mClouds: Cloud GPU platform

The main articles are:

• What is IaaS
• Cloud computing

2024: GPU Cloud Platform Launch

mClouds December 17, 2024 announced the launch of the GPU- platform. clouds It includes high-frequency processors AMD EPYC 9374F and performance video cards Nvidia L4, L40S and A16. Business will be able to speed up the processes by an order of magnitude: machine learning analysis, big data work with, neuronets design in and, BIM CAD video processing and 3D rendering. Thus, it will be possible to increase productivity and remain competitive in the market.

With the development of artificial intelligence and automation, business needs for powerful computing resources are growing exponentially. mClouds offers customers the ability to work with modern tasks in the cloud - without the need to buy expensive GPU servers and deploy their own infrastructure.

The key advantage of the solution is the architecture, which includes NVIDIA GPUs and AMD's flagship CPUs. This configuration provides high performance for fast parallel computing and high-speed tasks. The platform allows you to flexibly manage resources depending on the needs of the project.

Companies working with high-load tasks will be able to optimize work on computing models: training and testing of neural networks is accelerated by one and a half to two times. The combination of NVIDIA graphics cards and high-frequency EPYC processors reduces 3D animation and design rendering time. On the platform, you can quickly process large amounts of data and video with minimal delays so that your business can quickly solve important problems.

The main components and capabilities of the mClouds cloud GPU platform

• AMD EPYC 9374F is a top-end processor based on the GENOA Zen4 architecture with 32 cores. The clock speed is 3.85-4.1 GHz, the cache is expanded to 32 MB L2 and up to 256 MB L3.
• NVIDIA A16 with 64 GB of video memory, 5120 CUDA cores and 160 Tensor cores is suitable for professional BIM and CAD tasks.
• NVIDIA L40S is equipped with 48GB of video memory, 18,176  CUDA cores, 142 RT cores and 568 Tensor cores. This is optimal for complex computational models, rendering and processing large amounts of data.
• NVIDIA L4 with 24 GB of video memory, 7680 CUDA cores, 60 RT cores and 240 Tensor cores provides high performance in neural network learning and data analysis tasks.

The platform is located in a certified Tier III Gold data center  - NORD4 in Moscow. This guarantees the reliability and availability of services.

We're seeing computing power become a key growth driver for the business. Our AMD EPYC and NVIDIA-based platform takes customer computing to a completely different level. We provide customers with not just cloud resources, but a tool that will help them remain competitive in the face of digital transformation and accelerating AI adoption in business, "said Alexander Ivannikov, Development Director of Cloud Infrastructure Provider mClouds.