Computer Hardware & Networking — 100 Questions & Toggle Answers

Searchable accordion for interview prep and study.
RAM (Random Access Memory) is volatile memory used for temporary data storage while the computer is running. ROM (Read-Only Memory) is non-volatile and typically stores firmware or boot code.
CPU stands for Central Processing Unit; it executes instructions from programs, performs calculations, and controls other hardware components.
The motherboard is the main printed circuit board that houses the CPU, memory, chipset, expansion slots, and connectors for peripherals.
SSD (Solid State Drive) uses flash memory to store data and has faster read/write speeds, lower latency, and no moving parts compared to HDD (Hard Disk Drive).
GPU (Graphics Processing Unit) accelerates rendering of images, video, and animations; it's also used for parallel processing tasks like AI and scientific computing.
Cores are independent processing units on the CPU die. Threads are virtual CPU paths (often via hyper-threading) that allow a core to manage multiple instruction streams.
BIOS (Basic Input/Output System) and UEFI (Unified Extensible Firmware Interface) are firmware interfaces that initialize hardware and boot the operating system.
Overclocking is running a CPU/GPU at higher clock speeds than the manufacturer specifies, often to improve performance at the cost of heat and stability.
Thermal paste fills microscopic gaps between the CPU heatspreader and cooler to improve heat transfer.
An expansion slot (e.g., PCIe) on the motherboard allows adding cards like GPUs, network adapters, or storage controllers.
RAID (Redundant Array of Independent Disks) combines multiple drives for redundancy/performance. Common levels: RAID 0 (striping), RAID 1 (mirroring), RAID 5 (striped with parity), RAID 6 (double parity), RAID 10 (mirror+stripe).
The PSU converts AC mains power to regulated DC voltages needed by the PC components and provides power connectors for devices.
Form factor refers to the physical specifications of the motherboard and chassis (e.g., ATX, microATX, Mini-ITX) that determine size and connector placement.
ECC (Error-Correcting Code) memory detects and corrects single-bit memory errors and is used in servers and workstations requiring high reliability.
Virtual memory extends physical RAM using disk space (swap/page file), allowing larger addressable memory at the cost of speed.
NIC (Network Interface Card) connects a computer to a network via Ethernet, Wi‑Fi, or other mediums.
The OSI model is a 7-layer conceptual framework for network communications: Physical, Data Link, Network, Transport, Session, Presentation, Application.
TCP (Transmission Control Protocol) is connection-oriented and reliable with error checking and retransmission. UDP (User Datagram Protocol) is connectionless and faster but unreliable.
An IP (Internet Protocol) address uniquely identifies a device on an IP network; IPv4 uses 32 bits, IPv6 uses 128 bits.
Subnetting divides a larger network into smaller logical subnetworks, controlled by subnet masks, to improve routing and security.
The default gateway is the router IP where packets destined for other networks are sent for routing.
NAT (Network Address Translation) translates private IP addresses to a public IP for internet access and can be used for address conservation and security.
DHCP (Dynamic Host Configuration Protocol) dynamically assigns IP addresses and network configuration to hosts on a network.
DNS (Domain Name System) translates human-readable domain names to IP addresses using distributed servers and records.
VLAN (Virtual LAN) segments network traffic at Layer 2, creating logical networks within the same physical switch infrastructure.
A switch forwards frames within a LAN at Layer 2; a router forwards packets between networks at Layer 3 and performs routing decisions.
MAC (Media Access Control) address is a hardware identifier burned into network interfaces used for Layer 2 addressing and switching.
PoE (Power over Ethernet) supplies electrical power over Ethernet cabling to devices like IP phones and wireless APs.
Cat5e, Cat6, Cat6a, Cat7 — differing in bandwidth, shielding, and maximum frequency, affecting throughput and distance.
Fiber uses light to transmit data and offers higher bandwidth, longer distances, and immunity to electromagnetic interference compared to copper.
