GPU Left Unused for Months? Black Screen or Artifacts on First Power-Up — Long-Term Storage Revival Checklist

| HK GPU Fix

GPU Left Unused for Months? Black Screen or Artifacts on First Power-Up — Long-Term Storage Revival Checklist

Key Takeaways

A graphics card that worked before months in storage can show a black screen or artifacts on its first power-up. Storage can change contact surfaces, capacitor behaviour, fan lubrication and dried thermal paste. This checklist explains the most likely causes, the safe order to inspect and clean the card, and when inspection is the next step.

The Premise: It Worked Before Storage

This guide has a strict premise, and it matters. The card was working normally before it was put away. It then sat unused, without power, for several months or more. On its first power-up after that storage, it shows a black screen, no signal, artifacts, intermittent failure to be detected, or normal behaviour at idle followed by a crash under load.

That contrast — normal before storage, faulty on first use after storage — is itself a diagnostic signal. The most likely cause is something that changed during storage, not a new hardware fault that appeared out of nowhere.

Cases outside this guide: a card that already had a fault before it was stored; a fault that appeared after moving the computer; a fault that appeared after changing another part; and a fault in a machine that was in regular use during humid weather, where storage is only a side issue rather than the trigger.

Mechanism 1: The Contact Surfaces Changed While the Card Sat Idle

A graphics card relies on its gold fingers making contact with the spring contacts inside the PCIe slot. Gold itself does not rust easily, but the gold plating is very thin and can have microscopic pores. Moisture, salt, sulphides and oil from handling can pass through those pores and attack the nickel layer underneath, leaving a dull patch or a green or white deposit on the surface and raising the contact resistance.

With no power and no insertion or removal to provide friction, that oxide or contamination film can form quietly during storage. On the first power-up, the unstable contact can make the motherboard intermittently fail to detect the card, or make the screen go black partway through booting.

The distinguishing feature is intermittency. An oxidation fault is usually intermittent — sometimes the card is detected, sometimes it is not — and reseating the card may make it work again temporarily. That is different from a persistent failure to detect the card, and different again from a hidden PCB fracture, where the fault is triggered by temperature or vibration rather than by contact resistance. A single photograph of a dull patch on a gold finger is not enough, on its own, to convict the contact surface; the judgement needs the humidity timeline, repeated symptoms and both contact surfaces looked at together.

Mechanism 2: Aluminium Electrolytic Capacitors Lose Their Oxide Layer Without Power

This is the mechanism most readers will not have heard of, and it deserves the most careful explanation in this guide.

An aluminium electrolytic capacitor works because of an extremely thin layer of aluminium oxide acting as its dielectric. That oxide layer is formed by applying voltage, and it is not permanent. Without voltage applied, the electrolyte slowly dissolves the oxide layer back again, making the dielectric thinner and raising the leakage current.

Capacitor manufacturers' reference material gives a shelf life of roughly two to three years for aluminium electrolytic capacitors, and recommends reforming — restoring the oxide layer by raising the voltage slowly — before use after that period. Guidance published by ABB and summarised in technical discussion notes that storage of less than one year needs no reforming, storage of one to two years calls for about an hour at rated voltage, and storage beyond two years needs reforming, for example by raising the voltage gradually from a quarter of rated voltage. Another reference notes that leakage current in the first few minutes after voltage is applied following storage can be up to 100 times its normal value.

Two boundaries have to be kept clear. First, most capacitors on a graphics card are solid-state or ceramic types; aluminium electrolytic capacitors are only part of the mix. Capacitor ageing is one possible mechanism in this guide, not a verdict that can be applied to every card that has been stored for a long time. Second, reforming is a professional repair operation involving gradual voltage increases. It is not something a reader should attempt on a graphics card with a variable transformer or by altering voltages. This guide explains the principle and the practical point — tell the repair workshop how long the card was stored — and leaves the repair judgement to professional inspection. Capacitor failure with bulging or leaking is covered in the separate guide on capacitor faults, and the power-delivery stage is covered in the VRM guide. No capacitor lifespan figure is used here as a diagnostic standard; the two-to-three-year shelf life is capacitor manufacturers' general reference material, not a statement that a graphics card will fail after two years in storage.

Mechanism 3: Fan Bearing Lubricant and Dust

The lubricant inside a fan bearing can solidify or distribute unevenly when the fan has not turned for a long time. On first power-up, the fan may start slowly, struggle to start, or make an unusual noise, then run normally again. That is a first-start problem, and it is different from the fan-noise faults of a card in regular use, which are covered in the separate fan-noise guide.

