RAM (Random Access Memory) is a type of volatile memory used to store data and machine code currently being used by the computer, allowing for quick read and write operations. It is temporary storage that requires power to retain information . In contrast, ROM (Read-Only Memory) is non-volatile, permanent storage that holds essential instructions for boot-up processes and system firmware which are not meant to be modified frequently . RAM is used for data that need fast access and changes frequently, while ROM contains critical bootstrap instructions necessary for initial hardware checks and system configuration .

MAC (Media Access Control) addresses are unique identifiers assigned to network interfaces for communication at the data link layer (Layer 2) of a network, typically burned into hardware . They are crucial for local network communication by ensuring data packets reach the right physical device . IP (Internet Protocol) addresses, on the other hand, are logical addresses assigned by the network for identifying devices on a network at the network layer (Layer 3), enabling global internet communication . While MAC addresses handle local network routing within a physical network, IP addresses manage routing between networks across the internet .

The primary functions of the BIOS (Basic Input Output System) include initializing and testing hardware components through the Power-On Self Test (POST) process, loading the bootloader to start the operating system, providing a set of low-level routines for interfacing with hardware directly, and managing data flow between the computer's operating system and attached devices such as hard drives, video adapters, keyboard, and mouse . BIOS is critical because it ensures that hardware is operating correctly before handing over control to the operating system, thereby maintaining system stability and reliability .

The OSI (Open Systems Interconnection) model has seven layers: physical, data link, network, transport, session, presentation, and application, which abstract and describe the functions of a network system . The TCP/IP model, on the other hand, has four layers: link, internet, transport, and application, offering a more concise and practical design . The OSI model serves as a more conceptual framework that helps in understanding the inter-operation of network protocols, while the TCP/IP model, being less abstract, is widely used in real-world networks because it maps more directly to the actual stack in use . This means that network design using TCP/IP is often streamlined, focusing on practical deployment rather than theoretical models .

The primary network topologies include bus, star, ring, and mesh. The bus topology uses a single central cable to which all network devices are connected; it is easy to install but difficult to manage as it can become problematic if the central cable fails . Star topology connects all devices to a central hub or switch; it's easy to add devices and manage, but failure in the hub can bring down the entire network . Ring topology connects each device to two others, forming a circular pathway for data; it's relatively easy to install, but if any device fails, the entire network can be affected . Mesh topology, where every device is connected to every other device, provides high redundancy and reliability at the cost of higher installation complexity and expense .

SRAM (Static RAM) and DRAM (Dynamic RAM) differ primarily in their technology and speed. SRAM uses bistable latching circuitry to store each bit, making it faster and more reliable, as it doesn't require refreshing as DRAM does . DRAM, on the other hand, stores each bit in a separate capacitor within an integrated circuit, offering higher densities and cheaper cost, but it requires constant refreshing due to capacitor leakage. SRAM is typically used in CPU caches where speed is crucial, while DRAM is used for main memory where larger storage capacity is needed .

Online UPS (Uninterruptible Power Supply) systems provide continuous power from the inverter, ensuring zero transfer time during a power failure and offering better power conditioning and consistent voltage regulation . Offline UPS systems switch to battery power upon detecting main power failure, which can cause a brief transfer time, making the equipment vulnerable to short power fluctuations . Businesses opt for online UPS systems for critical systems or when sensitive equipment demands a stable and uninterrupted power supply, despite the higher cost, while offline UPS systems are suitable for environments where occasional power fluctuation is permissible and cost-efficiency is a priority .

IPv4, with its 32-bit addressing, can provide approximately 4.3 billion unique addresses, which are insufficient due to the rapid growth of internet-connected devices . IPv6, employing 128-bit addressing, exponentially increases the number of unique addresses and supports a vast array of devices far into the future . Furthermore, IPv6 includes features like better security protocols at the network layer and more efficient routing . However, the transition to IPv6 poses challenges including the need for reconfiguring network equipment, updating software, and ensuring compatibility with IPv4 systems, all of which involve significant time and financial investment .

Classful IP addressing divides IP address spaces into fixed classes (A, B, C), each with a predetermined size, which led to inefficient use of IP addresses . Classless IP addressing, introduced with CIDR (Classless Inter-Domain Routing), allows for more flexible allocation of IP addresses by representing them with a variable-length subnet mask, thus optimizing address space allocation . In modern networking, classless addressing has become essential for scaling the internet efficiently, as it reduces wastage of address spaces and allows for more precise network growth and management .

Subnetting in IP networks refers to dividing a larger network into smaller, more manageable segments or subnets . This practice optimizes network performance by reducing broadcast traffic, improving network security via isolation of network segments, and enhancing management by allowing specific IP ranges for different departments . Additionally, subnetting maximizes address space utilization by tailoring the subnet sizes to the actual needs, rather than wasting IP addresses in underutilized networks .