Home Hardware GCP

Google Cloud Platform

Google Cloud Platform (GCP) offers an extensive range of computational resources, ensuring versatility and efficiency for diverse computing needs. Currently, Inductiva provides access to a selection of resource families, including c2, c2d, c3, c3d, n2, n2d, and e2.

  • c2: 2nd-generation Intel Xeon Scalable processor (Cascade Lake) which offers up to 3.9 GHz sustained single-core max turbo frequency.
  • c2d: 3rd-generation AMD EPYC Milan processor offers up to 3.5 GHz max boost frequency.
  • c3: 4th generation Intel Xeon Scalable processors and offers a sustained, all-core turbo frequency of 3.0 GHz, 8 channels of DDR5 memory.
  • c3d: 4th generation AMD EPYC™ (Genoa) processor with a maximum frequency of 3.7 Ghz.
  • n2: Ice Lake (default for larger machines) or Cascade Lake (default for machines up to 80 vCPUs)
  • n2d: AMD EPYC Milan or AMD EPYC Rome processors that run with a base frequency of 2.25 GHz, an effective frequency of 2.7 GHz, and a max boost frequency of 3.3 GHz.

A more detailed explanation about each resource family can be found here and here.

For a complete list of supported resources do (using the CLI):

	

		
		

$ inductiva resources list available

or, using the Python API:

	

		
		

>>> import inductiva
>>> inductiva.resources.list_available_machines(provider="gcp")

While exploring the list of available resources, one can observe certain families offering configurations labeled as highmem or highcpu. While the standard provides a more well-rounded configuration (a decent amount of system memory per vCPU) its counterparts provide a more focused configuration depending on your needs. Where highmem provides more system memory per vCPU, highcpu provides less. The memory tradeoffs come at a lower cost for machines with less memory. For example, while both the c2d-highmem-4 and c2d-highcpu-4 offer equivalent processing power, the former commands a higher price due to its bigger memory capacity.

Presently, all resources deployed via GCP utilize simultaneous multithreading, a technology where two virtual CPUs (vCPUs) share a single physical CPU core.