Cloud Servers

Last Updated on 2023-03-01



LayerStack’s Cloud Servers give you secure, fast, and virtually limitless data access for your needs – however unique they may be. Take advantage of the superlative flexibility and scalability our server options offer and be inspired by the endless possibilities that they bring.

Each server plan varies in the amount of allocated resources, and hence its ability to tackle specific kinds of workloads. This page offers an in-depth look at the difference between each server type, the distinction between a shared and a dedicated CPU, and a summary of our plans and their use cases, so you can pick a solution that works best for you.

Shared CPU vs. Dedicated CPU

Regardless of the plan you opt for, RAM, disk storage and some other resources are always dedicated – meaning you are the only user of them. With CPU, on the other hand, you have two options: shared CPU (offered by our General Purpose plans) and dedicated CPU (Memory-Optimized or Premium Compute plans).

As the name suggests, plans with dedicated CPU give you virtual CPU instances (vCPUs) that are dedicated only to you and no one else. Dedicated CPU minimizes surprise in performance as your workloads are handled by the full capacity of the processing cores.

In shared CPU plans, the CPU power is shared with others in the same node. This makes it a more affordable option. While your server’s CPU power is shared among your neighbours, it is possible to enjoy full processing capacity when your neighbours’ usage is low.

Plans and Use Cases Summary

From enterprise applications to personal projects, LayerStack has you covered.

PLAN CPU TYPE vCPUs MEMORY STORAGE PROCESSOR 12-MTH PRICE
General Purpose Shared 1 - 18 2 - 64GB 60 - 1280GB - AMD EPYC Milan US$10.04 – US$235
Memory Optimized Dedicated 2 - 32 24 - 256GB 200 - 1024GB NVMe SSD - 3rd Gen AMD EPYC US$53 – US$923
Premium Compute Dedicated 4 - 16 8 - 32GB 200 - 600GB NVMe SSD - 4th Gen AMD EPYC Genoa US$28 – US$148
ARM-Based Shared 2 - 10 4 - 48GB 50 - 400GB - Ampere® Altra® Max US$5 – US$32
High Storage Shared 2 - 8 4 - 20GB 500 - 1500GB - AMD EPYC US$14.95 – US$42.95


Each plan comes with its own strengths catering to your specific demands.

GENERAL PURPOSE MEMORY OPTIMIZED PREMIUM COMPUTE
Day-to-day computing at a lower cost Enhanced storage for memory-demanding workloads High performance CPU for intense data processing and multitasking
• Web apps
• Blogs
• Microservices
• Virtual desktops
• Small-medium databases
• Back-office operations
• SQL or NoSQL databases
• Large in-memory databases
• Real-time big data processing
• Caching solutions
• Image rendering
• AI
• Gaming
• Machine learning
• CI/CD
• Batch processing
• Video encoding

Shared CPU: General Purpose

An all-round solution for day-to-day computing, the General Purpose Cloud Server plans are designed with an abundance of configurable vCPUs (1 to 18) and memory (2GB to 64GB). The myriad number of memory-to-vCPU ratio options gives you the highest level of autonomy to pick a configuration most suited to your applications. This flexibility also allows smart upscaling for your future needs.

Our General Purpose Cloud Server plans offer configurable storage and a mix of Intel Xeon Scalable (2.0 GHz, 3.5 GHz turbo), 1st Gen AMD EPYC and 2nd Gen AMD EPYC (2.0 GHz, 3.35 GHz turbo) processors, making them a cost-effective solution that equips you to handle low to medium workloads and occasional traffic surges. They are also ideal for tasks that require less from the system’s computing prowess. Workload examples include:


• Web apps
• Blogs
• Microservices
• Virtual desktops
• Small-medium databases
• Back-office operations

vCPUs are shared in all General Purpose plans. This maximizes resource utilization and therefore minimizes your cost while delivering reliable performance for its value.

Dedicated CPU: Memory optimized

Memory-Optimized is one of our two dedicated CPU offerings where you can enjoy the benefit of zero computing power steal. All of our Hong Kong Memory-Optimized Cloud Server plans are armed with 3rd Gen AMD EPYC processors (2.0 GHz, 3.675 GHz turbo) while the Singapore Memory-Optimized Cloud Server plans feature 2nd Gen AMD EPYC processors (2.0 GHz, 3.35 GHz turbo). With boosted memory ranging from 24 GB to 256 GB, this option is extremely resilient to memory-demanding tasks, such as:


• SQL or NoSQL databases
• Large in-memory databases
• Real-time big data processing
• Caching solutions
• Image rendering

With up to 32 dedicated vCPUs and NVMe SSD storage that possesses significantly higher read/write speed (IOPS), this memory-laden solution provides you with sufficient RAM (random access memory) and resources to avoid latency caused by excessive disk swapping and out-of-memory errors. You can benefit from a higher return on investment brought by the highly stable performance and ultra-low latency.

