CompTIA A+ Core 1 Domain 3: Hardware

JEDEC and manufacturer datasheets specify DDR5 VDD/VDDQ at 1.1 V nominal, lower than DDR4's 1.2 V, improving efficiency at higher data rates.

• AThis trap reuses the DDR4 operating voltage; DDR4 modules run at 1.2 V, whereas DDR5 dropped lower, so 1.2 V belongs to the previous generation, not DDR5.
• BThis is the low-voltage DDR3L figure, a reduced variant of standard DDR3; it is neither the DDR5 nominal nor the standard DDR3 value, making it a distractor.
• DThis is the standard DDR3 operating voltage; technicians who confuse the oldest of these three generations with the newest will wrongly pick 1.5 V for DDR5.

The 260-pin small-outline DIMM is the short, low-profile DDR4 module made for notebooks and small-form-factor systems, matching the laptop's socket.

• AThis is the full-length desktop/server DIMM; its 288-pin edge and length will not seat in a laptop's compact slot, so it is the wrong physical form factor here.
• CThis names a previous-generation full-size desktop DDR3 module; it is both the wrong generation and the wrong (desktop) form factor for a DDR4 laptop slot.
• DAlthough small-outline, this is the older DDR3 laptop module; its 204-pin notch keying is incompatible with a DDR4 notebook socket, so it cannot be installed.

ECC uses extra parity bits (72-bit vs 64-bit data path) to detect and correct single-bit errors on the fly, improving stability for servers and workstations.

• AThis confuses ECC with overclocking; the extra parity hardware adds reliability, not frequency, and ECC modules typically run at standard or slightly lower speeds, not faster.
• BChannel mode depends on the memory controller and how modules are populated; ECC's extra bits do nothing to create dual-channel pathways, so this is an unrelated trap.
• CThe additional bits store parity, not user data; ECC actually devotes width (72-bit vs 64-bit) to checking, so it does not add usable capacity, let alone double it.

Higher rotational speed brings data under the heads sooner, raising sustained throughput and lowering latency, so the 7200 RPM drive outperforms the 5400 RPM unit.

• BThis inverts the relationship; slower 5400 RPM platters reduce performance and are chosen for quiet, low-power use, not because they transfer data faster than 7200 RPM.
• CRPM directly affects rotational latency and data rate on mechanical drives, so claiming it is irrelevant ignores a primary HDD performance specification.
• DCache helps buffering but does not replace spindle speed; this trap fixates on one spec while ignoring that 7200 RPM rotation is the faster mechanical choice.

NVMe runs over multiple PCIe lanes with lower latency than legacy SATA, delivering several gigabytes per second, far exceeding SATA III's 600 MB/s ceiling.

• ASATA III caps the link at 6 Gb/s (about 600 MB/s), a ceiling modern SSDs saturate, so this interface limits throughput well below PCIe-based drives.
• BDespite the M.2 connector, this drive still rides the SATA bus and its 6 Gb/s limit; the form factor changes but the interface bandwidth does not improve.
• DUSB 3.2 Gen 1 tops out near 5 Gb/s and adds bridge overhead; as an external interface it cannot match internal PCIe NVMe bandwidth for a workstation drive.

RAID 5 stripes data with distributed parity across at least three disks, surviving one drive failure while using only one drive's worth of capacity for parity.

• ARAID 0 stripes data with no parity or mirror, so any single drive failure destroys the entire array; it offers speed but zero fault tolerance, failing the requirement.
• CRAID 1 mirrors disks and survives a failure, but it duplicates data and uses two drives, giving only 50% usable capacity rather than RAID 5's parity efficiency.
• DRAID 10 combines mirroring and striping but needs a minimum of four drives and also yields only 50% capacity, exceeding the three-disk, capacity-efficient requirement.

USB 3.2 Gen 2 (formerly USB 3.1 Gen 2) is the single-lane SuperSpeed+ tier defined at exactly 10 Gbps, matching the drive's required rate.

