Today, a large number of companies are building their own proprietary data and computing infrastructures for managing the varying workloads for AI, Machine Learning, Analytics, IoT, and 5G connectivity. These new trends have ensured that the demand for high-performance computing (HPC) infrastructure reaches its highest point in 2023. Traditionally, HPC systems were powered using local parallel computing machines connected to a data center. However, there has been a major shift in the HPC deployment trends in the remote working era. Today, cloud-based HPC systems have taken over the market due to their countless benefits to remote working teams. Seamless integrations, unlimited data storage, agility, faster deliveries, cost-effectiveness, and democratized access to enterprise data and analytics are some of the top reasons why customers prefer to invest in cloud HPC infrastructure for their varying computing needs. AWS leads the pack in this cloud HPC service and solutions market.
To meet the growing HPC demands of its customers, Amazon Web Services (AWS) has launched two new Amazon EC2 instances to manage the varying virtual workloads across multiple environments.
The two new instances of Amazon Elastic Computer Cloud or Amazon EC2 are:
- M7i-Flex
- M7i
Both instances are powered by the fourth-generation Intel Xeon Scalable Processors. These are custom-built to cater to the needs of EC2 workloads in the HPC systems. Compared to other Intel processors, this integration fires up the cloud performance in terms of computing speed by up to 15%.
Here’s a quick overview of what the new Amazon EC2 instances, M7i-Flex and M7i, offer to users.
| M7i-flex Instances (Category-first) | M7i Instances (Category-first) |
| Lower cost variant of M7i instances at 5% lower pricing | Offers price-performance benefits for workloads that need the largest instance sizes or continuous high CPU use |
| Ideal for applications that do not require full-capacity computing resources | 15% better price performance compared to M6i instances and DDR5 memory |
| Baseline CPU usage is 40%; maximum CPU performance for 95% of the time | Larger instance sizes (up to 48xlarge) / Bare metal sizes (metal-24xl, metal-48xl) |
| No change to your existing workloads or app infrastructure to deploy M7i-flex instances | Ideal for workloads that need continuous high CPU performances for larger applications |
| Available in five sizes | Available in nine sizes |
| Deliver 19% better price/performance than the previous generation of Intel processors | Best for CPU-based machine learning applications (up to 400 Gbps), gaming servers, video streaming apps |
| Ideal for use in web/ app servers, batch processing, enterprise resource and digital asset management, microservices, and databases | Provision to attach up to 128 EBS volumes of each M7i instances |
The newly launched instances of AWS EC2 use the 4th Generation Intel Xeon Scalable processors code name Sapphire Rapids. In addition to AWS, other users include Cisco, Cloudera, Dell Technologies, Google Cloud, HPE, Lenovo, Dropbox, Microsoft Azure, NVIDIA, Oracle Cloud, SAP, VMware, and others. This purpose-built workload-first approach to varying workloads in AI, data analytics, data storage, HPC, and security lowers the total cost of ownership (TCO) by up to 60%. Integrating with AWS EC2, users can further optimize their costs and resource allocation by delivering high-end computing capabilities within hybrid and Edge Cloud environments.
AWS customers can leverage four built-in accelerators powered by Sapphire Rapids processors.
These are:
- Advanced Matrix Extensions (AMX) to support workloads such as NLP, deep learning recommendation systems, and image recognition
- Intel Data Streaming Accelerator (DSA) to support data movements during storage, networking, and data-intensive transformations
- Intel In-Memory Analytics Accelerator (IAA) to support in-memory compression, decompression, and encryption
- Intel QuickAssist Technology (QAT) to offload encryption, decryption, and compression at reduced power consumption
As per an official statement published here, the Intel QAT, Intel IAA, and Intel DSA accelerators will be available on the m7i.metal-24xl and m7i.metal-48xl instances.
Top AWS Announcements: AWS Launches Infrastructure Region in Israel
Currently, the two instances would be offered to customers based out of specific AWS regions in the US and Europe with different pricing options for each instances. Customers such as Intuit, Autodesk, Snap, Cathay Pacific, and VW Group use AWS machine learning services.
Top EC2 Features include:
- Processors powered by processors from Intel, NVIDIA, and AWS
- Pre-configured services and orchestrations to integrate with an ever-growing list of operating systems such as Windows, CentOS, Linux, Ubuntu, SUSE, and Debian
- Pay for what you use pricing options
- Auto-scaling using predictive machine intelligence
- Optimized CPU configurations
- Pause and resume instances facilities
- Optimal storage capabilities across the Amazon EBS and EFS
- Dynamic cloud computing with Elastic IP Address
- Access to fo HPC clusters for high output and low latency
- Real-time maintenance updates during running, migration, and security patches
