Boosteroid is a cloud services platform that provides convenient access to computing power, storages and software products.
Cloud computer will allow to process video, create 3D-graphics, play games requiring high computer performance or carry out machine learning even on the device with the lowest performance. The only thing that user needs is a stable connection to the high-speed Internet.
Another Boosteroid solution is low cloud computing rent. Thanks to the lowest computing power rent Boosteroid will become affordable for every World Wide Web user.
Boosteroid app will provide the opportunity to use our services and select needed amount of computing power just in two mouse clicks. Users will pay for the service with BTR tokens, as well as USD and EUR.
Technical implementation of the Boosteroid project consists of software and hardware parts.
Hardware part will include a lot of heterogeneous computing systems, combined into a single high-speed infrastructure:
- - servers based on modern general-purpose Intel or AMD processors;
- - servers, which include NVIDIA or AMD graphics accelerators and graphics cards;
- - servers, which include Intel graphics accelerators;
- - high-speed switching equipment.
Server architecture will be based on:
- - Intel Xeon E5, E7 v4 or 5 processors;
- - AMD EPYC 7000 series processors;
- - NVIDIA GeForce GTX 1080 Ti graphics cards;
- - NVIDIA Tesla P100, Tesla V100 and AMD Radeon Instinct graphics accelerators;
- - Intel Xeon Phi 7000 series accelerators.
A fault-tolerant, distributed, redundant storage system will be used in the Boosteroid project. Also, block-level stores are provided. In addition to the main data stores, a long-term storage with a large amount of memory will be built.
10Gb and 100Gb Ethernet switches will be used as communication equipment. They support the operation of the system infrastructure, high-speed computing network and data warehouse replication.
Software part will be based on VMware software products, designed for server, storage and network virtualization, as well as IT-infrastructure transformation into stable automated computing environment.
This concept will allow to create many virtual computers within a computing cloud based on Intel Xeon processors and graphics accelerators (cards) NVIDIA with minimal capacity losses.