Dedicated CPU: Premium Compute

Our Hong Kong Premium Compute Cloud Server plans to equip you with the latest 4th Gen AMD EPYC Genoa processor, which comes with a 3.7 GHz base clock speed and a maximum turbo frequency of 4.4 GHz and packed with high-speed DDR5 memory, PCIe 5.0 support.
These plans are most appropriate for workloads that rely heavily on computing power and multitasking but are less demanding in terms of memory. The superior read/write speed (IOPS) of NVMe SSD storage and up to 16 dedicated vCPUs are no slouch when it comes to artificial intelligence (AI), machine learning, gaming, and other applications that can benefit from the fast, coherent performance and low latency.


Premium Compute plans are a perfect solution if your applications fall into these categories:

• AI
• Gaming
• Machine learning
• CI/CD
• Batch processing
• Video encoding

GPU add-on is also available exclusively to our Premium Compute Cloud Servers plans. GPU is a powerful accelerator designed to boost overall performance even further for data-parallel computation, machine learning and graphics-intensive projects.

High SAS Storage

LayerStack’s latest high storage cloud server plan is designed to maximize standard CPU and memory allocations, offering customers large SAS hard disks with HDD cache to host any type of data with ease. The High SAS storage cloud server plans are all equipped with 3rd Gen AMD EPYC processors (2.0 GHz, 3.675 GHz turbo). With boosted memory from 4GB all the way to 20GB. It comes with 2, 4, or 8 vCPU respectively. There are 3 options for the storage size to choose from: 500GB, 1TB and 1.5TB. All the plans are also safeguarded with Raid 10 for data redundancy, which is valuable when used for backup or data security. This plan option is well suited to cases such as:

  • Webhosting with more complex web app structures
  • FTP server hosting
  • Multimedia cloud storage with big data and resources

Arm-based

Arm-based cloud virtual machines, the Arm which stands for Advanced RISC Machine, is engineered to lower energy consumption as much as possible and to provide up to 30% better price performance by running efficiently per core and at a lower cost per core. All plans are powered by Ampere® Altra® Max Cloud Native Processors M128-30, based on ARM Neoverse N1 enterprise-level core architecture that delivers excellent cost-effective processing power with incredible speed and sustainable performance to maximize operational efficiency. It is designed specifically for running compute-intensive applications and cloud-native applications that have high memory requirements, such as:

  • Web, app, gaming servers
  • CI/CD development
  • Image/Video encoding and processing
  • Large-scale Java applications
  • AI, Machine Learning software development

We offer various Arm-based cloud virtual machines option plans for users to choose from, with ultra-fast NVMe SSD memory ranging from 8GB to 48GB, and also offer China Direct CN2 GIA route add-on for HK$50/US$6.5/SG$9.

Upgrade a Plan

Your needs may change over time, and we are here to make upscaling as seamless and easy as it can be. At any time of a service period, you can initiate an upgrade in the LayerPanel by choosing your desired plan. The upgrade is instant once the payment is confirmed, or you have an option to schedule the upgrade for a later date. The upgraded service in the rest of the billing cycle is charged on a pro-rata basis.

Shared bandwidth vs. Dedicated bandwidth

Bandwidth is the volume of data that the network can handle. The more the bandwidth, the more data can be transferred over a certain period of time, both to and from LayerStack’s data centers around the world.

With LayerStack, you have two bandwidth options – shared bandwidth and dedicated bandwidth.

Very similar to the distinction between shared and dedicated CPU, shared bandwidth is where you split the bandwidth – and the cost – with other users, making it a more economic choice.

Just like when you pick a shared CPU plan, you may still enjoy the benefit of full bandwidth when your neighbors are inactive.

Plans with dedicated bandwidth offer bandwidth that is reserved exclusively for your use with guaranteed network speed and stability for a small additional cost.

Unlimited Data Transfer and the Fair Use Policy

All LayerStack cloud packages offer unlimited data transfer under the “Fair Use Policy”. When your monthly data usage reaches the data capacity stated in each service plan, the cloud service continues but at a reduced bandwidth, hence a slower connection speed. This is to avoid service downtime and unwelcome surprises on your bills while maintaining a satisfactory level of network stability for all users. You can keep track of your data usage in LayerPanel. Remember, you always have the option of moving to a plan with a larger data capacity at any time. Data transfer between cloud servers over the private VLAN is unlimited and not charged.

Regional Availability

Related Tutorials