> ## Documentation Index
> Fetch the complete documentation index at: https://docs.gcore.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Types of Virtual Machines
Gcore Virtual Machines (VMs) offer a flexible, powerful, and scalable solution for hosting applications and services. Designed to meet a wide range of computing needs, VMs deliver optimal performance, reliability, and security.
The following VM types are available for different workflows and computing tasks:
* [Standard](/cloud/virtual-instances/types-of-virtual-machines#standard): Suitable for a wide range of basic workloads that require predictable compute performance.
* [GPU](/cloud/virtual-instances/types-of-virtual-machines#gpu): Virtual Machines with GPUs process certain types of tasks more efficiently than CPUs and are used for workloads that require the resources of graphics accelerators. These machines have dedicated graphics cards and are ideal for complex computations such as CAD, AI/ML, or rendering.
* [CPU-Optimized](/cloud/virtual-instances/types-of-virtual-machines#cpu-optimized): Best suited for CPU-intensive tasks that require predictable computing performance, such as batch processing of large data sets and video encoding.
* [Memory-Optimized](/cloud/virtual-instances/types-of-virtual-machines#memory-optimized): Ideal for tasks and projects that require intensive memory usage, such as databases.
* [High-frequency](/cloud/virtual-instances/types-of-virtual-machines#high-frequency): Best fit for applications that require single-threaded performance.
* [Basic VM](/cloud/virtual-instances/types-of-virtual-machines#basic-vm): Machines that work with partial core use and are designed to run applications that do not require high performance.
* [VMs with a Trusted Platform Module (TPM)](/cloud/virtual-instances/types-of-virtual-machines#tpm): Suitable for a diverse array of secure workloads that demand robust hardware-based security features.
* [Network optimised](/cloud/virtual-instances/types-of-virtual-machines#network-optimised): Designed for workloads that require high network throughput (up to 10 Gbps) and low latency.
The network bandwidth for all Virtual Machines, except Basic VM, is up to 1 Gbps. For Basic VM, bandwidth is up to 200 Mbps.
## Network specifications
The following table summarizes the network bandwidth and traffic policy for VMs:
| **VM type** | **Network speed** | **Incoming traffic** | **Outgoing traffic** |
| ------------------------------------------------------------------------ | ----------------- | -------------------- | -------------------- |
| Basic VM | Up to 200 Mbps | Free, unlimited | Free, unlimited |
| Standard, CPU-Optimized, Memory-Optimized,
GPU, High-frequency, TPM | Up to 1 Gbps | Free, unlimited | Free, unlimited |
| Network optimised (gX-net) | Up to 10 Gbps | Free, unlimited | Free, unlimited |
Bandwidth values indicate the maximum achievable speed for standard flavors. Network optimised VMs have bandwidth enforced through QoS policies. Actual throughput for standard VMs may vary depending on network conditions and workload.
Network traffic for all VMs is free and unlimited — no ingress or egress charges apply.
The following sections provide a detailed overview of each VM type and applicable use cases.
## Standard
These Virtual Machines are designed for a wide range of tasks that require predictable performance. They balance compute, memory, and networking resources and can be used as the default choice for most workloads.
**Use cases**
Applications that use these resources in equal proportions, such as web servers and code repositories.
Machines equipped with the industry's first cloud-native, **128-core ARM Ampere Altra Max Family** processors. Based on 64-bit ARM architecture that's optimized for cloud workloads, these VMs deliver predictable high performance to meet the demands of modern data centers and scalable cloud applications.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ----------------- | -------- | ------------------- |
| a1-standard-1-2 | 1 | 2 |
| a1-standard-2-4 | 2 | 4 |
| a1-standard-2-8 | 2 | 8 |
| a1-standard-4-8 | 4 | 8 |
| a1-standard-4-16 | 4 | 16 |
| a1-standard-8-16 | 8 | 16 |
| a1-standard-8-32 | 8 | 32 |
| a1-standard-16-32 | 16 | 32 |
| a1-standard-16-64 | 16 | 64 |
| a1-standard-32-64 | 32 | 64 |
Machines equipped with the **2nd Gen Intel® Xeon® Scalable Processors**. Optimized for demanding workloads such as high-performance computing, advanced analytics, and artificial intelligence.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ----------------- | -------- | ------------------- |
| g1-standard-1-2 | 1 | 2 |
| g1-standard-2-4 | 2 | 4 |
| g1-standard-2-8 | 2 | 8 |
| g1-standard-4-8 | 4 | 8 |
| g1-standard-4-16 | 4 | 16 |
| g1-standard-8-16 | 8 | 16 |
| g1-standard-8-32 | 8 | 32 |
| g1-standard-16-32 | 16 | 32 |
| g1-standard-16-64 | 16 | 64 |
| g1-standard-32-64 | 32 | 64 |
#### Windows-based VMs
Windows license is included.