• AUSB 2.0 (Hi-Speed) peaks at 480 Mbps, roughly one-twentieth of 10 Gbps, so it is far too slow and is a common low-end distractor for high-speed needs.
• BUSB 3.2 Gen 1 (formerly USB 3.0) is rated at 5 Gbps; technicians who assume any USB 3.x equals 10 Gbps will wrongly pick this 5 Gbps tier.
• CGen 2x2 reaches 20 Gbps by bonding two 10 Gbps lanes, so it overshoots the 10 Gbps requirement and needs a USB-C two-lane path the question did not specify.

Thunderbolt 4 over USB-C is certified to always deliver 40 Gbps of data, video, and power on one connection, meeting the dock's bandwidth requirement.

• BThis single-lane SuperSpeed+ tier maxes at 10 Gbps; users who see a USB-C port may assume full speed, but it is a quarter of Thunderbolt 4's 40 Gbps.
• CGen 2x2 reaches only 20 Gbps even using both USB-C lanes, so it falls short of the 40 Gbps the Thunderbolt-class dock requires.
• DUSB 2.0 tops out at 480 Mbps and predates USB-C high-speed and display tunneling, so it cannot approach 40 Gbps for a docking solution.

DisplayPort 1.4 uses HBR3 to provide 32.4 Gbps of link bandwidth and supports 8K at 60 Hz, exactly matching the stated requirement.

• AHDMI is a different connector family using TMDS/FRL signaling; the 32.4 Gbps HBR3 link rate and VESA-certified 8K60 path cited here are DisplayPort 1.4 specifications, not HDMI.
• BThis earlier revision predates HBR3 and is limited to roughly 21.6 Gbps, so it cannot reach the 32.4 Gbps link rate the 8K display requires.
• DVGA is an analog interface with no defined multi-gigabit digital link rate, so it cannot carry a 32.4 Gbps HBR3 DisplayPort signal or 8K resolution at all.

Cat 6A (500 MHz) is the category certified to support 10GBASE-T over the full 100-meter TIA channel, comfortably covering the 90-meter office run.

• ACat 5e is a 100 MHz category specified for 1000BASE-T (Gigabit); it is not rated for 10GBASE-T, so it cannot meet the 10 Gigabit requirement on this run.
• CCat 6 is a 250 MHz category rated for Gigabit; it does not reliably carry 10GBASE-T across a full 100-meter channel, so it is the wrong choice at 90 m.
• DCat 3 is a legacy voice-grade/10BASE-T category at 16 MHz; it is far below the bandwidth needed for 10 Gigabit Ethernet, making it an obvious distractor.

Standard ATX measures 305 mm x 244 mm, the largest of the common ATX-family desktop boards, providing the most expansion slots for a workstation.

• APer Intel's form-factor table microATX is the 244 mm x 244 mm square variant, shorter than ATX's 305 mm depth, so it is not the 305 x 244 mm board specified.
• BMini-ITX is the smallest standard board at 170 mm x 170 mm with limited slots; it is far below 305 x 244 mm, so it cannot be the largest of the three.
• CE-ATX is wider than standard ATX (extending the 244 mm depth), so its footprint does not equal the 305 x 244 mm dimensions that define standard ATX.

A PCIe x16 slot provides sixteen lanes, the widest standard link, giving graphics cards the most bandwidth, which is why GPUs are designed for the x16 slot.

• BA x1 slot carries a single lane and the least bandwidth; it suits sound or network cards, not a GPU that needs the board's widest sixteen-lane link.
• CA x4 slot offers four lanes, more than x1 but well short of x16; using it would mechanically and electrically bottleneck a card built for sixteen lanes.
• DConventional PCI is a slower parallel bus, not PCI Express, and lacks the lanes and bandwidth a modern graphics card requires, so it is the wrong slot entirely.

Intel 12th-generation Alder Lake desktop CPUs use the LGA1700 socket, where the pins reside on the socket and the processor seats with land contacts.

• AAM4 is an AMD socket for Ryzen processors; an Intel Alder Lake CPU will not physically fit it, so this cross-vendor pick is incorrect for a 12th-gen Intel chip.
• BAM5 is AMD's current LGA socket for Ryzen 7000-series CPUs; despite being LGA, it is an AMD platform and incompatible with Intel Alder Lake processors.
• DLGA1200 served Intel 10th/11th-generation CPUs; the larger LGA1700 superseded it for Alder Lake, so a 12th-gen chip will not seat in an LGA1200 board.

