Objective 3.5 is the broadest single objective in Domain 3, covering the central physical and firmware components that everything else in a computer connects to and depends on. The motherboard is the foundation that determines what size case you need, what CPU you can install, what expansion cards will fit, and what firmware-level security and configuration options are available. Get any of these compatibility decisions wrong, and the build simply won't work — which is exactly why this objective is scenario-heavy on the exam.
This guide follows the order of the objective itself: motherboard form factors, connector types, CPU/socket compatibility, BIOS/UEFI settings, virtualization and encryption support, CPU architecture, expansion cards, and cooling. Treat the early sections (form factor, sockets) as the physical foundation everything else builds on, and the later sections (BIOS/UEFI, cooling) as the configuration and maintenance layer on top of that foundation.
The Central Compatibility Chain
Building or upgrading a system is fundamentally a chain of compatibility checks: case → motherboard form factor → CPU socket → RAM type (objective 3.3) → expansion card slots → power supply connectors. A mismatch at any link breaks the whole chain. Most of this objective's scenario questions are testing whether you can correctly identify where in that chain a described problem actually lives.
A motherboard's form factor defines its physical dimensions, mounting hole layout, and (generally) how many expansion slots it can fit — which in turn determines what size case it requires and how much room is available for expansion cards, drives, and cooling.
Exam Angle
Remember the general size relationship — ATX > microATX > ITX — and that this same relationship usually holds for expansion slot count and overall flexibility. A smaller form factor board generally fits in a case designed for a larger form factor (a microATX board in an ATX case), but not the reverse. If a scenario emphasizes "smallest possible build" or "limited space," ITX is almost always the answer.
A motherboard exposes a range of connector types for expansion cards, storage, power, and front-panel/internal peripherals. Recognizing each connector by name and purpose is directly testable, often via photo identification.
PCIe Slot Size vs. Bandwidth
A larger physical PCIe slot (x16) doesn't necessarily mean every lane is electrically wired — some motherboards have physically x16-sized slots that are only electrically wired as x4 or x8 to save cost, often noted in the motherboard's documentation. Most cards will still function in a slot with fewer wired lanes than the card supports, just at reduced bandwidth — this is a real-world nuance worth knowing conceptually, even though the exam focuses mainly on recognizing slot types rather than this electrical detail.
A motherboard's CPU socket is the physical interface the processor plugs into, and it is brand- and generation-specific — a CPU will only physically fit (and only function correctly) in a socket it was specifically designed for. This is the single most important compatibility check when selecting a CPU and motherboard together.
Common Pitfall — Socket Compatibility Is Absolute
An AMD CPU will never fit an Intel socket, and vice versa — there is no cross-brand compatibility at all. Beyond brand, even within the same brand, different socket generations are typically incompatible with each other (an older AMD AM4 CPU will not fit a newer AM5 socket, for example). Always verify the exact socket type supported by the motherboard against the exact socket type required by the CPU before purchasing or attempting installation — specific socket names and numbers change over time and should be verified against current manufacturer documentation rather than memorized as fixed facts.
A multisocket motherboard contains two or more physical CPU sockets, allowing multiple processors to be installed and work together in the same system. This is exclusively a server/workstation-class feature, used in scenarios demanding extreme processing power, memory capacity, or core counts beyond what a single-CPU system can provide — virtualization hosts, high-performance computing, and large database servers are common examples. Consumer desktop and laptop motherboards are single-socket only.
Exam Angle
Expect direct matching: "Which CPU socket family does this AMD/Intel processor require?" tests brand-level recognition rather than requiring you to memorize every specific socket number, since those change with each new CPU generation. The conceptual takeaway the exam wants is simpler and more durable: brand and generation must match exactly between CPU and socket, and multisocket boards are a server-class feature, not something found in typical consumer builds.
The BIOS (Basic Input/Output System) and its modern successor UEFI (Unified Extensible Firmware Interface) are firmware built into the motherboard that initialize hardware and hand off control to the operating system during boot. UEFI is the modern standard, offering a graphical interface, support for larger boot drives, faster boot times, and more advanced security features compared to legacy BIOS — though the terms are still frequently used interchangeably in casual conversation, including occasionally on the exam itself.
Boot Password vs. BIOS Password — Don't Conflate Them
These are two distinct security layers that are easy to confuse. A boot password (sometimes called a "power-on password" or "user password") blocks the system from booting into the OS at all without it. A BIOS/UEFI password (sometimes called a "supervisor" or "administrator" password) specifically protects access to the firmware settings menu, preventing changes to boot order, security settings, etc. — but doesn't necessarily prevent the system from booting normally. A system can implement either, both, or neither.
Hardware virtualization support (commonly Intel VT-x or AMD-V) is a CPU feature, enabled/disabled via a BIOS/UEFI setting, that allows a hypervisor to run virtual machines efficiently by handling certain virtualization-related operations directly in hardware rather than relying entirely on slower software emulation. This setting must typically be manually enabled in BIOS/UEFI before virtualization software (such as Hyper-V, VMware, or VirtualBox running hardware-accelerated VMs) will function — a very common real-world troubleshooting scenario when virtualization software reports it cannot start a VM.
A TPM is a dedicated, tamper-resistant hardware chip (or firmware-based equivalent) built into or integrated with the motherboard, used to generate, store, and protect cryptographic keys used for disk encryption, secure boot verification, and other hardware-based security functions. Critically, the TPM stores its keys in a way that's tied to the specific physical hardware — if certain hardware changes occur, or someone attempts to move an encrypted drive to different hardware, the TPM can detect this and block key release, providing protection against many forms of offline/physical tampering.