Dust is the other half of this mechanism. Dust that settles during storage absorbs moisture, and damp dust can form a contamination film on contact surfaces and can also reduce cooling performance. The cleaning method is covered in the separate dust-cleaning guide and is not repeated here.

Mechanism 4: Thermal Paste and Thermal Pads Dry Out

Thermal paste and thermal pads dry out over time even when they are not used. On a card that has been stored for years, the first run under load may show higher core or hotspot temperatures than before. That is a supporting factor, not the main story of this guide. Replacement intervals are covered in the separate guides on thermal paste and on thermal pads, and are not repeated here. A card that has been stored for a long time cannot be reduced to a single cause of "the paste dried out".

Mechanism 5: The Motherboard CMOS Battery

A whole computer that has been stored for a long time may have a flat motherboard CMOS battery, which resets the BIOS settings — including PCIe settings, CSM and the boot order. The symptoms can look like a graphics card fault: no picture on boot, or settings that have changed. The root cause is the motherboard battery, not the card.

This belongs in the diagnostic flow as a separate branch: check whether the BIOS settings have been reset, and do not conclude that the graphics card is faulty on that evidence alone. BIOS settings resetting after a BIOS update, and after a CPU change, are covered in their own guides; the mechanism here belongs to the same family.

Storage changeWhat happensHow it usually shows
Contact surface oxidation or contamination filmContact resistance rises on the gold fingers and slot contactsIntermittent detection, black screen partway through boot, sometimes normal after reseating
Aluminium electrolytic capacitor oxide layer thinningDielectric thins and leakage current rises after long storage without powerInstability in the first minutes after power-up; reforming is a professional repair operation, not a home procedure
Fan bearing lubricant solidifyingFan starts slowly or noisily on first spinSlow start, unusual noise on first power-up, then normal running
Thermal paste and pads drying outHeat transfer worsens over time even without useHigher core or hotspot temperature on first load, compared with before storage
Flat motherboard CMOS batteryBIOS settings reset, including PCIe settingsNo picture on boot or changed settings, with the card itself healthy

The Revival Checklist: Follow This Order

Work through these steps in order. Do not skip ahead to a heavy load test, and do not keep taking the card apart and refitting it if the symptoms persist.

1. Look before powering on. Inspect the gold fingers for dull patches or green or white deposits, look inside the slot for dust or foreign objects, check the capacitors for bulging or leaking, and flick the fan blades gently by hand to see whether they turn freely. Do not force them. A single dull patch on its own does not prove the contact surface is at fault; the judgement needs the humidity timeline, repeated symptoms and both contact surfaces examined together.

2. Clean the contact surfaces gently. Clean the gold fingers using the correct method, covered in the separate gold-finger cleaning guide. The gold plating is very thin, so clean lightly and do not grind the plating away. Abrasive methods such as sandpaper or steel wool must not be used; they remove the plating and make the contact worse. After cleaning, refit the card firmly and make sure the retaining clip is properly latched.

3. Test with a minimum system. Keep only the hardware needed to boot, and see whether the card is detected. The detection flow is covered in the separate guide on a card not being detected. If another computer or another slot detects the card, the problem narrows to the original slot or the contact surfaces. If no computer and no slot detects it, only then does the card itself become the main suspect.

4. After the first power-up, do not run a heavy load immediately. Let the card idle and observe it first: watch whether the fan starts, and watch the temperature readings. Only when it is stable should the load be increased gradually. This is because capacitor leakage current is highest in the first few minutes after voltage is applied following storage — not because the card needs to be "run in".

5. If the symptoms persist, send the card for inspection. If cleaning, reseating and cross-testing still leave a black screen or artifacts, do not keep dismantling the card. When sending it for inspection, state how long it was stored, where it was stored — damp or dry — and that it was working normally before storage. That information is useful for the repair judgement. Inspection fees follow the published three-tier structure of HK$400–2,800, with the actual quote given after inspection.

Storage Environment Decides the Risk

A card stored in a dry, ventilated place and a card stored in a damp corner do not carry the same risk. A machine that is not used for a long time should not be left in a damp corner. Do not wrap a computer in cling film, and do not stuff desiccant into the case — both are discouraged in the separate humidity guide, and this guide does not repeat that material as a damp-proofing guide.

A long storage period is not, by itself, a reason to expect failure. Many cards sit unused for months and work perfectly when they are used again. This guide must not be read as a warning that a card stored for a few months will fail. In many cases, cleaning and reseating the card resolves the problem — a possibility, not a guarantee.