80 PLUS certifies the PSU converts AC to DC at a minimum of 80% efficiency at 20%, 50%, and 100% rated load, wasting less power as heat.

• AThe '80' refers to efficiency percentage, not wattage; 80 PLUS units ship in hundreds of watts, so reading it as an 80-watt cap misunderstands the rating entirely.
• CCertification measures conversion efficiency, not a usable-capacity derating; a certified PSU can deliver its full rated wattage, so this is a misreading of the badge.
• DThe standard sets electrical efficiency thresholds and says nothing about fan size; 80 PLUS units commonly use 120 mm or 140 mm fans, so this confuses the number with hardware.

The fuser uses heated rollers and pressure to melt and permanently bond the powdered toner into the paper fibers; a cold or failing fuser leaves toner loose so it smears.

• AThe transfer roller only relocates loose toner from the drum to the paper; it does not melt or bond it, so naming it confuses the transfer step with the fusing step that fixes toner.
• BThe charge roller conditions the drum before exposure and plays no part in fixing toner to paper, so blaming it for smearing confuses an early imaging stage with the final fusing stage.
• DThe drum holds the electrostatic latent image and attracts toner, but it never applies heat or pressure to the page, so it cannot be what fixes toner against smearing.

The laser scans the charged, light-sensitive drum surface, selectively discharging points to form the latent electrostatic image that later attracts toner; a worn drum produces repeating defects.

• BThe fuser acts at the very end of printing to bond toner with heat, long after the image is written, so it cannot be the surface the laser writes the latent image onto.
• CThe transfer roller charges the paper to attract toner during transfer; the laser never writes to it, so picking it confuses the transfer step with where the latent image forms.
• DThe pickup roller only moves paper into the printer and has no photosensitive coating, so it has no role in holding the laser-written electrostatic image of the page.

Piezoelectric heads use crystals that deform when voltage is applied to push ink from the nozzle without heat, matching the described mechanism used in Epson Micro Piezo print heads.

• AThermal inkjet is exactly the heat-based design the vendor says it avoids, so choosing it contradicts the stated no-heat, voltage-driven crystal mechanism that defines the piezo approach.
• BDye-sublimation is a heat-driven, ribbon-based photo technology, not a drop-on-demand inkjet, so it does not match a print head that mechanically flexes a crystal to spray liquid ink.
• CElectrophotographic printing is the laser/toner process, not an inkjet at all, so naming it ignores that the vendor clearly described liquid ink ejected from nozzles by a crystal.

Missing lines and white streaks with adequate ink indicate clogged nozzles; running a cleaning cycle clears them and a nozzle check pattern verifies the head was restored.

• AA fuser is a laser-printer component and does not exist in an inkjet, so replacing it cannot fix nozzle-related streaks and shows a mix-up between inkjet and laser hardware.
• CSpindle speed applies to spinning hard drives, not inkjet print heads, so adjusting it does nothing for clogged nozzles and confuses a storage specification with a printing problem.
• DImaging drums and toner belong to laser printers; swapping them on an inkjet wastes parts and ignores that streaking ink output points to nozzles that simply need cleaning.

An impact dot-matrix print head physically strikes an inked ribbon against the paper, so the force carries through carbon or carbonless multipart forms to produce several copies in one pass.

• AA laser printer presses a single sheet against a fuser and cannot strike through stacked carbon forms, so it cannot create the simultaneous duplicate copies the office needs.
• BInkjet deposits ink on the top sheet only and applies no striking force, so it cannot transfer an impression through multipart carbon forms to make several copies at once.
• DDirect thermal marks only the single heat-sensitive sheet touching the head and uses no striking impact, so it cannot produce carbon-copy duplicates on multipart business forms.

Direct thermal printing applies heat from the print head to specially coated thermochromic paper that darkens, so it needs no ink, toner, or ribbon, matching the requirement exactly.

