Computer Hardware Basics
Identify the components inside a computer and explain how they cooperate to process, store, and move data.
- Certification
- CompTIA A+
- Recommended study time
- 5h 35m
- Status
- Not started
Recommended study time
About 5h 35m in total, measured from the material on this page. At your session length of 45 minutes that is 8 sittings.
- Read the lesson21 min
About 2,782 words at a careful technical reading pace.
- Second pass with notes13 min
Re-read the harder parts and write your own notes.
- Work through the examples1h
3 worked examples and 9 practice questions.
- Recall from memory12 min
2 written recall questions.
- Practice decision12 min
One applied decision with feedback.
- Teach it back20 min
Write the topic in your own words.
- Real-world scenario15 min
Read the situation and justify your decision in writing.
- Hands-on practice2h
Labs, commands and configuration until you can do it unaided.
- Spaced review1h
4 short review sessions spread over the following weeks.
Learning objectives
- Identify the purpose of the CPU, RAM, motherboard, storage, power supply, and cooling system.
- Distinguish temporary working memory from persistent storage.
- Trace the basic path data follows while a program runs.
Start here
About 18 minutes of reading, in 8 short parts.
A computer is a small number of cooperating parts, each solving one problem: calculate, remember, store, connect, power, and cool. Once you can name each part and say what it does, most hardware faults stop being mysterious and become a question of which part is not doing its job.
Where you meet it: Every desk-side repair, every laptop that will not power on, every 'my computer is slow' ticket, and every hardware purchase decision.
The lesson, part by part
Open one part at a time. Each part stands on its own, so you can stop and come back.
Computer hardware is the physical machinery that executes software. The processor performs arithmetic and logic, memory holds what is currently being worked on, storage keeps data when the power is off, the motherboard wires everything together, the power supply converts wall electricity into the low voltages chips need, and the cooling system removes the heat that all of this produces.
The single most useful idea in hardware is the memory hierarchy. Data closest to the CPU is fastest and smallest: CPU registers, then cache (L1, L2, L3), then RAM, then an SSD or hard disk. Each step down is roughly an order of magnitude slower and an order of magnitude larger. Performance problems are usually a question of data sitting too far down that hierarchy for the work being asked of it.
Key ideas
If you remember nothing else from this topic, remember these.
- Every hardware fault eventually reduces to one of five things: power, connection, heat, capacity, or a failing component, and a methodical technician checks them in that order rather than guessing at parts to replace.
- The memory hierarchy is the single most predictive model in hardware: registers, then cache, then RAM, then SSD or HDD, each roughly an order of magnitude slower and larger than the one before it, so a slow machine is almost always working with data that is too far down that ladder.
- Clock speed and core count answer different questions: clock speed is how fast one lane of traffic moves, core count is how many lanes exist, and a workload that cannot be split across threads will never benefit from more cores no matter how many you add.
- A power supply is rated in watts but fails under load, not at idle, which is why intermittent shutdowns during gaming or heavy compiling point at the PSU well before they point at the CPU.
- SMART attributes on a drive, particularly reallocated sector count and pending sector count, give an early warning of mechanical or flash failure well before a drive stops mounting entirely.
- Thermal throttling is a designed safety behaviour, not a fault: a modern CPU deliberately slows itself down to survive rather than burning out, so 'the computer got slow under load' is a cooling symptom far more often than a broken-part symptom.
Diagnosing a workstation that shuts down under load
A worked example, step by step.
A designer's workstation powers off without warning whenever they render video, but runs fine for email and browsing. The user has already tried restarting it twice.
- 01Reproduce and recordStart a rendering job and watch Task Manager or htop. The shutdown happens after roughly 6 minutes, right after CPU usage sits above 90 percent for several minutes.
- 02Check thermal dataReboot, open the UEFI firmware or a monitoring tool, and read CPU temperature at idle: 42 C, which is normal, ruling out a dead fan at rest.
- 03Reproduce with monitoring openRun the render again with HWInfo open. Temperature climbs past 100 C in under two minutes and the system cuts power exactly at that point, well above the CPU's safe limit near 95 to 100 C.
- 04Inspect physicallyPower down, unplug the mains cable, discharge static by touching the case, and open the side panel. The CPU heatsink is packed with dust and the fan spins freely but moves almost no air.
- 05Clean and inspect the interfaceCompressed air clears the heatsink fins. Removing the cooler shows the thermal paste has dried into a cracked, chalky layer instead of a thin even film.
- 06Repaste and reseatClean both surfaces with isopropyl alcohol, apply a pea-sized dot of fresh thermal paste, and reseat the cooler with even, cross-pattern screw pressure.
