Tenders Are Invited For Super Computer - Cyber ​​Security Laborate - Target Computer

Super Computer - Cyber ​​Security Laborate - Target Computer. 1) Product Ownerships For The 28 Cyber Security Testing Machine Unit: It Is An Expectation That It Should Be Used In The Desktop Environment. Maximum 155X155x55mm - Usff. -The System Provides A Continuously High Ai Computing Performance With At Least 980 Tps Capacity In Infections And Fine Tuning Profiles, As Well As A Minimum Of ~ 0.9 Petaflop Fp4. -The System Should Have A High-Performance Cpu Based On Arm V9.2-A Architecture With At Least 18 Cores And 3.0 Ghz. - The System Has A Dedicated Gpu For Ai Tasks Containing At Least 5 Generations (Or Equivalent) Tensor Core Units And Supports The Fp4 Computing Format. - Cpu And Gpu Should Operate In Close Integration With A Common Memory Architecture. Nvidia Gb10 Grace -Blackwell Superchip Or Equivalent Solution Can Be Considered As Appropriate. - The System Should Have A High -Bandwidth System Memory With A Capacity Of At Least 120 Gb, Which Is Uniformly Available To Cpu And Gpu. - The System Has Internal Storage Facilities In The Form Of Nvme Ssd With A Capacity Of At Least 960 Gb, M.2 Attachment And A Minimum Pcie Gen4 × 4 Standard. - The System Provides High -Speed Cpu -Gpu Data Transmission Using A Coherent Single Memory Architecture With A Data Transmission Bandwidth Of At Least 650 Gb/S. Nvidia Nvlink C2c Or Equivalent Technology Can Be Considered As Appropriate. - The System Should Be Scaled To Run Large Language Models And Have An Interconnect Solution Providing At Least 160 Gb/S Network Bandwidth, Which Allows You To Connect Multiple Processor Units. The Solution Must Support The Running Of At Least 160 Billion Parameters, Depending On Workload And Quantization Settings. The Technology Of Nvidia Connectx-7 Or Equivalent Can Be Considered As Appropriate. - The System Has A Wired Network Connection With At Least 10 Gigabit Ethernet Lan Port. - The System Also Has Wireless Connection, At Least Intel Wi-Fi 7 (Gig +) 2 × 2 + Bluetooth 5.4 Support. - The System Should Have Energy Efficient And Reliable Cooling, Which Ensures The Sustainability Of Stable, High Performance In Compact Design. The Energy Requirements Of Some Units Of The Offered System Should Not Exceed 250 W. - The System Should Run On A Linux -Based Operating System. Ubuntu Linux Or Equivalent Distribution Can Be Considered As Appropriate. - Support The System Development Environment, Which Includes Leading Ai Framework (Eg Cuda Or Equivalent Calculation Platform, Pytorch, Tensorflow, Jupyter) And Provides Containerization And Cache Optimization Options. Nvidia Ai Software Stack (Nim ™, Blueprints, Tensorrt, Ollama, Complete Nvidia Ai Ecosystem Support) Or Equivalent To This