• BThermal transfer does use heat but melts a separate wax/resin ribbon onto the media, so it requires a consumable ribbon the warehouse explicitly wanted to avoid.
• CInkjet relies on liquid ink cartridges, directly contradicting the no-ink requirement, so choosing it ignores that direct thermal marks the paper with heat alone.
• DLaser printing consumes toner cartridges and a drum, so it fails the no-toner condition; it also is not designed around heat-sensitive label paper the way direct thermal is.

At its rated page count a LaserJet maintenance/fuser kit replaces the worn fuser and rollers; installing it and resetting the counter restores clean output and prevents jams.

• AA LaserJet has no inkjet print head or nozzles, so this action addresses the wrong technology entirely and cannot resolve fuser-and-roller wear on a laser engine.
• BDeep-cleaning cycles clear inkjet nozzle clogs, which do not exist in a laser printer, so the cycle does nothing for worn fuser and roller parts causing the repeating defects.
• CExtra memory only affects job buffering and spooling, not the mechanical wear of the fuser and feed rollers, so it cannot fix smears tied to end-of-life consumables.

Laser printers use toner, a fine dry powder that is attracted to the drum's latent image and then fused to the page by heat, rather than any liquid ink or ribbon.

• ALiquid ink is the consumable for inkjet printers, which spray droplets; a laser printer instead uses dry powder, so stocking ink would not fit the laser engines at all.
• CInked ribbons are consumed by impact dot-matrix printers whose pins strike them; a laser printer has no striking pins, so ribbons are the wrong supply for this technology.
• DHeat-sensitive paper is the medium for direct thermal printers, not a laser image-forming consumable, so choosing it confuses thermal receipt printing with electrophotographic laser printing.

External SATA (eSATA) extends the native SATA interface outside the case over shielded cables up to 2 meters and is hot pluggable, avoiding the bridge overhead of USB enclosures.

• AUSB 2.0 is a bridged interface that wraps ATA in USB protocol and tops out at 480 Mbps, so it neither extends the native SATA bus nor matches the described unbridged 2-meter link.
• BPS/2 is a legacy low-speed input connector for keyboards and mice, carries no storage data, and cannot connect a hard drive, making it irrelevant to external SATA storage.
• DThe parallel port is a slow legacy interface designed mainly for printers, not modern storage, so it cannot carry a SATA hard-drive signal or meet the stated requirements.

A NIC installs in a PCIe slot and provides the RJ45 port and controller to connect the system to a wired Ethernet network, directly replacing the failed onboard adapter.

• BA sound card only adds audio input/output and has no Ethernet controller or RJ45 port, so it cannot restore the wired network connectivity the user requires.
• CA capture card digitizes incoming video signals for recording or streaming and provides no network interface, so it does nothing to bring back wired Ethernet access.
• DA GPU outputs video to monitors and has no Ethernet port, so installing one addresses display performance, not the failed network connection the user actually needs fixed.

The Trusted Platform Module 2.0 is the hardware-based cryptoprocessor that provides a root of trust for keys and BitLocker, and Windows 11 requires it, so enabling it unblocks the upgrade.

• ASwitching the storage controller to RAID changes how disks are presented and has nothing to do with a security cryptoprocessor, so it will not satisfy the Windows 11 hardware-root-of-trust requirement.
• BSecure Boot validates boot-code signatures but is not the cryptoprocessor that stores keys; while also required, it cannot by itself provide the hardware root of trust the message demands.
• CPage file sizing is an operating-system memory setting unrelated to firmware security hardware, so adjusting it does nothing to resolve a missing trusted cryptoprocessor for the upgrade.

Secure Boot checks that firmware and the OS bootloader are digitally signed and trusted before they run, using signature-enforcement handshakes to block unsigned code and firmware rootkits at startup.

• AAttestation reports recorded measurements after boot but does not itself block an unsigned bootloader from running, so it is a monitoring function rather than the signature gate described.
• CFast Boot only speeds startup by skipping some initialization and performs no signature validation, so it provides no protection against tampered or unsigned boot components.
• DWake-on-LAN is a remote power-on feature unrelated to boot-code integrity, so enabling it does nothing to verify signatures or block rootkits during the boot sequence.

RAID 0 stripes data across the drives for the highest transfer rates and provides no redundancy, so a single drive failure loses everything, exactly matching the editor's performance-first choice.

