Virtual Server Plans
PacketStream Virtual Server offers five instance families tailored to different workloads. All prices are monthly and exclude VAT; the listed traffic is the base transfer allowance included each month.
Instance Naming
Instance names follow the family.size format. For example, g2a.large is the large size of the General Purpose family. Sizes scale from micro up to 5xlarge, with vCPU, memory, disk, and traffic increasing together as the size grows.
| Family | Prefix | Processor | Best for |
|---|---|---|---|
| Compact | c1i | Intel Xeon E5-v4 | Dev/test, small websites |
| General Purpose | g2a | AMD EPYC 3rd Gen | Web apps, APIs, small–mid databases |
| High-Frequency | hf1a | AMD 5000/9000 Series | Game servers, real-time, high-clock work |
| Compute Optimized | co2a | AMD EPYC 3rd Gen | Batch processing, heavy compute |
| Memory Optimized | mo2a | AMD EPYC 3rd Gen | In-memory DBs, caching, analytics |
Compact (c1i)
Entry-level instances that prioritize cost efficiency. A good fit for dev/test environments, low-traffic websites, and personal projects.
- Processor: Intel Xeon E5-v4
- Storage: SSD
| Instance | vCPU | Memory | Storage | Monthly Traffic | Monthly (KRW) | Monthly (USD) |
|---|---|---|---|---|---|---|
| c1i.micro | 1 | 1 GB | 50 GB SSD | 0.9 TB | ₩5,000 | USD 3.50 |
| c1i.small | 1 | 2 GB | 50 GB SSD | 0.9 TB | ₩7,500 | USD 5.00 |
General Purpose (g2a)
Balanced CPU, memory, and network — a solid fit for most production workloads. If you are unsure which family to choose, start with General Purpose.
- Processor: AMD EPYC 3rd Gen
- Storage: NVMe
| Instance | vCPU | Memory | Storage | Monthly Traffic | Monthly (KRW) | Monthly (USD) |
|---|---|---|---|---|---|---|
| g2a.micro | 2 | 4 GB | 100 GB NVMe | 0.9 TB | ₩16,000 | USD 11.20 |
| g2a.small | 4 | 8 GB | 100 GB NVMe | 1.8 TB | ₩32,000 | USD 22.40 |
| g2a.medium | 4 | 16 GB | 160 GB NVMe | 2.7 TB | ₩64,000 | USD 44.80 |
| g2a.large | 6 | 24 GB | 240 GB NVMe | 3.6 TB | ₩96,000 | USD 67.20 |
| g2a.xlarge | 8 | 32 GB | 320 GB NVMe | 4.5 TB | ₩128,000 | USD 89.60 |
| g2a.2xlarge | 12 | 48 GB | 480 GB NVMe | 5.4 TB | ₩192,000 | USD 134.40 |
| g2a.3xlarge | 12 | 64 GB | 640 GB NVMe | 6.3 TB | ₩256,000 | USD 179.20 |
| g2a.4xlarge | 16 | 96 GB | 960 GB NVMe | 7.2 TB | ₩384,000 | USD 268.80 |
| g2a.5xlarge | 24 | 128 GB | 1,280 GB NVMe | 8.1 TB | ₩512,000 | USD 358.40 |
High-Frequency (hf1a)
High-clock AMD processors deliver strong single-thread performance for latency-sensitive work. Ideal for game servers, real-time transactions, and other workloads where responsiveness matters — with generous traffic allowances.
- Processor: AMD 5000/9000 Series
- Storage: NVMe
| Instance | vCPU | Memory | Storage | Monthly Traffic | Monthly (KRW) | Monthly (USD) |
|---|---|---|---|---|---|---|
| hf1a.micro | 1 | 2 GB | 50 GB NVMe | 1.8 TB | ₩12,000 | USD 8.40 |
| hf1a.small | 2 | 4 GB | 100 GB NVMe | 3.6 TB | ₩24,000 | USD 16.80 |
| hf1a.medium | 4 | 8 GB | 100 GB NVMe | 5.4 TB | ₩48,000 | USD 33.60 |
| hf1a.large | 4 | 16 GB | 160 GB NVMe | 7.2 TB | ₩96,000 | USD 67.20 |
| hf1a.xlarge | 6 | 24 GB | 240 GB NVMe | 9.0 TB | ₩144,000 | USD 100.80 |
| hf1a.2xlarge | 8 | 32 GB | 320 GB NVMe | 10.8 TB | ₩192,000 | USD 134.40 |
| hf1a.3xlarge | 12 | 48 GB | 480 GB NVMe | 12.6 TB | ₩288,000 | USD 201.60 |
| hf1a.4xlarge | 16 | 64 GB | 640 GB NVMe | 14.4 TB | ₩384,000 | USD 268.80 |
Compute Optimized (co2a)
A high vCPU-to-memory ratio for compute-bound workloads such as batch processing, large builds, media encoding, and scientific computing.
