Briefly, the main important nuances

Before we move on to the question “When is it necessary to replace a GPU in a laptop?”, let’s first clarify some important points about what a video card is and how it works in a modern gaming laptop.

The definition of a gaming laptop includes any Razer, Acer Predator, Acer Nitro 5, Asus ROG, Asus TUF, Dell Alienware, HP Omen, Lenovo Legion, MSI Stealth, or Apache laptop with a full-fledged H-processor (for example, i7-8750H or i5-10300H), as well as Ryzen H or U series processors.

☝	Gaming laptops most often use a combination of integrated (built into the processor) and discrete (i.e., separate powerful GPU) graphics cards.

Almost always the laptop “starts” on the built-in video (since it consumes less energy), then the discrete video chip is briefly powered on so that the processor can ‘poll’ and “see” that there is a discrete GPU in the system, and then the power to the discrete video chip is turned off again.

☝	That is why such systems often do not even need to be reconfigured to work with integrated video (it will work anyway), but with an integrated video card in a gaming laptop processor, you will get a kind of “vegetable” that does not provide the necessary video system performance at all.

This architecture means that if there is a problem with the discrete video chip or its power supply during the timeout allocated for its detection, or if the power supply has not started and it has not been detected on the bus, the processor “decides” that the system does not have a discrete video chip and continues to operate without it.
Therefore, there are a number of reasons that can lead to the following:

❗	The video card is not listed in the device list.

to the beginning of the article
Usually, you don’t notice this right away, because the system may boot up as usual.
But after launching programs that load graphics, it becomes clear that something is wrong.
The frame rate drops sharply, and in tests, the “graphics renderer” column only shows Intel UHD Graphics.
It’s easy to verify the problem by simply going to Device Manager (left photo).

And instead of the usual picture (photo on the right), you will see that only Intel UHD Graphics remains in the list of video adapters (photo in the center).

☝	This should be done with any application that uses graphics enabled, because in some laptops, the video card power source is only activated in this case (necessary for the system to “see” GPU).

Reinstalling Windows and/or drivers does not help, and the “UMA only” option (built-in graphics only) is not enabled in the BIOS settings.

There are several reasons for this behavior of the video card, namely:

✅

CAUSE: One of the video chip power sources is not starting up and needs to be repaired. This could be either a malfunction of the power controller itself or a short circuit in its circuits (caused by an external object, something burnt out in the load or wiring). REQUIRES: First of all, carefully diagnose the problem to determine the reason why the source is not starting.

Don’t be surprised, there are several of them in the GPU system, and there are those without which the GPU will not appear at the allotted timeout when the system starts.
If there is no “response,” the system will decide that it is designed to work on the integrated video core in the processor, and discrete graphics will be off of the table, no matter what you do.
Unfortunately, we have heard many times in phone calls in such cases that the GPU in the laptop needs to be replaced, although in reality, with this symptom, it is not often necessary.
There is a more serious modification of this problem:

✅

CAUSE: The video card’s power supply is “blown.” This is a very worrying reason, because on some models, video memory power sources often break through the “upper” key from 19V to 1.35V. REQUIRES: In this case, not only almost all components of this power supply must be replaced, but also, for example, all video memory and GPU (if we are talking about a video memory power supply) if they are also “blown.”

To understand how bad this is, imagine a situation where an incompetent electrician mixes up the wiring in a house and supplies 380V to the sockets instead of 220V.
People who were “lucky enough” to encounter such a situation had to replace all the household appliances in their apartment.

☝

So, if system don’t see a discrete GPU GTX 1050, GTX 1060, GTX 1070, GTX 1080, GTX 1660, RTX 2060, RTX 2070, RTX 3070, RTX 3080, RTX 4060, RTX 4070, RTX 4080 in your system, don’t immediately assume that you need laptop GPU replacement. It may be caused by problems with its power supply. But don’t forget, there are worse scenarios where very expensive repairs may be required, involving the replacement of various parts, including the GPU, unfortunately.

In fact, we’ve seen this happen with Acer Predator Helios 300 laptops, which have a common issue where when overloaded and overheated in one of the video memory power supply phases, a key (transistor) blowout occurs and 19V power gets into the 1.35/1.5V power supply circuits.

In this circumstance, the 19V power supply will naturally begin to malfunction (break down), causing damage to the entire video memory (all memory modules) and its controller located in the graphics chip.
A GPU that is shorted in terms of video memory power supply may either disappear from the system or start to glitch when the video is loaded or even without load.