• BRAID 1 duplicates data for fault tolerance and halves usable capacity, so it prioritizes protection over the maximum striped throughput the editor specifically asked for.
• CRAID 5 adds parity for fault tolerance and needs at least three disks, so it does not match a two-drive, redundancy-free array built purely for speed.
• DRAID 10 requires four drives and mirrors for fault tolerance, exceeding the two-drive, no-redundancy scenario and adding protection the editor explicitly does not require.

RAID 1 mirrors identical data onto a second drive in real time, so if one disk fails all data remains immediately accessible on the other, matching the single-disk-capacity availability need.

• ARAID 0 has no redundancy, so any single drive failure destroys all data; that directly violates the requirement to keep running after one disk fails.
• BRAID 5 tolerates a failure but needs at least three disks and uses striped parity, so it is more complex than a simple two-drive duplicate of a single disk's data.
• DRAID 10 delivers mirroring plus striping but requires four drives, far more than the simple two-disk mirror needed to protect one disk's worth of data.

RAID 10 is a striped set whose members are each mirrored, using four drives to combine the read performance of RAID 0 with the fault tolerance of RAID 1, exactly as described.

• ARAID 0 provides only striping with no mirror, so it cannot deliver the redundancy the database requires and does not describe a four-drive stripe of mirrored pairs.
• BRAID 1 mirrors two drives but adds no striping for performance scaling, so it does not match a four-drive design that combines both mirroring and striping.
• CRAID 5 uses parity rather than full mirroring and is not built as a stripe of mirrored pairs, so it does not match the described four-drive nested mirror-and-stripe layout.

Hot-swap (hot plug) allows removing and inserting a SATA drive while the system is powered on, so a failed array member can be replaced without any downtime.

• AA cold spare can only be installed after shutting the server down, which directly contradicts the requirement to swap the drive while the system stays powered and online.
• CPXE boot is a network startup method for loading an OS image and has nothing to do with physically replacing a drive in a live system, so it does not enable the swap.
• DWake-on-LAN only powers a machine on over the network and provides no ability to remove or insert a drive during operation, so it is unrelated to live drive replacement.

The drum is first uniformly charged, the laser exposes a latent image, toner develops onto it, the image transfers to paper, and heat then fuses it, matching the documented order.

• BThis writes the drum before it is uniformly charged and fuses toner before it ever reaches the paper, which is physically impossible because exposure needs an existing charge and fusing must follow transfer.
• CPlacing developing first would apply toner before any latent image exists on the drum, ignoring that charging and exposing must create the electrostatic image before toner can adhere.
• DThis swaps exposing and developing so toner is applied before the laser writes the latent image, meaning there is no discharged pattern on the drum for toner to cling to.

The charging step uses a primary charge roller or corona wire to place a uniform negative charge across the drum surface, preparing it before the laser selectively discharges points.

• ADeveloping applies toner to the drum only after the latent image has been written, so it occurs later and cannot be the step that lays the initial uniform charge on the bare drum.
• BTransferring relocates developed toner from the drum onto the sheet near the end of the cycle, so it happens long after the drum is charged and is unrelated to conditioning it.
• DFusing melts toner into the paper with heat and pressure at the very end of printing, so it plays no part in placing an electrostatic charge on the imaging drum.

In developing, charged toner is drawn to the laser-discharged regions of the drum, converting the latent electrostatic image into a visible toner image ready for transfer.

• ACharging only lays a uniform electrostatic charge on the drum before exposure and applies no toner, so it cannot be the step where toner forms the visible image on the drum.
• BTransferring moves the already-developed toner image from the drum onto the paper, so it occurs after toner is applied rather than when toner first adheres to the drum.
• CFusing bonds toner to the paper using heat and pressure, so it acts on toner already on the sheet and not on toner being placed onto the drum's latent image.

In transferring, a positive charge applied to the paper from behind pulls the negatively charged toner off the drum and onto the sheet, matching the described mechanism.

• AFusing uses heated rollers to melt toner into the paper after it is already on the sheet, so it does not move toner from the drum onto the paper in the first place.
• CCharging conditions the drum with a uniform negative charge before the image is written, so it occurs early in the cycle and does not involve charging the paper.
• DExposing uses the laser to write the latent image onto the drum, so it forms the image rather than moving developed toner from the drum onto the paper.

