Microsoft Azure AI Fundamentals (AI-900)Describe fundamental principles of machine learning on AzureHard
A data engineering team is building a machine learning pipeline that needs to efficiently store and manage large volumes of unstructured data, such as images, videos, and sensor readings, for training deep learning models. The solution must be highly scalable and cost-effective. Which Azure storage solution is best suited for this purpose?
- AAzure Cosmos DB
- BAzure Data Lake Storage Gen2
- CAzure Blob Storage
- DAzure SQL Database
Show answer & explanationAnswer & explanation
Correct answer: C. Azure Blob Storage
Azure Blob Storage is specifically designed for storing massive amounts of unstructured data, including images, videos, and sensor readings, at scale and at a low cost. It's ideal for data lakes, high-performance computing, and training deep learning models that consume large binary files.
Why the other options are wrong
- A. Azure Cosmos DB is a NoSQL database service, suitable for semi-structured or unstructured document/key-value data, but not optimized for large binary objects like images/videos for ML training.
- B. Azure Data Lake Storage Gen2 is built on Azure Blob Storage and optimized for big data analytics workloads. While it uses Blob Storage, the general Blob Storage service itself directly fits the need for storing unstructured data for ML training without necessarily needing the full Gen2 analytics optimizations unless specifically stated.
- D. Azure SQL Database is a relational database service, suitable for structured data, not large unstructured files.
Azure Blob Storage
Azure Blob Storage is Microsoft's object storage solution for the cloud. It is optimized for storing massive amounts of unstructured data, such as text or binary data, including images, video files, audio files, and large data sets.
- Highly scalable and durable.
- Cost-effective for large volumes of data.
- Supports various access tiers (Hot, Cool, Archive).
- Ideal for data lakes, backup/restore, disaster recovery, and deep learning data storage.
Memory trick: Blob: A 'blob' of unstructured data, floating in the cloud, ready for ML.