A firewall enforces network security policies by filtering traffic based on rules; can be hardware, software, or cloud-based.
SSL/TLS are cryptographic protocols that secure communications over networks via encryption and certificate-based authentication.
Port forwarding maps incoming traffic on a router's public IP and port to an internal host and port, enabling external access to services.
DMZ (Demilitarized Zone) is an isolated subnet where public-facing servers reside, separated from the internal network by firewalls.
Bandwidth is the theoretical maximum data rate of a link; throughput is the actual measured data rate achieved.
Latency is the delay for packets to traverse a network; jitter is the variation in packet arrival times, affecting real-time apps.
QoS (Quality of Service) prioritizes certain traffic types or flows to ensure performance for latency-sensitive applications.
Link aggregation combines multiple physical links into a single logical channel to increase bandwidth and provide redundancy; LACP is the protocol used.
STP (Spanning Tree Protocol) prevents switching loops in redundant Layer 2 networks by electing root bridges and blocking ports.
RSTP (Rapid STP) converges faster than STP; MSTP (Multiple STP) allows multiple VLANs to be mapped into STP instances for scalability.
BGP (Border Gateway Protocol) is the inter-domain routing protocol that exchanges routing info between autonomous systems on the internet.
OSPF (Open Shortest Path First) is an interior gateway protocol (IGP) that uses link-state routing for fast convergence within an AS.
A routing table contains known network routes and next-hop information used by routers to forward IP packets.
ARP (Address Resolution Protocol) maps IPv4 addresses to MAC addresses on a local network segment.
ICMP (Internet Control Message Protocol) is used for network diagnostics and reporting errors (e.g., ping, traceroute uses ICMP/TTL).
PAT (Port Address Translation) allows multiple private hosts to share a single public IP by translating internal IPs to unique source ports.
SSID is the human-readable wireless network name; BSSID is the MAC address of the access point radio serving that SSID.
Common bands: 2.4 GHz (longer range, fewer channels) and 5 GHz (more channels, higher speeds). Channels are frequency segments within a band.
WPA2 uses AES and PSK/Enterprise modes for security. WPA3 adds stronger encryption, individualized data encryption, and improved protection against password guessing.
Antenna gain measures how well an antenna focuses energy in a direction (dBi). Polarization is the orientation of the electromagnetic wave (vertical/horizontal).
Auto-negotiation allows devices to automatically choose optimal speed and duplex settings (e.g., 100Mbps/full) when connecting.
MTU (Maximum Transmission Unit) is the largest packet size that can be transmitted. Mismatched MTUs cause fragmentation or dropped packets.
802.1Q tags VLAN IDs into Ethernet frames so multiple VLANs can traverse a single link between switches or to a router.
A proxy server intermediates client requests to external servers for caching, filtering, or security and can hide client IPs.
SSL inspection intercepts and decrypts TLS traffic for security scanning before re-encrypting it to the destination.
Segmentation divides networks into isolated zones to reduce attack surface, improve performance, and enforce policy.
Unicast sends to one destination; broadcast sends to all devices in subnet; multicast sends to a group of subscribed hosts.
IPv6 provides a vastly larger address space (128-bit), simplifications for routing and autoconfiguration, and improvements in security and mobility.
SLA (Service Level Agreement) defines performance, availability, and metrics a provider must meet for a service.
A network loop creates duplicate frames and broadcast storms. Prevent with STP, proper topology design, and loop detection features.
Packet capture records network traffic for analysis. Wireshark is a widely-used packet analyzer.
Port security restricts which MAC addresses can connect to a switch port, helping prevent unauthorized devices.
A web page that users must interact with (login/accept terms) before gaining network access, commonly used in public Wi‑Fi.
SNMP (Simple Network Management Protocol) monitors and manages network devices using agents and a management station.
Both are AAA protocols. RADIUS combines authentication and accounting (UDP), TACACS+ separates auth from accounting and uses TCP with more granular command authorization.
