AWS re:Invent 2023 Record after reading
There have some record update from AWS re:Invent on 2023. 在 2023 年的 AWS re:Invent 後有許多的服務提供了新功能，並且有許多的新服務也跟著上限，記錄些我熟悉的服務。
Compute
首先提及的是新實例類型 R8g，他使用了最新的 AWS Graviton 4，比前代 r7g or r6g 有效能上的提升。
96 Neoverse V2 cores, 2 MB of L2 cache per core, and 12 DDR5–5600 channels work together to make the Graviton4 up to 40% faster for databases, 30% faster for web applications, and 45% faster for large Java applications than the Graviton3.
[+] Join the preview for new memory-optimized, AWS Graviton4-powered Amazon EC2 instances (R8g) | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/join-the-preview-for-new-memory-optimized-aws-graviton4-powered-amazon-ec2-instances-r8g/
另外新 u7in 機型，提供高記憶空間，例如 u7in-32tb.224xlarge 有 896 vCPU ，並且提供高 32,768 GiB (32 TB) 的記憶空間。
u7in-xx.224xlarge has up to 12/16/24/32 TB memory in single instance.
[+] Introducing Amazon EC2 high memory U7i Instances for large in-memory databases (preview) | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/introducing-amazon-ec2-high-memory-u7i-instances-for-large-in-memory-databases-preview/
新 chips : AWS Graviton 4 & Trainium 2 (Trn2)
[+] AWS-Unveils-Next-Generation-AWS-Designed-Chips:
https://press.aboutamazon.com/2023/11/aws-unveils-next-generation-aws-designed-chips
AWS Graviton 4: AWS Graviton4 與其前代產品 AWS Graviton3 相比，支援的核心數量增加 50%，每個 CPU 的記憶體頻寬增加 75%。這些功能升級也帶來相當顯著的效能提升，包括網頁應用程式速度提高 30%，資料庫速度提高 40%，大型 Java 應用程式速度提高 45%。
Trainium 2 (Trn2): Trainium2 晶片專為訓練 AI 模型所建構，將支援 EC Trn2 執行個體。相較 2020 年 12 月推出的第一代 Trainium 晶片，第二代的效能提高 4 倍、能源效率提高 2 倍，包括 Databricks、Anthropic 計劃使用這款晶片投入 AI 模型開發。
支援 Oracle/Sun SPARC 作業系統
用 Charon marketplace AMI 來啟動 Linux 環境跑 SPARC 模擬器，然後才把服務轉上來。
[+] Virtualization for SPARC on AWS with Stromasys is now available:
https://aws.amazon.com/cn/about-aws/whats-new/2023/11/virtualization-sparc-aws-stromasys-available/
[+] Migrating Sun/SPARC Mission-Critical Applications from On-Premises to AWS Using SPARC Emulator | AWS Partner Network (APN) Blog:
https://aws.amazon.com/cn/blogs/apn/migrating-sun-sparc-mission-critical-applications-from-on-premises-to-aws-using-sparc-emulator/
AWS + NVIDIA
AWS 加深跟 NVIDIA 的合作，除了 P5e 執行個體配置 NVIDIA P5e(H200 GPUs) 外，也準備推出 NVIDIA G6e(L4 GPUs), G6e(L40S GPUs) 等新的實例類型 。
[+] Introducing three new NVIDIA GPU-based Amazon EC2 instances | AWS Machine Learning Blog:
https://aws.amazon.com/tw/blogs/machine-learning/introducing-three-new-nvidia-gpu-based-amazon-ec2-instances/
Storage
!!! 每一種服務有不同的讀寫延遲(Latency)，挑選適合的服務來使用。
Amazon S3 Express One Zone storage class
利用 single AZ 的方式，來以提高數據抓 (Read) 取速度 10 倍，有助於提升機器學習培訓、交互式分析和媒體內容創建等工作負載效率。
S3 Express One Zone can improve data access speeds by 10x and reduce request costs by 50% compared to S3 Standard and scales to process millions of requests per minute for your most frequently accessed datasets.
[+] Announcing the Amazon S3 Express One Zone storage class:
https://aws.amazon.com/tw/about-aws/whats-new/2023/11/amazon-s3-express-one-zone-storage-class/
EBS Snapshots Archive on AWS Backup
AWS Backup 也可以使用 EBS Snapshots Archive 的功能，讓備份與費用之間取得更進一步的節省。
[+] Amazon EBS Snapshots Archive is now available with AWS Backup | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/amazon-ebs-snapshots-archive-is-now-available-with-aws-backup/
EBS Snapshot Lock
[+] New — Amazon EBS Snapshot Lock | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/new-amazon-ebs-snapshot-lock/
該功能可以 “鎖定” 快照一個指定的時間區間，在這個時間點內是無法被刪除的，用來符合合規使用。
The new EBS Snapshot Lock feature helps you to meet your retention and compliance requirements without the need for custom solutions. You can lock new and existing EBS snapshots using a lock duration that can range from one day to about 100 years. The snapshot is locked for the specified duration and cannot be deleted.
EFS Archive
[+] Amazon EFS Archive | Amazon Web Services:
https://aws.amazon.com/tw/efs/storage-classes/archive/
Amazon Elastic File Storage (Amazon EFS) provides serverless, fully elastic file storage that makes it simple to set up and run file workloads on AWS. Amazon EFS Archive is a storage class that is cost- optimized for cold, rarely-accessed data — giving you a cost-effective way to retain even your coldest data so that it’s always available to power new business insights. EFS Archive delivers storage prices up to 50% lower compared to the cost-optimized EFS Infrequent Access (IA) storage class, and up to 97% lower compared to the high-performance EFS Standard storage class. The EFS Archive storage class costs only $0.008/GB-month1.`
=> EFS 提供 Archive 選項來節省費用。
End User Computing
WorkSpaces Thin Client
一個全新的 WorkSpaces 客戶端機置，只要 $195 美元，讓企業可以統一管理，讓終端使用者透過他可以直接使用 WorkSpaces 功能。
Amazon WorkSpaces Thin Client. The device itself is low cost ($195 in the United States)。
[+] New Amazon WorkSpaces Thin Client provides cost-effective, secure access to virtual desktops | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/new-amazon-workspaces-thin-client/
Cross-region data replication for Amazon WorkSpaces
自動跨 Region (cross-region) 來備份重要在 Amazon WorkSpaces 上的數據及快照。
Using cross-region data replication to provide business continuity for your Amazon WorkSpaces users. Snapshots are taken every 12 hours, replicated to the desired destination region, and are used to provide a recovery point objective (RPO) of 12–24 hours.
[+] Announcing cross-region data replication for Amazon WorkSpaces | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/cross-region-data-replication-for-amazon-workspaces/
Database
ElastiCache Serverless
ElastiCache Serverless 服務，ElastiCache 可能動態地去管理資源，來簡化客戶預估資源大小的困難。
You can use ElastiCache Serverless to operate a cache for even the most demanding workloads without spending time in capacity planning or requiring caching expertise. ElastiCache Serverless constantly monitors your application’s memory, CPU, and network resource utilization and scales instantly to accommodate changes to the access patterns of workloads it serves. You can create a highly available cache with data automatically replicated across multiple Availability Zones and up to 99.99 percent availability Service Level Agreement (SLA) for all workloads, which saves you time and money.