Telling Artifact Patterns Apart

If the artifacts have a structured pattern — coloured blocks, a checkerboard, or horizontal and vertical lines — a VRAM or core fault is more likely, and the separate guide on artifact patterns is the right next step. If the symptom is an intermittent black screen, or the card intermittently not being detected, a contact-surface fault is more likely. The two patterns should not be mixed together.

The same applies to second-hand cards described as new stock that has been sitting unopened for years. The same storage logic applies, and the inspection checklist for a second-hand card is covered in its own guide.

Common Misconceptions

"A card that has been stored for a long time is broken." Wrong. In many cases the fault is only contact-surface oxidation, and cleaning and reseating resolves it.

"The gold fingers have rusted." Wrong. Gold does not rust easily. The accurate description is a surface oxide or contamination film, with the nickel layer underneath attacked through microscopic pores in the plating.

"Sand the rust off and it will be fine." Wrong. The gold plating is very thin, and sanding it away makes the contact worse. Only gentle cleaning is appropriate.

"Run a heavy stress test as soon as it powers on, to see whether it still works." Wrong. Leakage current is highest in the first few minutes after storage, so the card should idle and be observed first, with load increased gradually only after it is stable.

"A capacitor that has been stored for a long time will definitely leak or bulge." Wrong. Leaking and bulging are visible late-stage states. Thinning of the oxide layer is an earlier change that cannot be seen with the naked eye, and the two must not be treated as the same thing.

"Replacing the CMOS battery fixes the graphics card." Wrong. The CMOS battery is only one branch of the diagnosis. After replacing the battery, the BIOS settings still have to be checked to confirm they are back to the correct values.

"Blow-drying it or leaving it in the sun will revive a damp card." Wrong. High heat and direct sunlight can damage components. The correct approach is visual inspection, cleaning and cross-testing — not heating.

When to Seek Inspection

Seek professional inspection if cleaning and reseating do not resolve the black screen or artifacts, if the card is not detected in any computer or slot, if artifacts have a structured pattern suggesting a VRAM or core fault, or if any capacitor shows bulging or leaking. Bring the card as it is, and note down how long it was stored, where it was stored, and that it was working normally before storage. That record shortens the inspection considerably. Inspection fees follow the published three-tier structure of HK$400–2,800, with the actual quote given after inspection. The workshop is at Room 1209, Dali Plaza, Shek Mun, Sha Tin. Telephone 2186 8010, WhatsApp 6558 6806. There is no on-site service.

FAQ

My card worked before I stored it for months. Now it shows a black screen. Is it dead?

Not necessarily. Storage can change the contact surfaces, capacitor behaviour, fan lubrication and dried thermal paste. In many cases the cause is contact-surface oxidation, and cleaning and reseating the card resolves it. Work through the checklist in order before concluding the card is faulty.

What does oxidation on the gold fingers actually mean?

Gold itself does not rust easily. Moisture, salt, sulphides and oil can pass through microscopic pores in the thin gold plating and attack the nickel layer underneath, leaving a dull patch or a green or white deposit that raises contact resistance.

Can I sand the gold fingers to clean them?

No. The gold plating is very thin, and abrasive cleaning removes the plating and makes the contact worse. Clean the gold fingers gently, without grinding the plating away.

Should I run a heavy stress test as soon as the card powers on after storage?

No. Capacitor leakage current is highest in the first few minutes after voltage is applied following storage. Let the card idle and observe the fan and temperature readings first, then increase the load gradually only after it is stable.

Do capacitors on a graphics card go bad just from sitting unused?

Aluminium electrolytic capacitors rely on a thin oxide layer that slowly dissolves without applied voltage. Manufacturers' reference material gives a shelf life of roughly two to three years for that capacitor type. That is a general reference for aluminium electrolytic capacitors, not a statement that a graphics card will fail after two years in storage, and most capacitors on a graphics card are solid-state or ceramic types. Reforming is a professional repair operation and must not be attempted at home.

My computer was stored for a long time and now shows no picture. Could it be the motherboard battery?

Yes. A flat CMOS battery can reset BIOS settings, including PCIe settings, which can look like a graphics card fault. Check whether the BIOS settings have been reset before concluding the card is faulty, and check the settings are back to the correct values after replacing the battery.

How do I tell a contact fault apart from a VRAM fault?

An intermittent black screen, or the card intermittently not being detected, points more towards a contact-surface fault. Structured artifacts — coloured blocks, a checkerboard, or horizontal and vertical lines — point more towards a VRAM or core fault. The two patterns need different diagnostic paths.

What information should I give the workshop if I send the card for inspection?

State how long the card was stored, where it was stored — damp or dry — and that it was working normally before storage. That information is useful for the repair judgement.

Related articles

Related services

If you need Power connector repair, see these service pages: