I will describe in this article the machine families, machine series, and machine types that you can choose from to create a virtual machine (VM) instance with the resources you need. When you create a VM, you select a machine type from a machine family that determines the resources available to that VM. There are several machine families you can choose from and each machine family is further organized into machine series and predefined machine types within each series. For example, within the N2 series in the general-purpose machine family, you can select n2-standard-4 machine type.
All machine series support preemtible VMs, with the exception of the M2 machine series:
1. General-purpose: best price-performance ratio for a variety of workloads.
2. Compute-optimized: highest performance per core on Compute Engine and optimized for compute-intensive workloads.
3. Memory-optimized: ideal for memory-intensive workloads, offering more memory per core than other machine families, with up to 12 TB of memory.
4. Accelerator-optimized: ideal for massively parallelized Compute Unified Device Architecture (CUDA) compute workloads, such as machine learning (ML) and high performance computing (HPC). This family is the best option for workloads that require GPUs.
This article uses the following terms:
1. Machine family: A curated set of processor and hardware configurations optimized for specific workloads. When you create a VM instance, you choose a predefined or custom machine type from your preferred machine family.
2. Series: Machine families are further classified by series and generation. For example, the N1 series within the general-purpose machine family is the older version of the N2 series. Generally, generations of a machine series use a higher number to describe the newer generation. For example, the N2 series is the newer generation of the N1 series.
3. Machine type: Every machine series has predefined machine types that provide a set of resources for your VM. if a predefined machine type doesn't meet your needs, you can also create a custom machine type.
Machine family and series recommendations
The following table provides recommendations for different workloads:
General-purpose machine family guide:
The general-purpose machine family offers several machine series with the best price-performance ratio for a variety of workloads.
Compute Engine offers general-purpose machine families that run on either x86 or Arm architecture.
x86:
1. Cost-optimized E2 machine series has up to 32 vCPUs with up to 128 GB of memory with a maximum of 8 GB per vCPU. The E2 machine series has a predefined CPU platform, running either an Intel processor or the second generation AMD EPYC Rome processor. The processor is selected for you when you create the VM. This machine series provides a variety of compute resources for the lowest price on Compute Engine especially when paired with committed-use discounts.
2. N2 machine series has up to 128 vCPUs, 8 GB of memory per vCPU, and is available on the Intel Ice Lake and Cascae Lake CPU platforms.
3. N2D machine series has up to 224 vCPUs, 8 GB of memory per vCPU, and is available on second generation AMD EPYC Rome and third generation AMD EPYC Milan platforms.
4. Tau T2D machine series provides an optimized feature set for scaling out. Each VM can have up to 60 vCPUs, 4 GB of memory per vCPU, and is available on third generation AMD EPYC Milan processors. The Tau T2D machine series has cluster-threading disabled, therefore a vCPU is equivalent to an entire core.
5. N1 machine series have up to 96 vCPUs, 6.5 GB of memory per vCPU, and are available on Intel Sandy Bridge, Ivy Bridge, Haswell, Broadwell, and Skylake CPU platforms.
The E2 and N1 series contain shared-core machine types. These machine types timeshare a physical core which can be cost-effective method for running small, non-resource intensive apps:
1. E2: offers 2 vCPUs for short periods of bursting.
2. N1: offers f1-micro and g1-small shared-core machine types which have up to 1 vCPU available for short periods of bursting.
Compute-optimized machine family guide
The compute-optimized machine family has the highest performance per core on Compute Engine and is optimized for compute-intensive workloads. Machine series in this family runs on either an Intel Scalable Processor (Cascade Lake) that can sustain up to 3.9 GHz all-core turbo, or the 3rd generation AMD EPYC Milan processor offering up to 3.5 GHz max boost frequency.
1. C2 VMs offer up to 60 vCPUs, 4 GB of memory per vCPU, and are available on the Intel Cascade Lake CPU platform.
2. C2D VMs offer up to 112 vCPUs, 4 GB of memory per vCPU, and are available on the third generation AMD EPYC Milan platform.
Memory-optimized and accelerator-optimized machine families guide:
The memory-optimized machine series that are ideal for OLAP and OLTP SAP workloads, genomic modeling, electronic design automation, and your most memory intensive HPC workloads. This family offers more memory per core than any other machine family, with up to 12 TB of memory.
1. M1 VMs offer up to 160 vCPUs, 14.9 GB to 24 GB of memory per vCPU, and are available on the Intel Skylake and Broadwell CPU platforms.
2. M2 VMs are available as 6 TB, 9TB, and 12 TB machine types, and are available on the Intel Cascade Lake CPU platform.
3. M3 VMs offer up to 128 vCPUs, with up to 30.5 GB of memory per vCPU, and are available on the Intel Ice Lake CPU platform.
The accelerator-optimized machine family is ideal for massively parallelized Compute Unified Device Architecture (CUDA) compute workloads, such as machine learning (ML) and high performance computing (HPC). This family is the optimal choice for workloads that require GPUs:
A2 VMs offer 12 to 96 vCPUs, up to 1360 GB of memory, and are available on the Intel Cascade Lake CPU platform.
Choosing VM properties to compare based:
1. Workload types,
2. CPU types,
3. Architecture,
4. vCPUs,
5. vCPU definition,
6. Memory,
7. Extended memory,
8. Sole tenancy,
9. Nested virtualization,
10. Custom machine types,
11. Confidential Compute,
12. Confidential Compute,
13. Disk interface type,
14. Local SSD,
15. Max local SSD,
16. Standard PDs,
17. Balanced PDs,
18. SSD PDs,
19. Extreme PDs,
20. Network interfaces,
21. Network performance,
22. High-bandwidth network,
23. Max GPUs,
24. Max GPUs,
25. Discounts,
26. Coremark score.
GPUs and VMs
GPUs are used to accelerate workloads. You can only attach GPUs to VMs using the N1 machine series or the A2 machine series. GPUs are not supported by other machine series.
VMs with lower numbers of GPUs are limited to a maximum number of vCPUs. In general, a higher number of GPUs lets you create instances with a higher number of vCPUs and memory.
Comments
Post a Comment