It is pleased to announce the new Dedicated GPU option specially designed for users with intensive and special computing needs. Users can harness the full power of dedicated GPU to achieve ultra-high performance computing cloud servers specialized in Data Parallel Computation, Machine Learning and Graphically-Intensive Projects.
Machine Learning
GPU is widely used in deep learning because of the strong capability of parallel computing. Compared to CPU, GPU has much more cores to learn representations automatically from data such as images, video or text. Thus, it helps to reduce machine learning training time from days to minutes, enabling researchers and developers to obtain faster, more accurate information from deep learning.
Graphically Demanding Applications
Besides, GPU is effective in video and image rendering. The long rendering time is always a major issue to creative professionals that tied up computing resources and stifled creative flow. Modern GPUs enable and accelerate the cloud servers to process video and graphic in higher-definition formats. Obviously, it is also powerful in computationally intensive gaming, providing gamers with smoother and higher-quality gaming experience.
Specification
LayerStack offers NVIDIA® Tesla T4 GPU for intensive users. T4 is a high rating GPU with 2560 CUDA cores and 320 Tensor cores, it has 16GB GDDR6 Memory and 40 RT core. Work together with best-in-class AMD EPYC2 CPU to deliver the best cloud performance for most demanding applications.
We very much value your feedback for future developments, please let us know your comments and suggestions on LayerStack Community.
To cope with diverse needs in cloud solutions and improve users’ cloud experience, LayerStack has made significant updates to the offers in Q4 2020. It newly launched two cloud servers plans, the High Frequency CPU Cloud Servers and Dedicated CPU Cloud Servers. It is no doubt that both of them are specially designed for CPU demanding workloads and applications, so what are the differences between the two new plans? This blog will identify the differences and provide some use cases for cloud users.
High Frequency CPU Cloud Servers
Launched in October 2020, it utilizes top-class 2nd Gen AMD EPYC with a turbo boost of 3.4 GHz which is the fastest architecture available on LayerPanel. It is specially designed for applications that rely heavily on computing speed and faster data processing. Moreover, dedicated GPU will be available soon for users who are seeking ultimate ML performance on High Frequency CPU Cloud Server.
CPU Comparison
According to the CPU comparison conducted recently, CPUs applied in LayerStack’s cloud servers are scoring 71362 and 67185 by PassMark which are way ahead of the processors used by other top cloud providers. Obviously, LayerStack’s High Frequency CPU Cloud Servers is utilizing top-level processors in the cloud industry. Users’ most demanding workloads can be handled rapidly with superior clock speeds, large bandwidth, and fast I/O transfers.
Dedicated CPU Cloud Server
Launched in Nov 2020, it is powered by another AMD EPYC2 which can be boosted up to 3.3GHz in turbo mode. The CPU cores are dedicated to specific servers that users can harness the full power of the CPU for their workloads. This servers plan is specially designed for long-running or latency-sensitive applications that require significant computing power to support.
Use Case
Although both new plans are CPU-related providing users with strong computing power, they are best suit in different cases. High Frequency CPU Cloud Servers is very effective in compute-intensive applications like machine learning, computer aided design (CAD) and video encoding. Dedicated CPU Cloud Server, on the other hand, is performing well in long-running projects, technical analysis, CI/CD and batch Processing.
CPU is the core power of cloud computing deciding the data-solving capability of the servers. Some CPU-intensive workload such as machine learning and financial calculations rely very much on the computing power. In fact, a CPU can be evaluated by several aspects, like cores/threads, clock speed, TDP, etc. According to the CPU comparison on PassMark, the CPU installed in LayerStack High Frequency plan outperform similar cloud servers of other top cloud providers. Reach the full article at Serchen Blog.
LayerStack is coming up with an exciting, one-of-a-kind High Frequency CPU Server Plan unmatched in the cloud-computing industry. Applications that depend upon higher computing power, faster network speeds, and lightning access to storage infrastructure will significantly benefit from the plan. The plan utilizes high-rating CPU – the AMD EPYC 7742 which boasts a superior performance with 64 cores and 128 threads. It comes with a high base clockspeed of 2.25 GHz, capable of a turbo boost up to 3.4 GHz. The 2nd generation EPYC processor is based on Zen 2 microarchitecture, with a TDP of 225 W and built on the 7 nm process technology.
Uses of High Frequency CPU Server
High Frequency CPU Servers handle high-performance computing (HPC) workloads easily due to their superior clock speeds, large bandwidth, and fast I/O transfers. The ability to perform quadrillions of calculations is immensely useful in data-sensitive applications, hyper-converged infrastructure, increased VM density, and processing complex calculations with humongous data in real-time.