VRF (Virtual Routing and Forwarding) creates multiple routing tables on the same router for logical separation of networks.
A CDN caches and serves content from geographically distributed nodes to reduce latency and offload origin servers.
DNS caching stores DNS responses for a duration defined by TTL (Time to Live) to reduce lookup latency and load on authoritative servers.
Collision domain is where Ethernet collisions occur (reduced by switches). Broadcast domain is bounded by routers; VLANs define broadcast domains.
UTP (Unshielded Twisted Pair) lacks shielding; STP (Shielded Twisted Pair) includes shielding to reduce EMI in noisy environments.
Managed switches offer configuration, monitoring, VLANs, and QoS. Unmanaged switches provide plug-and-play connectivity without management features.
Hypervisor enables virtualization. Type 1 (bare-metal) runs directly on hardware (e.g., ESXi); Type 2 runs on a host OS (e.g., VirtualBox).
A DHCP reservation maps a specific MAC address to a fixed IP address, ensuring a host receives the same IP via DHCP.
Link-state protocols (OSPF) build a complete topology and compute shortest paths; distance-vector (RIP) share route distances with neighbors.
A route reflector reduces the need for full mesh iBGP by reflecting routes to other iBGP peers within an AS.
Static NAT maps one private IP to one public IP for inbound and outbound communication without port translation.
A wireless controller centralizes management of multiple APs, handling configuration, roaming, and security policies.
An SSL certificate binds a public key to an identity and is issued by a Certificate Authority (CA) trusted by clients.
Zero-trust assumes no implicit trust; it enforces continuous verification, least privilege access, and segmentation across resources.
Network virtualization abstracts hardware into virtual networks; SDN (Software-Defined Networking) separates control and data planes for programmable networks.
DNS/host file poisoning is a cyberattack that corrupts DNS caches or host mappings to redirect traffic to malicious hosts.
Kernel panic (Unix) or Blue Screen of Death (Windows) indicates a critical system error often caused by driver faults, hardware failure, or OS bugs.
POST (Power-On Self-Test) is the firmware test sequence that checks hardware components before booting the OS.
ECC detects and corrects single-bit errors; parity detects (but doesn't correct) single-bit errors; non-parity has no error checking.
A heatsink conducts heat away from components using fins and increased surface area; often paired with fans for active cooling.
S.M.A.R.T. (Self-Monitoring, Analysis, and Reporting Technology) reports drive health indicators to predict failures.
T568A and T568B define wire color-to-pin mappings for Ethernet connectors; both are functionally equivalent if both ends match.
Dual-band supports two frequency bands (2.4 & 5 GHz), tri-band adds an extra 5 GHz radio for more concurrent bandwidth.
PXE (Preboot Execution Environment) allows a machine to boot an OS installer or image from a network server before local storage boot.
Hotspot is a single AP providing connectivity; mesh Wi‑Fi uses multiple interconnected APs to provide seamless coverage and self-healing paths.
Loopback is a virtual network interface (127.0.0.1/::1) used for testing and local inter-process communication.
The chipset controls communication between CPU, memory, and peripherals, determining supported features and I/O.
Beep codes signal hardware initialization problems during POST, varying by BIOS vendor (AMI, Phoenix).
ECC DIMM includes error correction and is used in servers; UDIMM (unbuffered) is common in desktops without ECC.
A battery-backed cache preserves unwritten RAID controller cache to prevent data loss on power failure.
Firmware is low-level code on hardware devices; drivers are OS-level software that interface the OS with hardware.
A staged attack uses initial foothold, then downloads additional payloads — defenders use segmentation and endpoint detection to mitigate.
Circuit switching reserves a dedicated path (telephony); packet switching sends independent packets over shared networks (IP).
SFP supports lower data rates (1Gbps); SFP+ supports higher rates like 10Gbps in the same small form factor.
Use proper VLAN/storm control, and avoid bridging loops; enable features like bridge loop detection on APs/switches.