- Processor: AMD EPYC 3rd Gen
- Storage: NVMe
| Instance | vCPU | Memory | Storage | Monthly Traffic | Monthly (KRW) | Monthly (USD) |
|---|---|---|---|---|---|---|
| co2a.micro | 4 | 4 GB | 100 GB NVMe | 0.9 TB | ₩28,000 | USD 19.60 |
| co2a.small | 8 | 16 GB | 160 GB NVMe | 1.8 TB | ₩112,000 | USD 78.40 |
| co2a.medium | 16 | 32 GB | 320 GB NVMe | 2.7 TB | ₩224,000 | USD 156.80 |
| co2a.large | 16 | 64 GB | 640 GB NVMe | 3.6 TB | ₩448,000 | USD 313.60 |
| co2a.xlarge | 32 | 64 GB | 640 GB NVMe | 4.5 TB | ₩448,000 | USD 313.60 |
| co2a.2xlarge | 32 | 128 GB | 1,280 GB NVMe | 5.4 TB | ₩896,000 | USD 627.20 |
Memory Optimized (mo2a)
A high memory-to-vCPU ratio for memory-hungry workloads such as in-memory databases, caching layers, and large-scale data analytics.
- Processor: AMD EPYC 3rd Gen
- Storage: NVMe
| Instance | vCPU | Memory | Storage | Monthly Traffic | Monthly (KRW) | Monthly (USD) |
|---|---|---|---|---|---|---|
| mo2a.micro | 4 | 16 GB | 100 GB NVMe | 1.8 TB | ₩56,000 | USD 39.20 |
| mo2a.small | 4 | 32 GB | 160 GB NVMe | 2.7 TB | ₩112,000 | USD 78.40 |
| mo2a.medium | 6 | 48 GB | 240 GB NVMe | 3.6 TB | ₩168,000 | USD 117.60 |
| mo2a.large | 8 | 64 GB | 320 GB NVMe | 4.5 TB | ₩224,000 | USD 156.80 |
| mo2a.xlarge | 12 | 96 GB | 480 GB NVMe | 5.4 TB | ₩336,000 | USD 235.20 |
| mo2a.2xlarge | 12 | 128 GB | 640 GB NVMe | 6.3 TB | ₩448,000 | USD 313.60 |
| mo2a.3xlarge | 16 | 192 GB | 960 GB NVMe | 7.2 TB | ₩672,000 | USD 470.40 |
| mo2a.4xlarge | 24 | 256 GB | 1,280 GB NVMe | 8.1 TB | ₩896,000 | USD 627.20 |
| mo2a.5xlarge | 32 | 384 GB | 1,280 GB NVMe | 9.0 TB | ₩1,344,000 | USD 940.80 |
Traffic Policy
The monthly traffic included with each instance is divided across 30 days and provided as a daily transfer allowance. For example, an instance with 9 TB per month can use roughly 300 GB per day.
- Exceeding the daily limit draws down additional traffic; once traffic is fully consumed, network speed is throttled for the rest of that day.
- You can purchase additional traffic from the console to extend the allowance.
Traffic Pricing
| Region | Routing Profile | Unit | Price (KRW) | Price (USD) |
|---|---|---|---|---|
| South Korea | General | 1 GB | ₩10 | USD 0.01 |
| South Korea | China Optimized | 1 GB | ₩45 | USD 0.045 |
| South Korea | China Direct | 1 GB | Contact sales | |
| Japan | General | 1 GB | ₩10 | USD 0.01 |
- China Optimized: Stable peering with Chinese carriers based on an eyeball network.
- China Direct: Direct interconnection with China Mobile, Unicom, and Telecom for low-latency delivery.
Storage Tiers
The default OS disk is provisioned on the Flash tier. For additional capacity, create a Block Storage volume and attach it to your virtual server.
| Tier | Backend | Characteristics | Best for |
|---|---|---|---|
| Archive | HDD | Lowest cost | Backups, rarely accessed data |
| Hybrid | HDD + SSD cache | Caches hot data on SSD | General-purpose workloads |
| Flash | SSD | General-purpose performance (default) | Most workloads |
| Express | NVMe | Highest I/O | Databases, high-I/O applications |
Supported Operating Systems
Linux
- Ubuntu 24.04 LTS, 22.04 LTS
- Debian 12 (Bookworm), 11 (Bullseye)
- Rocky Linux 9, 8
- AlmaLinux 9, 8
- Arch Linux (Rolling)
Windows Server
- Windows Server 2022 Standard
- Windows Server 2019 Standard
Windows Server licenses are provided under an SPLA agreement and incur additional cost.
Getting Started
You can create instances and manage billing from the console. For volume adoption or custom quotes, contact our sales team.
- Console: portal.packet.stream
- Sales: [email protected]