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------------ | -------- | ------------------- |
| g1w-standard-1-2 | 1 | 2 |
| g1w-standard-2-4 | 2 | 4 |
| g1w-standard-2-8 | 2 | 8 |
| g1w-standard-4-8 | 4 | 8 |
| g1w-standard-4-16 | 4 | 16 |
| g1w-standard-8-16 | 8 | 16 |
| g1w-standard-8-32 | 8 | 32 |
| g1w-standard-16-32 | 16 | 32 |
| g1w-standard-16-64 | 16 | 64 |
| g1w-standard-32-64 | 32 | 64 |
Machines powered with the **3rd Gen Intel® Xeon® Scalable Processors**. Designed for scalability, performance, and versatility across data center workloads — from entry-level to high-performance computing, virtualization, cloud, and enterprise applications.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------------ | -------- | ------------------- |
| g2-standard-1-2 | 1 | 2 |
| g2-standard-2-4 | 2 | 4 |
| g2-standard-2-8 | 2 | 8 |
| g2-standard-4-8 | 4 | 8 |
| g2-standard-4-16 | 4 | 16 |
| g2-standard-8-16 | 8 | 16 |
| g2-standard-8-32 | 8 | 32 |
| g2-standard-16-32 | 16 | 32 |
| g2-standard-16-64 | 16 | 64 |
| g2-standard-32-64 | 32 | 64 |
| g2-standard-32-128 | 32 | 128 |
#### Windows-based VMs
Windows license is included.
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------------ | -------- | ------------------- |
| g2w-standard-1-2 | 1 | 2 |
| g2w-standard-2-4 | 2 | 4 |
| g2w-standard-2-8 | 2 | 8 |
| g2w-standard-4-8 | 4 | 8 |
| g2w-standard-4-16 | 4 | 16 |
| g2w-standard-8-16 | 8 | 16 |
| g2w-standard-8-32 | 8 | 32 |
| g2w-standard-16-32 | 16 | 32 |
| g2w-standard-16-64 | 16 | 64 |
| g2w-standard-32-64 | 32 | 64 |
Machines powered with the **Intel Xeon E3-E5 Family** processors. Designed for a wide range of workloads and reliable performance.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ---------------- | -------- | ------------------- |
| g0-standard-1-2 | 1 | 2 |
| g0-standard-2-4 | 2 | 4 |
| g0-standard-4-8 | 4 | 8 |
| g0-standard-8-16 | 8 | 16 |
## GPU
Production-grade Virtual Machines designed for AI and compute-intensive workloads.
**Use cases**
Artificial Intelligence and Machine Learning, Computer-Aided Design (CAD), rendering.
Machines powered with the **2nd Generation Intel Xeon Scalable Processors** and **Nvidia Graphic Cards**. Nvidia GPUs enable parallel processing of multiple matrix operations and calculations. As a result, GPUs complete AI training tasks much faster than traditional CPUs.
Well-suited for compute-intensive workloads, including dynamic programming algorithms, video rendering, and scientific simulations.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| --------------- | -------- | ------------------- |
| g1-gpu-4-16-1 | 4 | 16 |
| g1-gpu-8-32-1 | 8 | 32 |
| g1-gpu-8-32-2 | 8 | 32 |
| g1-gpu-16-64-4 | 16 | 64 |
| g1-gpu-32-128-4 | 32 | 128 |
#### Windows-based VMs
Windows license is included.
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ---------------- | -------- | ------------------- |
| g1w-gpu-4-16-1 | 4 | 16 |
| g1w-gpu-8-32-1 | 8 | 32 |
| g1w-gpu-8-32-2 | 8 | 32 |
| g1w-gpu-16-64-4 | 16 | 64 |
| g1w-gpu-32-128-4 | 32 | 128 |
Machines powered with the **3rd Gen Intel® Xeon® Scalable Processors**. Designed for scalability, performance, and versatility across data center workloads — from entry-level to high-performance computing, virtualization, cloud, and enterprise applications.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------------ | -------- | ------------------- |
| g2-gpu-4-16-a2-1 | 4 | 16 |
| g2-gpu-8-32-a2-1 | 8 | 32 |
| g2-gpu-8-32-a2-2 | 8 | 32 |
| g2-gpu-16-64-a2-4 | 16 | 64 |
| g2-gpu-32-128-a2-4 | 32 | 128 |
Machines equipped with the **Intel® Xeon® E Family (E-2288G)** processor.