Exam Angle — TPM and BitLocker
TPM is most commonly referenced on the exam in connection with Windows BitLocker full-disk encryption, which by default relies on the TPM to securely store its encryption keys and verify system integrity before unlocking the drive at boot. TPM is also a hard requirement for modern Windows operating system installations on supported hardware — a frequently tested real-world compatibility checkpoint.
A HSM (Hardware Security Module) is a dedicated, often external or rack-mounted physical device specifically built to generate, store, and manage cryptographic keys and perform cryptographic operations (encryption, decryption, digital signing) at scale, typically for an entire organization rather than a single device.
TPM vs. HSM — Scale Is the Key Difference
Think of a TPM as protecting the cryptographic keys for one specific device — it's built into or attached to that single motherboard. An HSM, by contrast, is typically a separate, dedicated, often network-attached or rack-mounted device protecting cryptographic operations for many systems or an entire organization — commonly used by certificate authorities, payment processors, and enterprises with heavy compliance or large-scale key-management requirements. Both protect cryptographic keys; they differ primarily in scale and deployment context.
x86 refers to the instruction set architecture originally developed by Intel, historically referring to 32-bit processors and software. x64 (also called x86-64 or AMD64) is the 64-bit extension of that same architecture family, now the standard for virtually all modern desktop, laptop, and server CPUs from both Intel and AMD. A 64-bit (x64) CPU and operating system can address vastly more system RAM than a 32-bit system was ever capable of, among other performance and capability advantages.
ARM (Advanced RISC Machine) is a fundamentally different processor architecture family, built on RISC (Reduced Instruction Set Computer) design principles — using a smaller, simpler set of instructions compared to x86/x64's more complex instruction set (CISC-style design). ARM processors are widely known for excellent power efficiency relative to performance, which is why they dominate smartphones, tablets, and increasingly appear in laptops and even some servers/desktops (such as Apple Silicon Macs) where battery life or power efficiency at scale is a priority.
| Factor | x86/x64 | ARM |
|---|---|---|
| Design philosophy | CISC-influenced, complex instruction set | RISC — reduced, simpler instruction set |
| Power efficiency | Generally higher power consumption | Generally much more power-efficient |
| Typical devices | Traditional desktops, laptops, servers | Smartphones, tablets, increasingly some laptops/servers |
| Software compatibility | Vast existing software library built for this architecture | Requires ARM-native software, or emulation/translation of x86 software |
A modern CPU contains multiple processing cores — independent processing units within a single physical CPU package, each capable of executing instructions somewhat independently. More cores generally allow better performance on workloads that can be split across multiple simultaneous tasks (multithreaded applications, running many programs at once), though single-threaded application performance depends more on individual core speed/efficiency than on core count alone. Common consumer configurations include dual-core, quad-core, hexa-core (6), octa-core (8), and higher core counts increasingly common on enthusiast and workstation-class CPUs.
More Cores ≠ Automatically Faster for Every Task
A higher core count helps most with workloads designed to run in parallel — video editing/rendering, virtualization, heavily multitasked environments. But software that isn't written to take advantage of multiple cores (some older or simpler applications) won't see proportional benefit from additional cores, and in some cases a CPU with fewer but faster/higher-clocked cores can outperform a higher-core-count CPU on single-threaded tasks. Matching CPU core configuration to the actual workload matters more than simply maximizing core count.
Expansion cards add capabilities to a system beyond what's built into the motherboard, typically installed into a PCIe slot (Part 2, above).
| Card Type | Purpose |
|---|---|
| Sound card | Provides dedicated audio processing, often offering higher fidelity, more inputs/outputs, or lower latency than a motherboard's built-in (onboard) audio |
| Video card (GPU) | Provides dedicated graphics processing, essential for gaming, video editing/rendering, 3D work, and machine learning workloads beyond what integrated graphics can handle |
| Capture card | Captures video/audio input from an external source (a gaming console, camera, or another computer) for recording or live streaming |
| Network interface card (NIC) | Adds or upgrades network connectivity — covered in depth in objective 2.5 — commonly added when a motherboard's onboard NIC is insufficient (e.g., adding 10 Gigabit Ethernet, or adding Wi-Fi to a desktop without built-in wireless) |
CPUs (and other components, like GPUs) generate significant heat under load, and effective cooling is essential to prevent thermal throttling (the CPU automatically slowing itself down to avoid damage) or outright hardware failure.
Common Pitfall
Installing a heat sink without thermal paste, with too little paste, or with old/dried paste left over from a previous installation are among the most common real-world causes of unexpected overheating and thermal throttling after a CPU or cooler swap — always clean off old paste completely and apply a fresh, appropriately-sized application before reseating any heat sink.
Final Exam Reminders
Form factor size order = ATX (largest) > microATX > ITX (smallest); smaller boards generally fit in larger cases, not vice versa.
CPU socket compatibility = brand AND generation must match exactly; AMD never fits Intel, and old generations rarely fit new sockets.
Multisocket = server/workstation only, never consumer desktops/laptops.
Secure Boot = blocks unsigned bootloaders. Boot password = blocks booting at all. BIOS/UEFI password = blocks firmware setting changes. Three distinct protections.
TPM = single-device key storage, ties to BitLocker and modern Windows requirements. HSM = organization-scale, dedicated, often external.
x86/x64 = traditional CISC-influenced architecture. ARM = RISC, power-efficient, dominant in mobile.
Thermal paste = always reapply fresh when reseating any heat sink — never reuse old, dried paste.
Liquid cooling = generally outperforms air cooling under sustained heavy load.