☝

It is important to understand, the procedure for figuring out whether the video chip has been damaged by a 19V power supply failure and/or video memory, and finding out how many memory strips have been damaged, is a well-established and logically structured procedure, rather than blindly soldering everything in a random order. If you start replacing everything in a row or start up power supplies in random order after replacement, you can repeatedly “burn” entire sets of memory and new video chips one after another. Or, for example, you can first “burn out” half of the working memory on the customer’s board without first properly repairing the power supply, thereby increasing the cost of repairs by a considerable amount for no reason.

Another problem that is also common in gaming laptops is

❗	Code 43. The device cannot be started.

to the beginning of the article
This symptom is even more frequent.
This error indicates that although the video chip was able to start up and initialize in the system, it detected malfunctions in its operation, which it reports with this error.

This is how it looks in real life on a laptop:

Since a video system is a complex unit consisting of several modules, problems can vary.
Let’s start with the most frequent ones:

✅

ПРИЧИНА: The video chip detected problems when scanning the video memory. Interestingly, it is not necessary to immediately suspect the memory, although it can also malfunction. It may not only be faulty video memory, but also a problem with the contact pads of the video memory chip(s) or GPU with the board, or a problem with the pads on the board itself (which is quite rare). REQUIRES: First and foremost, a very careful diagnosis is needed to determine the cause.

It should be understood that, first of all, such an error can occur even with a problem with a single signal line.

This could be a problem with the solder ball contact under the video memory chip or video chip, or with the contact with the board.

That means, to deal with this issue, you need to clearly understand how to pinpoint the exact location of the malfunction, at least up to the video memory module.

And then the fun part starts.

You have to take it apart and solder it, because in the circuit “memory module – contact – solder ball – contact – track on the board – contact – solder ball – contact – video chip,” any link can cause such a problem.

☝

That’s why in these situations, it’s all about the experience, intuition, and steady hands of the engineer who’s working on your laptop. Because there’s often a lot of soldering to do, and someone who’s good at soldering can solder a lot and eventually find the faulty part and fix the laptop. But someone who doesn’t know how to solder can consistently increase the number of problems and eventually bring them to a confident “now only replacing the motherboard will help you, although at first, it may not even have been necessary to replace the GPU in the laptop.”

And the saddest story begins when a Lenovo laptop with its compound finds itself at the center of this scenario.
We show what it looks like after that in this video:

 
In addition, code 43 may appear if:

✅

CAUSE: One of several video chip power sources is failing. This point requires no explanation. There are no unnecessary power supplies in the video system. If even the secondary power supply is missing, the chip itself can be identified in the system due to the presence of the main one, but after scanning its modules, it will “see” that there is a module that did not receive its secondary power supply and did not “respond”, so naturally the final result of the check will be negative and an error will appear. REQUIRES: First of all, a very attentive diagnosis is needed to find this non-functioning source and determine why it is not working.

The saddest thing is that even such a simple problem often becomes the reason for the “torture” of the motherboard.
The art of diagnostics requires not only a basic understanding of the motherboard’s circuitry, but also patience.
This is because it often takes several looks at the datasheet, turning the board over, finding all the sources on it, taking the necessary measurements, and comparing them with the values that should be there in order to “measure” the entire chip.
That is why most unskilled “specialists” start with the old legend that you first need to heat something up.
Some of them are not even aware of what exactly needs to be heated (and in modern gaming laptops, it has become pointless to heat anything), so they simply start heating everything in turn.
Some don’t stop at heating, but also start soldering.
With the quality of soldering that I wrote about above.
The result is also shown in the video above.
So I have to note again that:

☝	In this case, the experience, intuition, and steady hands of the engineer who gets your laptop become decisive. And in this case, choosing the wrong service often leads to increased laptop repair costs and even forced motherboard replacement, as there are no other options left.

Let’s move on.
It’s not often, but sometimes code 43 is caused by:

✅

CAUSE: Failure of the GPU itself. If this happens, it is usually due to systematic neglect of the need to clean the gaming laptop, resulting in constant overheating of the video chip during gaming. REQUIRES: checking error statistics in video memory tests, measuring all power sources and, if necessary, replacing the GPU.

I am not listing this reason last by accident.
The fact is that the experience of our service, which has already replaced more than a hundred video chips and accumulated statistics on symptoms, shows that cases where the chip itself is to blame are the least common.
In our opinion, they account for a maximum of 30%, if not less.

☝	This is the main reason why you should not immediately look for a service with the requirement “I need to replace the video chip in my laptop.”.