Targeted at entry-level and medium applications, these VMs offer a balance of performance, cost-effectiveness, and power efficiency. Suitable for operations that don't require high computing resources or memory-extensive tasks.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ----------------- | -------- | ------------------- |
| g1-gpuhf-4-16-1 | 4 | 16 |
| g1-gpuhf-8-32-1 | 8 | 32 |
| g1-gpuhf-8-32-2 | 8 | 32 |
| g1-gpuhf-16-64-4 | 16 | 64 |
| g1-gpuhf-32-128-4 | 32 | 128 |
## CPU-Optimized
Production-grade machines designed for high workloads with intensive CPU usage.
**Use cases**
Big data packet processing, video coding, huge builds.
Machines powered with the **2nd Gen Intel® Xeon® Scalable Processors**. Optimized for demanding workloads like high-performance computing, advanced analytics, and artificial intelligence.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------ | -------- | ------------------- |
| g1-cpu-2-2 | 2 | 2 |
| g1-cpu-4-4 | 4 | 4 |
| g1-cpu-8-8 | 8 | 8 |
| g1-cpu-16-16 | 16 | 16 |
| g1-cpu-32-32 | 32 | 32 |
#### Windows-based VMs
Windows license is included.
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------- | -------- | ------------------- |
| g1w-cpu-2-2 | 2 | 2 |
| g1w-cpu-4-4 | 4 | 4 |
| g1w-cpu-8-8 | 8 | 8 |
| g1w-cpu-16-16 | 16 | 16 |
| g1w-cpu-32-32 | 32 | 32 |
Machines powered with the **3rd Gen Intel® Xeon® Scalable Processors**. Designed for scalability, performance, and versatility across data center workloads — from entry-level to high-performance computing, virtualization, cloud, and enterprise applications.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------ | -------- | ------------------- |
| g2-cpu-2-2 | 2 | 2 |
| g2-cpu-4-4 | 4 | 4 |
| g2-cpu-8-8 | 8 | 8 |
| g2-cpu-16-16 | 16 | 16 |
| g2-cpu-32-32 | 32 | 32 |
Machines powered with the cloud-native **128-core ARM Ampere Altra Max Family** processor. Optimized for cloud workloads, these VMs deliver predictable high performance to meet the demands of modern data centers and scalable cloud applications.
**Use cases**
Applications that take advantage of scale-out and elastic cloud architectures.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------ | -------- | ------------------- |
| a1-cpu-2-2 | 2 | 2 |
| a1-cpu-4-4 | 4 | 4 |
| a1-cpu-8-8 | 8 | 8 |
| a1-cpu-16-16 | 16 | 16 |
| a1-cpu-32-32 | 32 | 32 |
#### Windows-based VMs
Windows license is included.
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------- | -------- | ------------------- |
| a1w-cpu-2-2 | 2 | 2 |
| a1w-cpu-4-4 | 4 | 4 |
| a1w-cpu-8-8 | 8 | 8 |
| a1w-cpu-16-16 | 16 | 16 |
| a1w-cpu-32-32 | 32 | 32 |
## Memory-Optimized
Production-grade Virtual Machines designed for workloads that rely heavily on RAM for optimal performance.
**Use cases**
Databases, real-time analytics, and memory-intensive AI and Machine Learning tasks.
Machines powered with the **2nd Gen Intel® Xeon® Scalable Processors**. Optimized for demanding workloads, such as high-performance computing, advanced analytics, and artificial intelligence.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ---------------- | -------- | ------------------- |
| g1-memory-2-16 | 2 | 16 |
| g1-memory-4-32 | 4 | 32 |
| g1-memory-8-64 | 8 | 64 |
| g1-memory-16-128 | 16 | 128 |
| g1-memory-32-256 | 32 | 256 |
#### Windows-based VMs
Windows license is included.
