Free tools Windows power users keep installed
One-click scans. No signup required.
iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Choose a VDS by measuring what your workload actually uses at busy times, then validate a candidate configuration under realistic load. There is no reliable universal rule such as “one vCPU per site”: CPU allocation, concurrent demand, memory pressure, storage I/O, network conditions, and the provider’s resource definitions all affect capacity.
How do you choose a VDS size?
Work from evidence rather than a plan name. First establish what the application does, how demand changes over time, and what performance users need. Then measure the current system, identify its limiting resource, and test a candidate size that leaves room for peaks.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
TBZFNZXF FN063-VDS.6N.V7P7 Fan | $985.68 | Buy on Amazon |
- Describe the workload. List the application and services, typical and peak traffic or job patterns, concurrent users or requests, data size and growth, customer geography, and any GPU or low-latency needs. Separate steady demand from scheduled or event-driven spikes.
- Set success criteria. Define acceptable response time, job completion time, availability, and recovery expectations. Include deployments, backups, startup activity, batch work, and synchronized traffic or login bursts where relevant.
- Measure a representative busy period. Track CPU saturation and queueing, memory pressure and paging, disk capacity and I/O latency, network throughput and latency, and application-level response time. A single snapshot—or a raw “used memory” figure without operating-system context—can mislead.
- Find the bottleneck. CPU-heavy work needs suitable processing performance and concurrency; memory-bound applications need working-set capacity; databases and frequent disk operations need storage IOPS and latency as well as space; distributed workloads may be network-bound. Add GPU capacity only if the application benefits from it.
- Match findings to a provider’s actual definitions. Check whether a vCPU is a shared thread, dedicated thread, or physical core; whether RAM is reserved; what storage medium and I/O characteristics are specified; and what network limits and usage policies apply. Confirm whether each resource can be changed independently.
- Pilot and validate. Test the candidate configuration with representative traffic or a load simulation, increasing concurrency or workload volume gradually. Monitor both resource indicators and user-facing performance, including peaks, deployments, backups, or startup bursts.
- Plan the next move. Before committing, check upgrade and downgrade paths, likely downtime, migration method, IP or region changes, storage restrictions, and backup arrangements. Set a review trigger around sustained resource pressure or missed performance objectives, not one temporary spike.
Microsoft’s sizing guidance for Azure Virtual Desktop and Remote Desktop Services session hosts supports pilot or simulation testing and consideration of peak logon demand. Its session-host examples are initial estimates for that context, not general VDS benchmarks. Microsoft’s session-host sizing guidance states: “The examples in this article are generic guidelines, and you should only use them for initial performance estimates.”
How many vCPUs does your VDS need?
Estimate how much work runs in parallel and benchmark the application. A CPU-bound task that is mostly single-threaded may benefit more from faster individual cores than from a larger vCPU count. Parallel jobs can benefit from additional cores, but only to the extent the software can use them and the provider actually allocates the capacity as expected.
#1 Best Overall
- Check provider terminology: “vCPU” can mean different things, including a hardware thread or a dedicated allocation. Confirm generation or performance class where documented.
- Watch for CPU saturation, queueing, and virtualization contention indicators such as steal or ready time when available.
- Do not assume doubling cores doubles capacity. Microsoft’s session-host guidance describes diminishing returns and synchronization overhead in its own context; that principle cautions against linear assumptions but does not establish a universal VDS ratio.
For scale, Microsoft reports a typical CPU scaling factor of 1.5–1.9 in its session-host capacity context. It also reports 15–20% capacity cost and 10–20% higher response times in virtualization overhead and latency comparisons based on internal testing for that guidance. The page does not state a publication year for these figures, and none should be treated as a general hosted-server estimate. Read the scope and methodology context in Microsoft’s guidance.
How much RAM should you allocate?
Estimate the concurrent working sets of the operating system, application processes, caches, and database. Observe memory pressure and paging during busy periods, and allow for bursts and maintenance tasks. Add RAM when measurements show memory is constraining performance; more memory will not resolve a CPU, storage, or network bottleneck.