Enabling duplex tells the driver to print on both sides of each sheet automatically, halving paper use, which is exactly what the office requested.

• ACollate only controls whether multiple copies emerge as ordered sets, so it affects copy sequencing rather than whether printing occurs on both faces of a sheet of paper.
• BOrientation merely rotates the page layout between portrait and landscape and has no effect on whether the printer uses both sides of each sheet of paper.
• DN-up reduces multiple document pages onto a single side of one sheet to save paper, but it never prints on the reverse side automatically as duplex does.

With collate enabled the printer outputs one full ordered copy at a time (1,2,3,4,5; 1,2,3,4,5), rather than printing all the page ones, then all the page twos.

• ADuplex controls whether printing occurs on both faces of the paper and does nothing to determine whether copies are produced as ordered sets or as grouped identical pages.
• CDraft mode only lowers print density to conserve toner and has no influence on the order in which copied pages are produced or assembled into sets.
• DReverse order merely flips the printing sequence of pages within a single job and does not assemble multiple copies into separate complete ordered sets.

Brightness, expressed as color and white light output in ANSI or ISO lumens, quantifies a projector's total light output, the key spec for performance in a bright room.

• AContrast ratio compares the brightest white to the darkest black the projector can show, describing image depth and richness rather than the total quantity of light emitted.
• BNative resolution states the fixed pixel grid of the panel, describing detail and sharpness of the image, not how bright the projected picture will appear in a lit room.
• CThrow ratio relates projection distance to image width for positioning the unit, so it governs sizing and placement rather than the amount of light output.

Throw ratio is distance divided by image width, so multiplying it by the desired image width gives the exact mounting distance the installer needs to set.

• BLumens quantify the projector's light output for a given ambient condition, so they guide brightness selection rather than calculating how far back to place the unit.
• CAspect ratio is the width-to-height proportion of the picture, such as 16:9, and does not relate projection distance to the resulting image width on the screen.
• DContrast ratio measures the difference between the lightest and darkest output and has no bearing on how far the projector must sit from the screen.

SATA defines a 7-pin signal (data) segment carrying the differential transmit/receive pairs and a separate 15-pin power segment, matching the connectors on the drive.

• ASATA data is not a 4-pin interface; the 4-pin design is the legacy Molex power plug, so this confuses an old power connector with the SATA data link.
• BThis reverses the two segments, assigning the wider 15-pin power connector to data and the narrow 7-pin signal connector to power, which is exactly backwards.
• DNative SATA drives use a 15-pin power connector, not the legacy 4-pin Molex, so pairing the 7-pin data link with a 4-pin power plug misstates the SATA power segment.

The CMOS coin-cell battery powers the RTC RAM that stores BIOS/UEFI configuration and the date and time, keeping them intact while the PC is unplugged.

• AThe standby rail only supplies wake circuitry while the unit remains plugged in, so once the power cord is removed it cannot retain the firmware settings or clock.
• CThe TPM stores cryptographic keys for features like BitLocker and Secure Boot, not the date, time, and hardware configuration parameters that live in CMOS.
• DThe SSD holds user data and its own controller firmware, but it does not store the motherboard's BIOS/UEFI hardware configuration or real-time clock values.

Clearing the RTC RAM erases the saved BIOS configuration, so the system reverts to optimized default settings, undoing the unstable overclock and allowing it to POST.

• AClearing CMOS locks nothing; it erases the stored configuration, after which the boot order simply returns to its default value and can be freely reconfigured again.
• BClearing CMOS resets firmware settings held in RTC RAM but does not touch keys stored in the TPM, so it is unrelated to regenerating cryptographic material.
• CClearing CMOS only erases the saved configuration in RTC RAM; it does not rewrite, update, or restore the BIOS/UEFI firmware image stored in the flash chip.

The front-panel/system panel header (e.g., JFP1 or F_PANEL) routes the chassis power switch, reset switch, power LED, and HDD activity LED to the board.