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ----------------- | -------- | ------------------- |
| g1w-memory-2-16 | 2 | 16 |
| g1w-memory-4-32 | 4 | 32 |
| g1w-memory-8-64 | 8 | 64 |
| g1w-memory-16-128 | 16 | 128 |
| g1w-memory-32-256 | 32 | 256 |
Machines powered with the **3rd Gen Intel® Xeon® Scalable Processors**. Designed for scalability, performance, and versatility across data center workloads — from entry-level to high-performance computing, virtualization, cloud, and enterprise applications.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ---------------- | -------- | ------------------- |
| g2-memory-2-16 | 2 | 16 |
| g2-memory-4-32 | 4 | 32 |
| g2-memory-8-64 | 8 | 64 |
| g2-memory-16-128 | 16 | 128 |
| g2-memory-32-256 | 32 | 256 |
#### Windows-based VMs
Windows license is included.
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ----------------- | -------- | ------------------- |
| g2w-memory-2-16 | 2 | 16 |
| g2w-memory-4-32 | 4 | 32 |
| g2w-memory-8-64 | 8 | 64 |
| g2w-memory-16-128 | 16 | 128 |
| g2w-memory-32-256 | 32 | 256 |
Machines powered with the cloud-native **128-core ARM Ampere Altra Max Family** processor. Optimized for cloud workloads, these VMs deliver predictable high performance to meet the demands of modern data centers and scalable cloud applications.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ---------------- | -------- | ------------------- |
| a1-memory-2-16 | 2 | 16 |
| a1-memory-4-32 | 4 | 32 |
| a1-memory-8-64 | 8 | 64 |
| a1-memory-16-128 | 16 | 128 |
| a1-memory-32-256 | 32 | 256 |
## High-frequency
These VMs have a high processor clock speed, starting at 3.7 GHz in the base configuration. Suitable for applications that require single-threaded performance, using a single CPU core for optimized performance and speed.
Configurations with a large amount of memory are also well-suited for working with large databases.
**Use cases**
Financial and probabilistic analytics, automation of computational processes.
Machines equipped with the **2nd Gen Intel® Xeon® Scalable Processors**. Optimized for demanding workloads such as high-performance computing, advanced analytics, and artificial intelligence.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ---------------- | -------- | ------------------- |
| g1-highfreq-1-2 | 1 | 2 |
| g1-highfreq-2-4 | 2 | 4 |
| g1-highfreq-4-8 | 4 | 8 |
| g1-highfreq-8-16 | 8 | 16 |
#### Windows-based VMs
Windows license is included.
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ----------------- | -------- | ------------------- |
| g1w-highfreq-1-2 | 1 | 2 |
| g1w-highfreq-2-4 | 2 | 4 |
| g1w-highfreq-4-8 | 4 | 8 |
| g1w-highfreq-8-16 | 8 | 16 |
## Basic VM
Virtual Machine with shared CPU usage. Cost-effective solution featuring multiple Intel CPU models, ideal for light web servers and microservices. Suitable for operations that don't require high computing resources or memory-extensive tasks.
**Use cases**
File sharing or storage, backup services, web hosting of small websites, or private VPN deployment.
Powered with multiple Intel CPU models.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ---------------- | -------- | ------------------- |
| g1s-shared-1-0.5 | 1 | 0.5 |
| g1s-shared-1-1 | 1 | 1 |
| g1s-shared-1-2 | 1 | 2 |
| g1s-shared-2-2 | 2 | 2 |
| g1s-shared-2-4 | 2 | 4 |
Powered with multiple Intel CPU models.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------------- | -------- | ------------------- |
| g2s-shared-1-2-25 | 1 | 2 |
| g2s-shared-1-2-50 | 1 | 2 |
| g2s-shared-2-4-50 | 2 | 4 |
| g2s-shared-2-4-80 | 2 | 4 |
| g2s-shared-4-8-80 | 4 | 8 |
| g2s-shared-4-8-160 | 4 | 8 |
| g2s-shared-8-16-160 | 8 | 16 |
| g2s-shared-8-16-320 | 8 | 16 |
## TPM
Virtual Machines with a **Trusted Platform Module** (TPM) are suitable for a diverse range of workloads that demand robust hardware-based security features.
**Use cases**
* **BitLocker Drive Encryption**: Secure entire disks with TPM-protected encryption keys, preventing unauthorized access even if the system is compromised.
* **Secure boot**: The system boots using only software trusted by the manufacturer. TPM provides a secure environment to store cryptographic keys that verify the integrity of the boot process, detecting and preventing unauthorized changes to the boot loader or firmware.