- 07Re-run the same testRepeat the exact render job. Temperature now peaks at 78 C under sustained full load and holds steady.
- 08Verify and documentRun the render twice more over 30 minutes with no shutdown, then note in the ticket the dust buildup, the dried paste, the before and after temperatures, and the parts used.
Outcome: The shutdown was thermal protection triggered by a dust-blocked heatsink and degraded thermal paste, not a failing CPU or PSU; cleaning and repasting restored safe operating temperatures and resolved the fault permanently.
Hardware Facts Worth Memorising
Worth keeping at hand while you work.
- SATA III
- 6 Gbps theoretical, roughly 550 MB/s real-world SSD throughput
- NVMe over PCIe 4.0
- Up to roughly 7000 MB/s sequential, using the PCIe lanes directly instead of the SATA controller
- DDR4 vs DDR5
- Different notch position and voltage; physically cannot be swapped between motherboards
- SO-DIMM
- Laptop memory form factor, physically smaller than a desktop DIMM
- 80 PLUS Bronze/Gold/Platinum
- PSU efficiency rating at typical load; higher tier wastes less power as heat
- ATX / microATX / Mini-ITX
- Motherboard form factors from largest to smallest, each needing a matching case
- POST
- Power-On Self-Test; firmware checks CPU, RAM, and video before handing off to the bootloader
- SMART reallocated sector count
- Non-zero and rising means the drive is actively failing
- M.2 key types
- B-key and M-key define which slot a given M.2 SSD or Wi-Fi card physically fits
- TDP
- Thermal design power in watts; the rough heat output a cooler must dissipate
- ESD wrist strap
- Grounds the technician to prevent static discharge from damaging chips
- CMOS battery (CR2032)
- Keeps firmware settings and the clock when the machine is unplugged; dead battery causes wrong date and lost boot order
Common misunderstandings
What most beginners get wrong here.
More RAM always makes a computer faster.
RAM only helps once it was the bottleneck; a CPU-bound or storage-bound machine sees no improvement from added memory.
A bigger power supply wattage is always better.
An oversized PSU wastes efficiency at low load and adds cost with no benefit; the PSU should be sized to the actual component draw plus headroom, not maximised.
If the fans are spinning, cooling is fine.
A spinning fan can still be a dust-blocked heatsink moving almost no useful air, or dried thermal paste preventing heat transfer in the first place.
A blue screen means the hard drive is failing.
Blue screens are more often caused by faulty RAM or a bad driver; a memory test and driver review should come before replacing storage.
Any USB-C port can charge and transfer data at full speed.
USB-C is a connector shape only; the actual protocol, power delivery wattage, and speed depend on what standard the port implements underneath.
Exam traps
How the question writers try to catch you out.
- A scenario describes random reboots under heavy load only, not at idle: this points to power delivery or thermal issues, not RAM, which the exam expects you to isolate by the word 'under load.'
- Questions naming SATA versus NVMe expect you to know NVMe uses the PCIe bus directly and bypasses the older AHCI/SATA controller bottleneck.
- A question describing 'no display, but fans spin and drive light is on' is testing POST failure diagnosis: reseat RAM and GPU before assuming a dead board.
- CompTIA expects DDR generations described as non-interchangeable due to physical notch and voltage differences, not just 'faster or slower.'
- Questions about ESD precautions expect the wrist strap or self-grounding step before touching any component, every time, not just for 'sensitive' parts.
Check yourself
Answer in your head first, then reveal. This is not scored.
A machine is slow only when many applications are open at once, and Task Manager shows high memory usage with low CPU usage. What is the likely fix?
Why does replacing a spinning hard disk with an SSD produce such a large perceived speed improvement on old hardware?
A user's laptop won't power on at all, but it was working yesterday. What are the first two things to check?
Why should you strip a system to minimum hardware when troubleshooting an intermittent fault?
What does a rising SMART reallocated sector count indicate, and what should you do?
Quick reference
A condensed summary of the lesson above, for revision.
What It Is
Computer hardware is the physical equipment that runs software. The motherboard connects components; the CPU executes instructions; RAM holds active data; storage retains files when power is off; the power supply converts incoming electricity; and cooling removes heat. Input, output, and network devices let the system interact with people and other computers.
Why It Matters
Support technicians must translate symptoms into likely component failures. A computer that powers off under load may have heat or power trouble, while a machine that slows when many applications open may lack RAM. Hardware knowledge also prevents unsafe upgrades: components must fit the motherboard, power budget, physical case, and intended workload.
How It Works
- The CPU fetches instructions and operates on data held in RAM.
