nerdexam
Amazon

MLS-C01 · Question #224

A company is using Amazon SageMaker to build a machine learning (ML) model to predict customer churn based on customer call transcripts. Audio files from customer calls are located in an on-premises…

The correct answer is A. Order and use an AWS Snowball Edge Compute Optimized device with an NVIDIA Tesla module. To quickly transcribe petabytes of on-premises audio data requiring GPUs and transfer the resulting transcriptions to S3, an AWS Snowball Edge Compute Optimized device with an NVIDIA Tesla module is the most efficient solution.

Data Engineering

Question

A company is using Amazon SageMaker to build a machine learning (ML) model to predict customer churn based on customer call transcripts. Audio files from customer calls are located in an on-premises VoIP system that has petabytes of recorded calls. The on-premises infrastructure has high-velocity networking and connects to the company's AWS infrastructure through a VPN connection over a 100 Mbps connection. The company has an algorithm for transcribing customer calls that requires GPUs for inference. The company wants to store these transcriptions in an Amazon S3 bucket in the AWS Cloud for model development. Which solution should an ML specialist use to deliver the transcriptions to the S3 bucket as quickly as possible?

Options

  • AOrder and use an AWS Snowball Edge Compute Optimized device with an NVIDIA Tesla module
  • BOrder and use an AWS Snowcone device with Amazon EC2 Inf1 instances to run the
  • COrder and use AWS Outposts to run the transcription algorithm on GPU-based Amazon EC2
  • DUse AWS DataSync to ingest the audio files to Amazon S3. Create an AWS Lambda function to

How the community answered

(29 responses)
  • A
    62% (18)
  • B
    24% (7)
  • C
    3% (1)
  • D
    10% (3)

Why each option

To quickly transcribe petabytes of on-premises audio data requiring GPUs and transfer the resulting transcriptions to S3, an AWS Snowball Edge Compute Optimized device with an NVIDIA Tesla module is the most efficient solution.

AOrder and use an AWS Snowball Edge Compute Optimized device with an NVIDIA Tesla moduleCorrect

AWS Snowball Edge Compute Optimized devices are designed for large-scale data transfer and on-premises processing, offering compute capabilities with optional NVIDIA Tesla GPUs. This enables the GPU-intensive transcription algorithm to run locally, avoiding the slow VPN connection for petabytes of raw audio, and then the devices can be shipped back to AWS for efficient ingestion of processed data into S3.

BOrder and use an AWS Snowcone device with Amazon EC2 Inf1 instances to run the

AWS Snowcone is a smaller device for edge computing and data transfer, but it does not offer GPU capabilities suitable for the required intensive transcription workload.

COrder and use AWS Outposts to run the transcription algorithm on GPU-based Amazon EC2

AWS Outposts extends AWS infrastructure on-premises, allowing GPU-based EC2 instances, but transferring petabytes of data over a 100 Mbps VPN to Outposts and then to S3 would be prohibitively slow for the initial data ingestion.

DUse AWS DataSync to ingest the audio files to Amazon S3. Create an AWS Lambda function to

AWS DataSync would ingest data over the existing 100 Mbps VPN connection, which is too slow for petabytes of data, and it does not provide on-premises GPU compute for transcription before ingestion.

Concept tested: Large-scale on-premises data processing and transfer with GPUs

Source: https://aws.amazon.com/snow/snowball-edge/features/

Topics

#Data Transfer#Edge Computing#GPU Compute#Hybrid Cloud

Community Discussion

No community discussion yet for this question.

Sign up to join the discussion
Full MLS-C01 Practice