• BThe CPU fan header supplies power to and PWM-controls the processor cooler, so it does not connect the case's power switch, reset switch, or status indicator LEDs.
• CThe 24-pin connector delivers main power from the PSU to the board and carries no terminals for the chassis buttons or front-panel status LEDs.
• DThe USB header routes the case's front USB ports for data and charging, so it does not carry the power switch, reset switch, or drive-activity LED signals.

Hyper-Threading (Intel) or SMT (AMD) lets each physical core present two logical processors, so six cores appear to the OS as twelve schedulable threads.

• AThe chip physically contains only six cores; threads are logical scheduling units, so no additional physical cores are added or unlocked while the processor is running.
• BThread count does not change clock frequency; simultaneous multithreading shares one core's execution resources between two threads rather than doubling its GHz.
• DLogical threads share the same physical cache; the thread count reflects how many execution contexts a core exposes, not any increase in cache capacity.

Hyper-Threading/SMT gives each physical core two logical processors, so an 8-core CPU presents 16 logical processors to the operating system's scheduler.

• AThese features raise clock speed when thermal and power headroom allow, so they affect how fast cores run but not how many logical processors the OS detects.
• CThe on-die memory controller manages RAM access and bandwidth, so it has no effect on the number of logical processors the operating system enumerates.
• DDual-channel memory widens the data path to RAM for more bandwidth, but it does not create additional logical processors for the scheduler to use.

An integrated GPU inside the processor renders the display and feeds the motherboard's video outputs, letting the system drive a monitor with no add-in graphics card.

• AMainstream CPUs contain no separate PhysX co-processor; the video output here is generated by graphics hardware built directly into the processor itself.
• BModern chipsets do not contain a display engine; on current platforms the graphics and display outputs are integrated into the CPU, not the PCH.
• CA riser only extends or redirects PCIe slots and contains no graphics processor, so it cannot generate the video signal that drives the connected monitor.

Intel 'F' SKUs ship without integrated graphics, so the motherboard's video outputs are inert until a discrete graphics card is installed to drive the display.

• BEven if the port were enabled, it still could not output video because an 'F' processor lacks the integrated graphics required to generate a signal for that port.
• CIntel and AMD processors are never combined in one socket; the 'F' Intel part simply omits its own iGPU and instead needs a separate discrete graphics card.
• DMemory type does not create graphics capability; the 'F' processor has no integrated GPU regardless of whether DDR4 or DDR5 modules are installed.

When the CMOS battery is dead, the RTC RAM loses power off-line, so the clock resets and saved BIOS settings are erased each time the PC is unplugged.

• AA weak PSU usually causes random shutdowns or no-boot conditions, not the specific loss of the clock and saved firmware settings that points to failed backup power.
• BThe date and BIOS settings are lost before the operating system even loads, so an OS fault cannot explain configuration that disappears at the firmware level.
• DToo little RAM affects performance and application stability, but it does not cause the firmware clock and stored BIOS configuration to reset between power cycles.

A short-throw projector has a low throw ratio, so it can cast a large image from a short distance, exactly fitting a room where the unit must sit close to the screen.

• AContrast ratio improves perceived image depth and black levels but does nothing to let a projector produce a large image from such a short distance to the screen.
• CMore lumens raise brightness for ambient light but do not enlarge the image when the projector is forced to sit close to the screen in a small room.
• DHigher native resolution adds pixel detail but does not change the distance-to-width relationship that determines how big the image is at a short throw distance.

ASUS and MSI both direct users to populate DIMM_A2 and DIMM_B2 first for two modules, placing one stick in each channel so the controller enables dual-channel mode.

• APopulating both slots of channel A leaves channel B empty, so the memory controller falls back to single-channel operation and the second channel's bandwidth goes completely unused.
• CLoading only channel B's two slots leaves channel A unpopulated, which keeps the system in single-channel mode and forfeits the doubled bandwidth that correct dual-channel population provides.
• DSlot position absolutely determines channel pairing, so installing two sticks in adjacent same-channel slots can leave the board running single-channel even though both modules are detected.

An RDIMM places a register that buffers command and address signals between the DRAM and controller, reducing electrical load and letting servers scale to higher, more stable memory capacities.

