AWS recently announced the general availability (GA) of Amazon EC2 M7i and M7i-flex instances, equipped with custom 4th Gen Intel Xeon Scalable processors (code name Sapphire Rapids). The Amazon EC2 M7i and M7i-flex instances are instance types intended for general-purpose workloads providing a balance of compute, memory, and networking resources.
Amazon EC2 M7i and M7i-flex instances are designed for workloads that use resources equally, such as web servers and code repositories. Compared to the EC2 M6i instances announced two years ago, the company claims a 19% better price/performance for the M7i-flex instances and 15% better price performance for the M7i instances.
The M7i-flex instances are a lower-cost variant of the M7i instances. They are available in the five most common sizes with dimensions ranging from 2 vCPU and 8 GiB memory (m7i-flex.large) to 32 vCPU and 128 GiB memory (m7i-flex.8xlarge) – each having up to 12.5 Gbps network bandwidth and up to Gbps EBS bandwidth. Ideal, according to the company, for running general-purpose workloads such as web and application servers, virtual desktops, batch processing, micro-services, databases, and enterprise applications.
On a Reddit thread, the M7i-flex raises questions about being lower-cost, as a respondent seligman99 asks:
Given that it's on the 4th gen Xeon, I wonder if it has something to do with how they’re selling the High/Low Priority Cores to users?
With other Mutjny responding:
That’s actually a great point. These only go up to 8xlarge, and there is a chart that lists them as "Base Performance / vCore: 40%," so I think that’s very strongly what it is. Except these m7i-flex are Intel Xeon Platinum 8488C, which shows is a Sapphire Rapids chipset, which would seem to contra-indicate these are low-performance cores in a mixed core CPU.
On the other hand, M7i instances are available in nine sizes ranging from 2 vCPU and 8 GiB memory (m7i.large) to 192 vCPU and 768 GiB memory (m7i.48xlarge) – with increasing network bandwidth up to 50 GB and up to 40 Gb ESB bandwidth. According to the company, these instances are recommended for workloads such as large application servers and databases, gaming servers, CPU-based machine learning, and video streaming.
Also, with built-in accelerators like Intel Advanced Matrix Extensions (Intel AMX) in the Sapphire Rapids processors, Intel states in a press release:
Built-in accelerators like Intel Advanced Matrix Extensions (Intel AMX) offer a much-needed alternative in the market for customers with growing AI workload demand. 4th Gen Xeon with AMX can also meet inference performance metrics for large language models (LLMs) below 20 billion parameters, making LLMs both cost-effective and sustainable to run on general-purpose infrastructure.
The other built-in accelerators in the Sapphire Rapids processors are:
- Intel's Data Streaming Accelerator (DSA) enhances performance for storage, networking, and data-intensive tasks by efficiently handling data movement between CPU, memory, caches, network devices, and storage devices.
- In-Memory Analytics Accelerator (IAA) boosts database and analytic workloads' speed and potential power efficiency through high-throughput in-memory compression, decompression, and encryption.
- QuickAssist Technology (QAT) relieves processor cores by offloading encryption, decryption, and compression tasks, reducing power consumption while facilitating merged compression and encryption within a single data flow.
Furthermore, future additions to the M7i family will include bare-metal sizes suited to high-transaction and latency-sensitive workloads.
Alongside AWS, Azure and Google Cloud offer a wide selection of instance types and varying combinations of storage, CPU, memory, and networking capacity, allowing organizations to scale their resources to match the demands of their specific workloads. For instance, Microsoft offers various Virtual Machines for general-purpose workloads; the latest is the Dv5-series, equipped with the third-generation Intel Xeon Platinum processor. In comparison, Google Cloud has E2, N2, N2D, and N1 general-purpose machines and C3 instances released last October, including Sapphire Rapids processors.
Currently, the EC2 M7i-flex and EC2 M7i instances are available in the AWS Regions: US East (Ohio), US East (N. Virginia), US West (Oregon), and Europe (Ireland). Furthermore, pricing details of EC2 instances can be found on the pricing page.