The EPYC server series provides the computing backbone for Google Cloud, Amazon Web Services, and Microsoft’s Azure services. Research labs use clusters of high-frequency CPU servers for predicting and tracking storms and testing new products. Live event streaming, rendering mind-blowing special effects, and 3D imaging need powerful CPUs. Artificial intelligence and machine learning use HPC in applications ranging from detecting credit card fraud to teaching self-driving vehicles. Real-time stock trends and automated trading platforms use high-frequency servers extensively.
Comparison of CPU’s used by LayerStack and other top cloud providers.
LayerStack uses AMD’s powerful EPYC series processors in all its plans. According to Passmark’s CPU benchmarking test, LayerStack CPUs’ score is significantly higher than the CPUs selected by most of the top cloud providers, we have picked a few providers and compare the CPU efficiency.
*The comparison below is for reference purpose only, the information may vary due to different cloud server deployment, LayerStack bears no responsibility for any listed information.
LayerStack
The Standard Plan uses AMD EPYC 7662 and 7702, with clockspeeds between a base 2.0 GHz and max turbo 3.35 GHz. They both have 64 cores and supports 128 threads simultaneously, with a TDP up to 225W.
The High Frequency CPU Plan uses the AMD EPYC 7742 which is highly scored by PassMark Software.
DO Cloud Provider
The Standard Plan uses Intel Xeon E5-2650 v4, with clockspeeds between a base 2.2 GHz and turbo 2.9 GHz. Launched in Q2 2016, the Xeon has 12 cores and supports 24 threads, with a TDP of 105W.
The CPU-optimized Plan uses Intel Xeon Gold, with clockspeeds between a base 2.3 GHz and turbo 3.7 GHz. Launched in Q4 2017, it has 18 cores and supports 36 threads, with a TDP of 140W.
L Cloud Provider
The Standard Plan uses both Intel Xeon E5-2697 v4, with clockspeeds between a base 2.3 GHz and turbo 3.6 GHz, and AMD EPYC 7501 between 2.0 GHz and 3.0 GHz. The Xeon has 18 cores and supports 36 threads, while the 7501 has 32 cores and supports 64 threads.
The Dedicated-CPU Plan uses AMD EPYC 7601, with clockspeeds between a base 2.2 GHz and turbo 2.7 GHz. Launched in Q4 2017, the 7601 has 32 cores and supports 64 threads, with a TDP of 180W.
Comparison Result
LayerStack’s EPYC 7702 and 7742 have the highest CPU Mark scores, followed by L Brand’s EPYC 7601 and 7501. DO Brand’s Xeon E5-2560 scores the lowest.
To cope with the demanding workloads, LayerStack maximize the cloud performance with AMD EPYC to deliver the best cloud computing service to our valued customers. Today, AMD dominates the cloud computing market, the 2nd Gen AMD EPYC™ Processors series is specifically designed for cloud-based solutions which sets a higher standard for cloud deployments. Innovative design makes AMD EPYC™ #1 in performance across industry to propel cloud users’ workloads simply and cost-effectively.
LayerStack Cloud Configuration & Use Case
LayerStack cloud servers are configured with top-class AMD EPYC aiming to provide leading cloud efficiency in the industry. It delivers outstanding values for users to manage most of the general and CPU-intensive workloads such as machine learning, video editing, compiling programs and AD serving.
LayerStack standard cloud servers uses 2nd Gen AMD EPYC 64 Core CPU. All selected CPU are with 64 cores and 128 threads and launched after Q3 2019. They runs with base clock speed 2.0 GHz and max. turbo boost 3.4 GHz. Moreover, LayerStack’s coming high frequency CPU servers are configured with AMD EPYC 7742 with higher operating frequency and turbo speed.
CPU Comparison By Geekbench Browser
The new High Frequency CPU Cloud Server (Launch on 27/10/2020) is specially for CPU-intensive cloud usage. We have conducted a simple comparison for the CPU used by top cloud providers with Geekbench Browser on 23rd Oct , a cross-platform processor benchmark to evaluate single-core and multi-core performance. We purchased the most similar optimized-CPU servers with 2 vCPU and 4GB Ram directly from each provider for comparison. It can be seen that LayerStack High Frequency CPU Server is rated the highest with 7919 multi-core score. Vutlr is ranked second with 7820 points. AWS Lightsail and DigitalOcean are at the third and the fourth place with 6246 and 4670 points respectively.