• AA UDIMM connects the DRAM directly to the controller with no register, so it offers less signal stability and cannot scale to the high module counts and capacities large servers require.
• BSO-DIMM only describes the compact physical form factor used in notebooks and small systems and says nothing about registering signals, so it does not provide the buffering the requirement specifies.
• DHigh-clock non-ECC kits target desktop performance and lack both the register that buffers signals and the error correction that servers rely on for stability at very large memory capacities.

NVMe runs over the PCIe interface for low latency and scalable performance, making PCIe NVMe SSDs roughly six to seven times faster than a SATA SSD on the same workload.

• BThe SATA interface tops out at 6Gb/s, a ceiling reached back in 2009, so it cannot match the latency or throughput of a drive attached directly to the faster PCIe bus.
• CA mechanical SATA hard drive is limited by both the SATA bus and spinning-platter mechanics, so its latency and throughput fall far below any solid-state drive on the PCIe interface.
• DA USB thumb drive uses slow flash on a peripheral bus, so it cannot serve as a low-latency, high-throughput boot and scratch disk the way a PCIe-attached NVMe SSD can.

ATX 3.0 power supplies provide the 12VHPWR/12V-2x6 connector that delivers up to 600 W to a single add-in card, matching the flagship GPU's 16-pin input directly.

• AAn 80 PLUS rating measures energy-conversion efficiency, not which connectors a PSU provides, so a Bronze label gives no assurance the unit includes the required 16-pin GPU connector.
• BTwo 6-pin connectors supply only 75 W each, far short of 600 W, and an older ATX 2.x supply lacks the native 16-pin cable this high-power graphics card requires.
• CThe 8-pin EPS12V connector feeds the CPU's voltage regulators, not the graphics card, so using it for the GPU is both insufficient and electrically inappropriate for the load.

Inkjet printers excel at high-quality color images and photos, blending inks for smooth gradients and supporting glossy photo paper and borderless printing for exactly this use case.

• AA mono laser prints only black toner and is optimized for crisp text, so it cannot reproduce the wide color gamut and smooth photographic gradients the user wants on glossy paper.
• BA thermal printer marks heat-sensitive label stock in a single color for shipping labels and receipts, so it is wholly unsuited to producing full-color photographic images on glossy photo paper.
• DA dot-matrix printer strikes an inked ribbon through a pin head for multipart forms, producing coarse low-resolution output that cannot render detailed full-color photographs at all.

Monochrome laser printers provide superior text quality at a lower cost per page and print 20-40 pages per minute, making them ideal for high-volume office document output.

• AA photo inkjet is built for color images and prints relatively slowly with a higher cost per page from ink cartridges, making it a poor fit for high-volume black text printing.
• CA thermal receipt printer only marks narrow heat-sensitive rolls for point-of-sale slips, so it cannot produce the standard full-page office documents the workgroup needs in volume.
• DA plotter is designed for large engineering and architectural drawings on oversized media, so it is the wrong tool and far too costly for everyday letter-size text printing.

For CPUs pushed to the limit with overclocking, an AIO liquid cooler is the better option, moving heat to a radiator whose size and fans set its cooling capacity.

• BA small bundled cooler is sized for stock thermal output, so it lacks the radiator capacity to dissipate the much greater heat a heavily overclocked, power-hungry CPU produces under load.
• CA passive heatsink relies on natural convection alone and cannot shed the high sustained heat of an overclocked CPU, so the processor would quickly thermal-throttle under heavy load.
• DA case fan only moves ambient air through the chassis and does not draw heat off the CPU die, so on its own it cannot replace a dedicated CPU cooler for an overclocked chip.

An M.2 2280 'gumstick' SSD seats flat in the board's M.2 slot held by one screw, carrying power and data through the slot with no separate cables.

• AA 3.5-inch HDD is a large mechanical drive that needs a drive bay plus separate SATA data and power cables, so it cannot seat directly on the motherboard in a cramped mini-ITX case.
• BA 2.5-inch SSD still requires a mounting point and two cables, a SATA data lead and a power lead, so it does not satisfy the cable-free, board-mounted requirement of this build.
• CAn external USB SSD sits outside the chassis and connects through a peripheral port, so it is not an internal, board-mounted drive and adds overhead compared with a direct slot connection.