How much disk space and I/O do you need?
Plan storage capacity separately from storage performance and data protection. Add the operating system, applications, databases, logs, uploads, temporary files, backups, and expected data growth to the capacity estimate. Then assess read/write patterns, latency, IOPS, and throughput. An SSD or NVMe label alone does not promise a particular service level.
- Find out whether storage is local or networked, persistent or ephemeral, and included in plan changes.
- Check whether disk size can be increased independently and whether doing so requires a migration or reinstall.
- Verify backups and recovery separately; a disk allocation is not itself a backup plan.
As a provider-specific example, Selectel’s VDS documentation, updated September 18, 2026, describes fixed CPU/RAM/local-disk configurations using local NVMe in RAID 10. Disk size is not independently configurable within that product’s configuration. Selectel publishes local-volume figures of 12,800 read IOPS, 6,400 write IOPS, and 300 MB/s throughput; these are vendor specifications for that service, not independent benchmarks or values to expect from another VDS. See Selectel’s VDS configurations and storage details.
What network limits should you check?
Estimate inbound and outbound transfer, burst rates, concurrent connections, and traffic patterns. Compare the port rate with included traffic or fair-use terms, public and private network options, region, latency to users and dependencies, and DDoS protections. A high advertised port rate does not establish low latency or guaranteed sustained end-to-end throughput.
For interactive or distributed workloads, consider latency, packet loss, and jitter alongside bandwidth. Microsoft identifies those as user-experience factors in its session-host scenarios; the relevant thresholds for your application depend on its own needs. Microsoft’s session-host guidance discusses those network factors.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Do you need GPU or specialized hardware?
Choose a GPU only when the application benefits from graphics acceleration or GPU compute and the provider supports the necessary device and driver stack. Rendering, 3D, and simulation are examples Microsoft gives for GPU needs in session-host workloads, but the actual application’s requirements should determine the choice. Microsoft’s guidance lists GPU alongside CPU, memory, network, and storage as a session-host capacity factor.
How do VDS provider plans differ?
Plan labels are not standardized sizing units. The following published examples show how packaging and allocation semantics differ; they are provider-specific snapshots, not recommendations for a particular workload.
| Provider and product | Published configuration range or examples | Resource and change details | Network and storage details |
|---|---|---|---|
| Selectel VDS | Examples include 1 vCPU / 1 GB RAM / 10 GB disk; 2 vCPU / 4 GB / 50 GB; 4 vCPU / 8 GB / 80 GB; and 6 vCPU / 12 GB / 120 GB. | Configurations are fixed rather than independently customizable; configuration can be changed after creation. | Documentation lists local NVMe SSD in RAID 10 and bandwidth up to 3 Gbps. The page publishes local-volume performance of 12,800 read IOPS, 6,400 write IOPS, and 300 MB/s throughput. |
| Contabo Cloud VDS S–XXL | 6–24 vCores, 24–96 GB RAM, and 180–720 GB NVMe. | Contabo counts vCores with AMD SMT enabled at two vCores per physical core, and says CPU cores and RAM are exclusively reserved while the host remains multi-tenant. Storage changes require reinstall and erase data; snapshots are unavailable for this line. | Ports range from 250–1,000 Mbit/s. |
Selectel’s figures come from documentation updated September 18, 2026. Contabo’s range reflects product documentation checked October 7, 2026; provider plans can change, so confirm current specifications and terms before purchase. Selectel VDS documentation · Contabo Max Performance VPS documentation · Contabo server portfolio guide
When should you scale up?
Scale when measurements show sustained pressure or the workload repeatedly misses its response-time, completion-time, or availability target. A temporary peak may call for workload scheduling or burst capacity rather than a permanent larger configuration; diagnose the limiting resource before changing plans.
Re-test after resizing. Confirm that the change improved the user-facing objective and did not simply move the bottleneck—for example, from CPU to disk I/O or from memory pressure to network limits. Keep monitoring as concurrency and data grow, and revisit the size when workload patterns or provider definitions change.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