[+] Amazon ElastiCache Serverless for Redis and Memcached is now available | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/amazon-elasticache-serverless-for-redis-and-memcached-now-generally-available/
Aurora Limitless Database
Aurora Limitless Database 提供兩種類型 `Sharded tables` 及 `Reference tables`，來組成新的 `DB shard group (Limitless Database)`，來分散取的負載。
Aurora Serverless資料庫服務之所以能實現線上擴充容量，DeSantis表示，主要是仰賴虛擬化技術。AWS開發了一個稱之為Caspian的虛擬層，由Caspian Heat Management System來管理實體主機如何分配虛擬記憶體給資料庫。Aurora所使用的每一臺實體主機配置容量256GB的記憶體，雖然Aurora資料庫開啟服務可以使用256GB記憶體，但實體主機的記憶體並不會因此而被完全占用，Caspian管理軟體會依據每個Aurora資料庫服務運作所需配置適當的實體記憶體容量，因此每臺實體主機皆可承載多個Aurora資料庫服務，如此就可以極大化利用運算資源。
Amazon Aurora Limitless Database, a new capability supporting automated horizontal scaling to process millions of write transactions per second and manage petabytes of data in a single Aurora database.
[+] Join the preview of Amazon Aurora Limitless Database | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/join-the-preview-amazon-aurora-limitless-database/
RDS for Db2
全託管的數據庫 `IBM DB2`，來用提供特定需要的客戶，來搬移客戶的數據庫到雲端。
IBM DB2 is an enterprise-grade relational database management system (RDBMS) developed by IBM.
[+] Getting started with new Amazon RDS for Db2 | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/getting-started-with-new-amazon-rds-for-db2/
三個 Vector (向量) search / engine 提供 (DocumentDB/MemoryDB and OpenSearch)
Vector search and engine 的功能應該會有 AI/ML 的數據分析，能夠提供到幫忙，他不同於傳統 relational database 存取數據的方式，值得花時間去研究。
[+] Vector search for Amazon DocumentDB (with MongoDB compatibility) is now generally available | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/vector-search-for-amazon-documentdb-with-mongodb-compatibility-is-now-generally-available/
Vector search is an emerging technique used in machine learning (ML) to find similar data points to given data by comparing their vector representations using distance or similarity metrics. Vectors are numerical representation of unstructured data created from large language models (LLM) hosted in Amazon Bedrock, Amazon SageMaker, and other open source or proprietary ML services. This approach is useful in creating generative artificial intelligence (AI) applications, such as intuitive search, product recommendation, personalization, and chatbots using Retrieval Augmented Generation (RAG) model approach. For example, if your data set contained individual documents for movies, you could semantically search for movies similar to Titanic based on shared context such as “boats”, “tragedy”, or “movies based on true stories” instead of simply matching keywords.
[+] AWS announces vector search for Amazon MemoryDB for Redis (Preview):
https://aws.amazon.com/tw/about-aws/whats-new/2023/11/vector-search-amazon-memorydb-redis-preview/
Amazon MemoryDB for Redis now supports vector search in preview, a new capability that enables you to store, index, and search vectors. MemoryDB is a database that combines in-memory performance with multi-AZ durability. With vector search for MemoryDB, you can develop real-time machine learning (ML) and generative AI applications with the highest performance demands using the popular, open-source Redis API. Vector search for MemoryDB supports storing millions of vectors, with single-digit millisecond query and update response times, and tens of thousands queries per second (QPS) at greater than 99% recall. You can generate vector embeddings using AI/ML services like Amazon Bedrock and SageMaker, and store them within MemoryDB.
[+] Vector engine for Amazon OpenSearch Serverless is now available | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/vector-engine-for-amazon-opensearch-serverless-is-now-generally-available/
The vector engine uses OpenSearch Compute Units (OCUs), compute capacity unit, to ingest and run similarity search queries. One OCU can handle up to 2 million vectors for 128 dimensions or 500,000 for 768 dimensions at 99 percent recall rate.
