EC2 updates in AWS re:Invent 2022
AWS re:Invent 2022 有許多EC2實例類型、及技術的更新，這邊做些記錄。
https://aws.amazon.com/tw/blogs/aws/top-announcements-of-aws-reinvent-2022/#Compute
#1 Nitro version 5 (New version)
Nitro system card 是底層虛擬化的”硬體卡片”，新的一代 Nitro card 可以提升效能外，也提升介接其他硬體/服務的能力及速度。
There are three components describing the Performance Gains of Nitro v5 :
- PPS — Nitro v5 has a 60% higher PPS.
- Latency — 30% lower latency.
- Performance — It has 40% better performance per Watt
!!! 跟前幾代的比較，例如第6代的機型，是使用 Nitro version 4。
!!! Nitro system 包含了以下東西。
* Nitro Card for VPC
* Nitro Card for EBS
* Nitro Controller
* Nitro Security Chip
* Nitro Hypervisor (KVM-based)
# 2 New instance types of generation 7: C7gn, R7iz, and Hpc7g
第7代新的實例類型，除了 ARM 類型的 c7g, c7gn 外，也有 INTEL 的 CPU r7iz，這些都是使用最新的 Nitro version five
C7gn — ARM with Nitro version 5
https://aws.amazon.com/blogs/aws/new-amazon-ec2-instance-types-in-the-works-c7gn-r7iz-and-hpc7g/
C7gn Instances are designed for your most demanding network-intensive workloads: network virtual appliances (firewalls, virtual routers, load balancers, and so forth), data analytics, and tightly-coupled cluster computing jobs. They are powered by AWS Graviton3E processors and will support up to 200 Gbps of network bandwidth, along with 50% higher packet processing performance. The c7gn instances will be available in multiple sizes with up to 64 vCPUs and 128 GiB of memory. We are launching the preview today and you can Sign Up Today to join in.
https://aws.amazon.com/ec2/instance-types/c7g/?nc1=h_ls
C7g Instances — ARM with Nitro version 5
https://aws.amazon.com/ec2/instance-types/c7g/
Amazon Elastic Compute Cloud (Amazon EC2) C7g instances, powered by the latest generation AWS Graviton3 processors, provide the best price performance in Amazon EC2 for compute-intensive workloads.
Hpc7g — ARM with Nitro version 5
Hpc7g Instances are also powered by AWS Graviton3E processors, with up to 35% higher vector instruction processing performance than the Graviton3. They are designed to give you the best price/performance for tightly coupled compute-intensive HPC and distributed computing workloads, and deliver 200 Gbps of dedicated network bandwidth that is optimized for traffic between instances in the same VPC. The hpc7g instances will be available in multiple sizes with up to 64 vCPUs and 128 GiB of memory. I’ll have more information to share on these instances in early 2023.
R7iz — Intel CPU with Nitro version 5
https://aws.amazon.com/about-aws/whats-new/2022/11/introducing-amazon-ec2-r7iz-instances/
R7iz Instances are powered by the latest 4th generation Intel Xeon Scalable Processors (code named Sapphire Rapids) and run at a sustained all-core turbo frequency of 3.9 GHz. With high performance and DDR5 memory, these instances are a perfect match for your Electronic Design Automation (EDA), financial, actuarial, and simulation workloads. They are also great hosts for relational databases and other commercial software that is licensed on a per-core basis. The r7iz instances will be available in multiple sizes with up to 128 vCPUs and 1 TiB of memory. We are launching the instances in preview today and you can Sign up Today to participate.
#3 New instance types of generation 6
第6代的實例類型，除了 ARM 類型外，其他例如 INTL 及 AMD 的 CPU 類型，也都上線上，用來取代前一代 5系統(例如 c5/m5/r5….)常用機型。
!!! 請注意，新機型的數量一開始會比較少，所以在轉換前要先測試、並且分批來轉換。例如原本使用的是 m5(x86)，可以考慮改使用 m6i (INTEL)或是 m6a (AMD)的機型。
General Purpose: m6i、m6in、m6a
Computer Optimized: c6i、c6a、hpc6a
Memory Optimized: r6i、r6in、r6a
Hpc6id
https://aws.amazon.com/about-aws/whats-new/2022/11/announcing-amazon-ec2-hpc6id-instances/
https://aws.amazon.com/ec2/instance-types/hpc6i/
Amazon Elastic Compute Cloud (Amazon EC2) Hpc6id instances, powered by 3rd Generation Intel Xeon Scalable processors, offer cost-effective price performance for memory-bound and data-intensive high performance computing (HPC) workloads in Amazon EC2. These instances are built on the AWS Nitro System, providing 200 Gbps for low-latency inter-node communication.
# 4 Other new instances
其他例如 Deep learning / ML 所使用的機型，也有更新。
Inf2
https://aws.amazon.com/about-aws/whats-new/2022/11/aws-announces-amazon-ec2-inf2-instances-preview/
https://aws.amazon.com/ec2/instance-types/inf2/Amazon Elastic Compute Cloud (Amazon EC2) Inf2 instances are purpose built for deep learning (DL) inference. They are designed to deliver high performance at the lowest cost in Amazon EC2 for your most demanding DL applications. You can use Inf2 instances to run your inference applications for natural language understanding, language translation, video and image generation, speech recognition, personalization, fraud detection, and more.
I4i
https://aws.amazon.com/about-aws/whats-new/2022/11/aws-announces-torn-write-prevention-ec2-i4i-instances-ebs-amazon-rds/
https://aws.amazon.com/ec2/instance-types/i4i/
Amazon EC2 I4i instances are powered by 3rd generation Intel Xeon Scalable processors and feature up to 30 TB of local AWS Nitro SSD storage. Nitro SSDs are NVMe-based and custom-designed by AWS to provide high I/O performance, low latency, minimal latency variability, and security with always-on encryption. I4i instances deliver up to 30% better compute price performance, 60% lower storage I/O latency, and 75% lower storage I/O latency variability compared to I3 instances.
#5 SRD — Multi-Path through network
透過 Scalable Reliable Datagram (SRD) protocol 的技術，來提升內部網路的傳輸能力，進而降低服務之間的存取的Latency。
- Reduces P99 latency of traffic flows by up to 50% and P99.9 latency by up to 85% (in comparison to TCP)
- Increasing the maximum single-flow bandwidth from 5 Gbps to 25 Gbps.
Amazon’s own SRD makes use of multi-path routing and doesn’t transmit packets in order, but can tidy things up when packets arrive out of order. DeSantis claimed it will retransmit dropped packets “in microseconds, not milliseconds” and speed up networks hosted on the AWS cloud.
!!! 而這個技術會導入 EBS io2 volume、EFA、ENA Express 等等，接下來 AWS manager 代管的服務，例如 ElastiCache… 應該都會慢慢導入。
EBS io2 volumes will be running on SRD
- 90% lower EBS tail latency with SRD.
- latency 會再下降，而客戶端實例不用改，直接用就可以了。
Elastic Network Adapter — ENA Express
Introducing ENA Express
https://aws.amazon.com/blogs/aws/new-ena-express-improved-network-latency-and-per-flow-performance-on-ec2/
Today we are launching ENA Express. Building on the Scalable Reliable Datagram (SRD) protocol that already powers Elastic Fabric Adapters, ENA Express reduces P99 latency of traffic flows by up to 50% and P99.9 latency by up to 85% (in comparison to TCP), while also increasing the maximum single-flow bandwidth from 5 Gbps to 25 Gbps. Bottom line, you get a lot more per-flow bandwidth and a lot less variability.
