactly the compute shape you need.

To make that even easier, OCI Data Science recently launched support for E3 Flex shapes that let you specify the number of cores and memory for the compute shape. E3 standard instances are based on the AMD EPYC 7742 processor, which has a base clock frequency of 2.25 GHz, a max boost of up to 3.4 GHz, and 64 cores (which AMD calls Tetrahexaconta-core).

While the previous generation shape limited you 1,2,4,8, or 16 cores, now you can choose from 1 to 64 OCPUs. Also, you get up to 1 TB in memory. The E3 instances have 16GB of memory per OCPU, which is double the ratio of the AMD E2 shapes.

Do you need even more flexibility for your workload? With this update, you can deactivate your notebook session, choose a different shape configuration, then reactivate the notebook session.

Let’s take a closer look at the simple process of customizing your shapes.

Choosing your compute shape families

When you log into OCI Data Science, you will see a new UI that allows you to pick different shape families.

When you log into OCI Data Science, you will see a new UI that allows you to pick different shape families.

 

Upon choosing the AMD Rome or the E3 Flex Shape, you can customize the number of OCPUs and memory in your instance.

Upon choosing the AMD Rome or the E3 Flex Shape, you can customize the number of OCPUs and memory in your instance.

 

That’s it — you’ve customized your E3 instance to deliver just the amount of performance you need.

Speed-tested on our benchmarks

Just how much faster are the E3 shapes? We’ve found the second generation AMD EPYC processor server performs impressively compared to existing VM instances on our standardized benchmarks. Compared to the X7 Standard instances, the E3 Standard instance delivers a 60% increase in integer performance per core, a 41% increase in floating point performance per core, and a 52% increase in STREAM Triad bandwidth per core.

Feel free to let us know how your workloads are running on OCI Data Science and the types of problems you’re solving. We may feature you on this blog!

 

Wendy Yip

Data Scientist


Previous Post

Accelerate your model build process with the Intel® Extension for Scikit-learn

Praveen Patil | 3 min read

Next Post


Graph Machine Learning with PyPGX and OML4Py

Rhicheek Patra | 14 min read