Analytics
Amazon Q
Amazon Q: AI 驅動的個人助理，透過生成式 AI 來協助企業裡的員工完成日常任務。
[+] Amazon QuickSight Q — 商業智慧服務 — Amazon Web Services:
https://aws.amazon.com/tw/quicksight/q/
[+] 亞馬遜終於推出聊天機器人「Amazon Q」，比ChatGPT強在哪？｜天下雜誌:
https://www.cw.com.tw/article/5128307
Amazon Q的目標是幫助企業裡的員工完成日常任務，例如摘要策略文件、填寫內部支援表單，和回答有關公司政策的問題。
Amazon Q + QuickSight
[+] New Amazon Q in QuickSight uses generative AI assistance for quicker, easier data insights (preview) | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/new-amazon-q-in-quicksight-uses-generative-ai-assistance-for-quicker-easier-data-insights-preview/
Amazon Bedrock
Amazon Bedrock : 大型語言訓練模型 : 是一項全受管服務，透過單一 API 提供來自 AI21 Labs、Anthropic、Cohere、Meta、Stability AI 和 Amazon 等領先 AI 公司的各種高效能基礎模型（FM）選擇，以及提供建置生成式 AI 應用程式所需的各種廣泛功能，在維護隱私和安全的同時簡化開發。
[+] 使用基礎模型建置生成式 AI 應用程式 — Amazon Bedrock — AWS:
https://aws.amazon.com/tw/bedrock/
CodeWhisperer
CodeWhisperer 經過數十億行程式碼訓練，可根據您的註解和現有程式碼，即時產生從程式碼片段到完整功能的程式碼建議。略過耗時的編碼任務，並利用不熟悉的 API 加速建置。
[+] AI 程式碼產生器 — Amazon CodeWhisperer — AWS:
https://aws.amazon.com/tw/codewhisperer/
[+] Amazon投資40億美元給OpenAI競爭者Anthropic | iThome:
https://www.ithome.com.tw/news/158977
新聞:
Amazon 投資給 OpenAI競爭者Anthropic，將借其力研發AWS服務及自家AI晶片。
Security
[+] Amazon Detective adds new capabilities to accelerate and improve your cloud security investigations | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/amazon-detective-adds-investigations-and-finding-group-summaries-to-help-you-investigate-security-findings/
[+] Detect runtime security threats in Amazon ECS and AWS Fargate, new in Amazon GuardDuty | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/introducing-amazon-guardduty-ecs-runtime-monitoring-including-aws-fargate/
[+] IAM Access Analyzer updates: Find unused access, check policies before deployment | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/iam-access-analyzer-updates-find-unused-access-check-policies-before-deployment/
[+] AWS Identity and Access Management Access Analyzer — Amazon Web Services:
https://aws.amazon.com/tw/iam/features/analyze-access/
IAM Access Analyzer 使用可證明的安全性來提供對您資源的公有和跨帳戶存取的全面問題清單。可證明的安全性依賴自動推理技術，即應用數學邏輯來協助解決有關基礎設施的關鍵問題，包括 AWS 許可。
Others
Amazon Time Sync Service
[+] Introducing the Amazon Time Sync Service:
https://aws.amazon.com/about-aws/whats-new/2017/11/introducing-the-amazon-time-sync-service/?nc1=h_ls
Posted On: Nov 29, 2017
The Amazon Time Sync Service provides a highly accurate and reliable time reference that is natively accessible from Amazon EC2 instances.
[+] Amazon Time Sync Service now supports microsecond-accurate time:
https://aws.amazon.com/tw/about-aws/whats-new/2023/11/amazon-time-sync-service-microsecond-accurate-time/
Amazon Time Sync Service 在 2017 推出，並在 2023 更進一步支援 ms 等級的時間差。
re:Post Private
[+] Increase collaboration and securely share cloud knowledge with AWS re:Post Private | AWS News Blog:
https://aws.amazon.com/tw/blogs/aws/increase-collaboration-and-securely-share-cloud-knowledge-with-aws-repost-private/
AWS re:Post Private 提供企業內部的知識分享工具，使用介面就如同 AWS re:Post 一樣。
