Google ACE Practice Question: A data science team needs a VM with 96 vCPUs and…

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Select the closest larger predefined N2 machine type

  Why it's wrong here

  Predefined N2 machine types bundle vCPU and memory in fixed ratios, such as highmem providing 8 GB per vCPU. At 96 vCPUs, the closest larger predefined option is 768 GB of memory (highmem), which over-provisions by 144 GB versus the 624 GB requirement, increasing cost and potentially pushing into higher pricing tiers. A standard type would only provide 384 GB, so you cannot hit the exact requirement with a predefined family, making this choice wasteful.

• ✓

  Create a custom machine type with exactly 96 vCPUs and 624 GB RAM

  Why this is correct

  Custom machine types on N2 let you specify vCPU count and memory independently, with memory configurable in granular steps up to 8 GB per vCPU. Setting exactly 96 vCPUs and 624 GB RAM satisfies the workload's memory footprint without over-allocation, and you are billed only for those specific resources. This is the only option that precisely matches the stated requirement while avoiding the cost and waste of a larger predefined instance.

• ✗

  Split the workload across multiple smaller VMs and coordinate manually

  Why it's wrong here

  Manually splitting the workload across smaller VMs introduces coordination overhead and data synchronisation latency that undermines the single-system memory consistency required for the 624 GB RAM footprint. This approach is tempting because it mirrors a common strategy for scaling stateless, horizontally partitioned tasks like batch processing or web serving, where independent smaller instances can be orchestrated without shared-memory dependencies.

• ✗

  Contact Google Cloud support to request a new predefined machine type

  Why it's wrong here

  Google Cloud does not offer a mechanism to request a new predefined machine type for individual customers; support can assist with quotas and service issues but cannot add custom SKUs to the catalog. Since custom machine types already provide the exact 96 vCPU/624 GB combination, asking support to build a new predefined type is an unnecessary step that adds delay without solving the actual requirement. The scenario calls for selecting an existing offering, not extending the platform's predefined catalog.

Go deeper

Related to this question