* **Virtual smart cards**: Provide secure authentication without the need for physical hardware.
* **Credential Guard**: Isolate and protect user credentials using virtualization-based security (VBS).
* **Digital Rights Management (DRM)**: Protect digital content from unauthorized access and distribution with TPM-stored encryption keys.
Machines powered with the **3rd Gen Intel® Xeon® Scalable Processors**. Designed for scalability, performance, and versatility across data center workloads — from entry-level to high-performance computing, virtualization, cloud, and enterprise applications.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** |
| ------------- | -------- | ------------------- |
| g2-tpm-1-2 | 1 | 2 |
| g2-tpm-2-4 | 2 | 4 |
| g2-tpm-2-8 | 2 | 8 |
| g2-tpm-4-8 | 4 | 8 |
| g2-tpm-4-16 | 4 | 16 |
| g2-tpm-8-16 | 8 | 16 |
| g2-tpm-8-32 | 8 | 32 |
| g2-tpm-16-32 | 16 | 32 |
| g2-tpm-16-64 | 16 | 64 |
| g2-tpm-32-64 | 32 | 64 |
| g2-tpm-32-128 | 32 | 128 |
## Network optimised
Virtual Machines designed for workloads that require high network throughput and low latency. Network optimised VMs provide up to 5 Gbps for smaller configurations and up to 10 Gbps for larger configurations, compared to up to 1 Gbps on standard flavors.
These VMs run on dedicated hypervisor nodes with hardware-level SR-IOV network offloading, delivering near bare-metal networking performance while maintaining full security feature support including security groups and anti-spoofing protection.
All network optimised flavors have multi-queue virtio networking enabled, which distributes packet processing across multiple vCPUs. This significantly increases throughput on larger flavors.
GPU and AI VM flavors (`gX-ai`) also run on the same network-optimised infrastructure and benefit from the same networking capabilities.
**QoS tiers**
| **Flavor size** | **Bandwidth** | **Burst** |
| ------------------------------ | ------------- | --------- |
| Up to 8 vCPU / 32 GB RAM | 5 Gbps | +1 Gbps |
| 16 vCPU / 32 GB RAM and larger | 10 Gbps | +2 Gbps |
**Use cases**
* **High-throughput load balancers**: Services handling large volumes of concurrent connections or high-bandwidth traffic, such as media streaming, CDN origin, or API gateways.
* **Network appliances and virtual routers**: Workloads that need to route, inspect, or transform large amounts of traffic without CPU becoming a bottleneck.
* **Real-time data pipelines and streaming analytics**: Applications requiring consistent low-latency packet delivery, such as financial data processing or log ingestion at scale.
* **Game servers and multiplayer backends**: Latency-sensitive applications where packet rate and jitter matter as much as raw bandwidth.
* **VPN concentrators and security gateways**: Services terminating many VPN tunnels or performing deep packet inspection at line rate.
**Limitations**
* VxLAN-encapsulated networks receive partial hardware offload. VLAN-backed networks achieve close to 100% hardware offload.
Machines powered with **AMD EPYC™ 4th Gen** processors. Flavors up to 8 vCPU / 32 GB RAM use the 5 Gbps tier; larger flavors use 10 Gbps.
#### Linux-based VMs
| **VM type** | **vCPU** | **RAM/memory (GB)** | **Network bandwidth** |
| -------------- | -------- | ------------------- | --------------------- |
| g3a-net-4-8 | 4 | 8 | 5 Gbps |
| g3a-net-4-16 | 4 | 16 | 5 Gbps |
| g3a-net-8-16 | 8 | 16 | 5 Gbps |
| g3a-net-8-32 | 8 | 32 | 5 Gbps |
| g3a-net-16-32 | 16 | 32 | 10 Gbps |
| g3a-net-16-64 | 16 | 64 | 10 Gbps |
| g3a-net-32-64 | 32 | 64 | 10 Gbps |
| g3a-net-32-128 | 32 | 128 | 10 Gbps |
| g3a-net-64-128 | 64 | 128 | 10 Gbps |
| g3a-net-64-256 | 64 | 256 | 10 Gbps |
| g3a-net-96-128 | 96 | 128 | 10 Gbps |
| g3a-net-96-256 | 96 | 256 | 10 Gbps |
| g3a-net-96-490 | 96 | 490 | 10 Gbps |