Prepstellar
Get the app

DP-750 · Practice set 7 of 11

Configuring Compute Features: 10 practice questions

10 questions · Untimed · Free

10 free DP-750 practice questions on Configuring Compute Features, with an explanation for every answer. Untimed. The full mock exam and the timed version are in the app.

Set 7 · Configuring Compute Features · 10 questions Read the lesson

  1. Question 1 of 10

    Which operations are directly accelerated by Photon on supported compute?

    1. ACluster permission grants and workspace bindings
    2. BSQL queries and DataFrame operations
    3. CGPU driver installation and CUDA compilation
    4. DRDD operations and R notebook execution
    Show the answer

    Photon replaces traditional Spark execution components with optimized native code for SQL and DataFrame work; it is not an access-control or GPU-environment feature.

    Next → 1 / 10

  2. Question 2 of 10

    What else is selected when a Databricks Runtime version is chosen?

    1. AA compute permission level
    2. BA Unity Catalog workspace binding
    3. CIts associated Apache Spark version
    4. DA separate serverless subscription
    Show the answer

    The runtime packages Spark and platform optimizations together, so runtime selection determines the Spark version; access and workspace controls are independent.

    Next → 2 / 10

  3. Question 3 of 10

    Which runtime includes preinstalled machine learning libraries, GPU drivers, and CUDA frameworks?

    1. AA serverless SQL warehouse runtime
    2. BDatabricks Runtime ML
    3. CAn instance-pool preloaded runtime only
    4. DA standard Databricks Runtime
    Show the answer

    Databricks Runtime ML provides the prepared software stack for ML and GPU work; a standard runtime does not represent that specialized bundle.

    Next → 3 / 10

  4. Question 4 of 10

    Which workload pattern is most likely to benefit from Photon?

    1. AA one-row notebook calculation that finishes immediately
    2. BChanging cluster permission levels for a group
    3. CRepeated large-table joins and aggregations
    4. DNeural-network training on GPU instances
    Show the answer

    Photon is most valuable for substantial SQL and DataFrame scans, joins, and aggregations; trivial work has little opportunity for acceleration and GPU compute is incompatible.

    Next → 4 / 10

  5. Question 5 of 10

    Why is an LTS runtime a suitable choice for operational job compute?

    1. AIt removes the need to test production upgrades
    2. BIt emphasizes extended compatibility and stability
    3. CIt automatically converts every cluster to serverless
    4. DIt enables Photon on GPU clusters
    Show the answer

    An LTS release supports a stability-first production posture, while current runtimes favor the newest features; production changes should still be tested.

    Next → 5 / 10

  6. Keep the ones you got wrong

    In the app, every question you miss comes back exactly when you’re about to forget it.

    Get the app

  7. Question 6 of 10

    Which compute behavior prevents direct runtime-version selection?

    1. AServerless uses an automatically upgraded versionless runtime
    2. BDatabricks Runtime ML removes Apache Spark
    3. CClassic compute requires explicit runtime selection
    4. DDedicated access mode fixes all clusters to one specific LTS version
    Show the answer

    Serverless abstracts version management through a versionless runtime; classic compute is the choice when the team must select and manage the version.

    Next → 6 / 10

  8. Question 7 of 10

    For which initial ML experiment can a large single-node resource be appropriate?

    1. AA shared SQL dashboard with many concurrent users
    2. BA framework whose workload does not distribute across workers
    3. CA large Spark job that requires horizontal worker scaling
    4. DA cluster that must tolerate driver loss through worker redundancy
    Show the answer

    A single-node resource avoids unnecessary shuffle for a non-distributed ML framework, but it cannot provide horizontal Spark scaling or driver redundancy.

    Next → 7 / 10

  9. Question 8 of 10

    A production job performs wide-table scans and complex aggregations on CPUs. It does not use GPUs, and the team wants faster SQL and DataFrame execution without changing application logic. Which feature should be enabled?

    1. AA dedicated group access mode only
    2. BA newer Spark version with Photon disabled
    3. CDatabricks Runtime ML with GPU drivers
    4. DPhoton acceleration
    Show the answer

    Photon directly accelerates the stated CPU-based SQL and DataFrame operators, and the absence of GPUs removes its incompatibility constraint.

    Next → 8 / 10

  10. Question 9 of 10

    A team is moving a daily production pipeline to a newer runtime. The pipeline needs stability, but a recent Spark feature is also required. Which approach best manages both constraints?

    1. AKeep the old runtime and enable a different access mode
    2. BUse the newest non-LTS runtime directly in production
    3. CChoose an LTS runtime containing the feature and test it in development first
    4. DMove to a versionless runtime and select its Spark release
    Show the answer

    The runtime determines Spark features, while an LTS release and development validation address production stability; access mode cannot add Spark functionality.

    Next → 9 / 10

  11. Question 10 of 10

    A notebook trains an image-recognition neural network. It needs CUDA and GPU drivers, while SQL acceleration is not a requirement. Which feature combination is valid?

    1. AA standard runtime on GPU compute with Photon enabled
    2. BDatabricks Runtime ML on GPU compute with Photon enabled
    3. CA standard CPU runtime with Photon as the GPU framework
    4. DDatabricks Runtime ML on GPU compute with Photon disabled
    Show the answer

    Runtime ML supplies CUDA and GPU drivers, and Photon must be disabled because it is unsupported on GPU-enabled clusters; the SQL acceleration tradeoff is irrelevant here.

    Next → 10 / 10

  12. You’ve finished this set

    That’s 10 questions on Configuring Compute Features. In the app the ones you miss come back exactly when you’re about to forget them.

    Back to the course Next lesson: Compute Libraries and Access PermissionsGet the app
Keep your progress · Get the app

The whole course, on your phone

Lessons you can read, audio you can listen to on the way to work, and practice that remembers what you got wrong.

Get the app