- The motherboard buses connect processing, memory, storage, and peripheral controllers.
- The PSU supplies regulated power while cooling moves heat away from components.
Where You See It
- Desktop workstations, laptops, servers, point-of-sale terminals, and repair benches.
Key Terms
- CPU
- The processor that fetches and executes program instructions.
- RAM
- Fast, volatile memory used by programs that are currently running.
- Storage
- Persistent media, commonly an SSD or HDD, that retains data without power.
- Motherboard
- The main circuit board that connects and coordinates system components.
- PSU
- The power supply unit that converts wall power into regulated voltages.
Examples
- Opening a browser loads program files from an SSD into RAM, then the CPU executes its instructions.
- Installing a graphics card requires a compatible slot, enough case space, adequate cooling, and sufficient PSU capacity.
Common Problems
- Loose power or data cables
- Insufficient RAM under workload
- Thermal throttling from blocked airflow
How It Fails
- A failed PSU may produce no power or sudden shutdowns.
- Faulty RAM may cause crashes or corrupted calculations.
- A failing drive may become slow, report errors, or stop booting.
How to Troubleshoot
- Confirm the symptom and recent changes.
- Check power, cables, indicators, and temperatures.
- Test one suspected component at a time and document the result.
Practical Knowledge
- Match RAM generation and form factor to the motherboard.
- Use antistatic precautions and disconnect power before internal work.
Exam Coverage
- Component purposes and compatibility
- Storage, memory, power, and cooling symptoms
- Safe hardware procedures
Interview Questions
- How would you distinguish a RAM shortage from a storage-capacity problem?
- What would you check first when a desktop has no signs of power?
Worked examples
Each calculation is shown one step at a time, then you try it yourself before revealing the answer.
Convert binary to decimal
Convert the 8-bit binary number 11000000 to decimal.
- 1. Write the place valuesAn 8-bit number has fixed place values, left to right: 128, 64, 32, 16, 8, 4, 2, 1.
- 2. Line the bits up under the place values128→1, 64→1, 32→0, 16→0, 8→0, 4→0, 2→0, 1→0.
- 3. Keep only the place values above a 1The 128 column and the 64 column hold a 1. Every other column holds 0 and contributes nothing.
- 4. Add them128 + 64 = 192.
Answer: 11000000 = 192. This is why a /24 subnet mask octet of 11000000 reads as 192 in dotted decimal.
Now you try
Convert 10101010 to decimal.
Convert 11111111 to decimal.
Convert 00011100 to decimal.
Convert binary to hexadecimal
Convert 11011110 to hexadecimal (the form used by MAC and IPv6 addresses).
- 1. Split into 4-bit groups11011110 becomes 1101 and 1110.
- 2. Convert each group with place values 8, 4, 2, 11101 = 8 + 4 + 1 = 13. 1110 = 8 + 4 + 2 = 14.
- 3. Replace values over 9 with letters10=A, 11=B, 12=C, 13=D, 14=E, 15=F. So 13 = D and 14 = E.
- 4. Join the digitsD followed by E.
Answer: 11011110 = DE in hex, which is 222 in decimal. Each pair of hex digits is exactly one byte, which is why a MAC address is six hex pairs.
Now you try
Convert 10101111 to hex.
Convert hex 3C to binary.
Convert hex FF to decimal.
Explain a 'missing' drive capacity
A customer buys a 1 TB drive and Windows reports 931 GB. Has the shop cheated them?
- 1. Manufacturer unitsDrive makers use decimal units: 1 TB = 1,000,000,000,000 bytes.
- 2. Operating system unitsWindows reports binary units but labels them GB: 1 GiB = 1,073,741,824 bytes.
- 3. Do the division1,000,000,000,000 ÷ 1,073,741,824 ≈ 931.
- 4. Account for formattingThe filesystem and recovery partitions then consume a further few gigabytes.
Answer: No. The bytes are all there; the two sides count in different units. Say this to the customer in plain terms rather than quoting powers of two.
Now you try
How much usable space would a 2 TB drive report?
How many bytes in 1 MiB?
Why does a phone advertise 128 GB but show around 112 GB?
Watch and read
Verified official and reputable sources for this topic. Links open in a new tab.
Video training
Reading and courses
Lesson notes and bookmark
Notes and bookmarks for this lesson, saved with everything else you have marked.
No notes on this item yet.
Learning progress
0% across six evidence areas. Reading alone does not change progress.
Prerequisites
None. This topic starts from first principles.
Next steps
- 01Locate each major component in a desktop or teardown diagram.
- 02Compare the installed RAM, CPU, and storage reported by an operating system.