Cloud servers with dedicated cores
A cloud server with dedicated cores is a cloud server with vCPUs that are tied to the threads of specific physical processor cores rather than switching between threads of different cores.
Dedicated cores guarantee that the cloud server uses processor resources without competing for computing power with other cloud servers. This isolation:
- provides the cloud server with stable performance;
- protects the cloud server from hardware vulnerabilities (such as Spectre, Meltdown, and others) caused by other cloud servers on the same physical node.
Dedicated core configuration is available when creating a cloud server or when changing the cloud server configuration.
For a cloud server with dedicated cores, the Hyper-Threading (SMT) option is enabled by default. You can also configure mandatory resource placement on a single NUMA node.
Use cases
Suitable for high-load and performance-critical systems that require predictable and stable vCPU performance without the influence of other cloud servers:
- for running resource-intensive databases;
- for processing analytical tasks and tasks with high computational complexity;
- for running CI/CD servers;
- for hosting game servers and real-time applications requiring low latency;
- for deploying microservices and high-request-rate APIs;
- for ensuring stable monitoring and logging performance during high event rates.
Hyper-Threading
Hyper-Threading (SMT or hyper-threading) is a technology that allows a single physical processor core to process two threads simultaneously. In this case, each pair of vCPUs of a cloud server corresponds to two threads of one physical processor core.
This technology improves performance for multi-threaded tasks, such as running web servers and enterprise applications, or rendering and video processing. Learn more about the technology and its operating principles in the Selectel blog article What is Hyper-Threading.
On cloud servers with dedicated cores, Hyper-Threading (SMT) is enabled by default. You cannot enable Hyper-Threading (SMT) for cloud servers without dedicated cores.
For databases, high-performance computing (HPC), or software with per-physical-core licensing, we recommend testing system performance with the Hyper-Threading (SMT) option enabled and disabled. If necessary, the option can be disabled:
- for stable performance in single-threaded scenarios;
- for task isolation at the physical core level;
- for license compliance.
When disabling the Hyper-Threading (SMT) option, threads are disabled — each cloud server vCPU corresponds to an entire physical processor core. The number of vCPUs on the cloud server is halved.
For example, you create a cloud server with the Hyper-Threading (SMT) option enabled and 8 vCPUs. The cloud server will use 4 physical cores, each of which is operating in Hyper-Threading (SMT) mode, meaning it processes 2 threads. Thus, the cloud server has 8 vCPUs.
If you disable the Hyper-Threading (SMT) option, the number of vCPUs on the cloud server will decrease to 4 and will correspond to the number of physical cores rather than the number of threads. The number of occupied physical cores will not change, so the cost of the cloud server will remain the same.
Before disabling Hyper-Threading (SMT), be sure to test application performance under load — some applications may run slower without Hyper-Threading (SMT).
Mandatory resource placement on a single NUMA node
NUMA (Non-Uniform Memory Access) is an architecture for multi-processor systems. NUMA nodes are architectural blocks in multi-processor systems where a node includes one processor and local RAM, which the processor within the node can access quickly and with low latency.
When placing resources on a single NUMA node, the cloud server operates only within that node—it uses the processor cores and local memory tied to that node. This helps minimize memory access latency.
You can enable mandatory cloud server resource placement on a single NUMA node when creating a cloud server or changing the cloud server configuration only with dedicated cores and 4 or more vCPUs. You cannot enable mandatory resource placement on a single NUMA node for cloud servers without dedicated cores.
If it is not possible to place cloud server resources on a single NUMA node, the server will not be created.
Limitations
Cloud servers with dedicated cores are available only in pool segments ru-3b, ru-7a, ru-7b, and all segments of the multi-AZ pool ru-6.
Dedicated core configuration is available only in the Standard and HighFreq lines. A list of fixed-configuration flavors in these lines that support dedicated cores can be viewed in the table List of fixed-configuration flavors in all pools.
Cost
The cost of a cloud server with dedicated cores, the Hyper-Threading (SMT) option, and the mandatory placement on a single NUMA node option can be viewed at selectel.ru.
The cost of a cloud server with dedicated cores does not change:
-
when the Hyper-Threading (SMT) option is enabled or disabled;
-
when the mandatory placement on a single NUMA node option is enabled or disabled.