Because by asking such questions, you yourself are provoking high repair costs.
And how much could it be if, for example, a GTX 1660 chip can cost around $150-200 on eBay?
And you are provoking unnecessary soldering of your board and chip, even if it is not necessary!
At the same time, the blame for this unnecessary soldering falls on your shoulders, because you wanted it yourself.
As a result, you may end up with a non-working laptop, having paid for the chip, for the soldering work (or for the diagnosis “well, the guys tried”) with a poorly soldered (burnt) chip on poor flux (as in the video above).
And at the same time, you will be left without the possibility of making any claims, because all this was done at your request!

👇

So here is a sequence of actions if you see the message “This device has been stopped because it reported a malfunction. (Code 43)” in the list of devices near the video adapter in your laptop.:

1) Try removing the device and reinstalling the drivers. This process does not guarantee 100% that you will eliminate the software component of the problem, so if it does not help, I recommend reinstalling the system completely. If you have a lot of programs and your own data, you can install another hard drive (after removing your own with the system, believe me, it’s safer that way). 2) If this does not help and you are an experienced user, you can update the BIOS firmware. The connection with the problem is that when the BIOS is updated, the video BIOS located in the system is also updated. If the OS previously updated the firmware itself and something went wrong, this could also lead to code 43. But this is already in the category of very specific scenarios. 3) If that doesn’t help, take it to a service center.

As practice has shown, replacing the video chip in a GTX 1050 laptop with code 43 is quite common.
One can even say that this is a typical malfunction with such symptoms.
In chips with a larger package size (GTX 1060, GTX 1080, GTX 1650, GTX 1660, RTX 2060, RTX 2070, RTX 2080, etc.), this is less common (with the exception of the GTX 1070, the number of replacements for which is also beginning to attracting attention).
So you won’t be able to deal with this problem on your own.
If you contact a “conveyor” or “superficial” service with the wrong wording, you are likely to get a sad result, as in the video above.
So contact us, write to us, call us.
At the very least, we will definitely help you with decent advice.
Then you can decide where to go and what to look for based on your circumstances.
I almost forgot: in some cases, code 43 can be caused by:

✅

CAUSE: Other very specific circumstances. Short-circuiting power sources during disassembly (for example, without first disconnecting the battery) Removal of SMD components from the motherboard due to careless handling, various types of deformation of the laptop, leading to the separation of chips from the board (hitting the keyboard or floor) and, of course, spillage. REQUIRES: When visiting the service center, be sure to inform the technician working on your laptop and conduct a thorough diagnosis with elements of luck (believe me, in such cases, luck sometimes helps even more than experience).

❗	The video card is present, but judging by the low frame rate in games, the system is running on integrated graphics.

to the beginning of the article
This is a very specific symptom.
It is not encountered often, but even a specialist can be stumped by it.
We have seen this many times, because laptops with such malfunctions that were brought to us for repair were previously returned from other services with the statement “probably, the video chip in the laptop needs to be replaced, but we are not sure”))).
And it can manifest itself in a wide range of ways, from a slight drop in frame rate or a drop in the performance of the discrete video chip to the performance of the graphics built into the processor.

☝

To understand why this may occur, please note that the discrete graphics card itself (i.e., the part of the board on which it is located) on a laptop motherboard can consume more than half of the power from the power adapter. For example: the Acer Nitro 5 with GTX 1050 is powered by a 135W power supply, while the Clevo Sager with the same processor but with a GTX 1070 graphics card requires a whopping 230W to operate! That means the difference in power consumption between the GTX 1050 and GTX 1070 is only about 100 W! Meanwhile, everything else in the laptop consumes the same or even less power.

So it is obvious that the video system consumes a large amount of power when in operation.

☝	By the way, there are some nuances regarding the correct use of batteries in gaming laptops! (An article will be available on our website soon).

This leads to the first reason for such behavior:

✅

CAUSE: Insufficient power supply to the video card. The essence of this phenomenon is that the video card power consumption monitoring system detects a power shortage in the system when the video card is launched in full mode. Of course, this operating condition is not accepted by the power controller, and a switch to the built-in graphics occurs, causing a sharp drop in the frame rate. REQUIRES: connecting a power source of sufficient capacity.

The simplest reason for this is that the power supply unit is providing insufficient power (less than the required amount specified on the laptop case).
This symptom is primarily encountered by those who have purchased a non-original or “weak” power supply unit.
It should be noted that such problems can also be caused by more complex issues with components in the main power supply circuits, usually 19V on the board, namely:

✅

CAUSE: Problems with input transistors, power connector contact with the board, integrity of the connector itself, connector in the power supply cable, etc. The essence of the problem is usually that increased resistance occurs at the problem location, which prevents the passage of currents and power transmission. REQUIRES: and again requires careful and accurate diagnosis, rather than mindless heating or soldering everything that comes to hand!

This problem may be accompanied by the following symptoms:
a) periodic battery recharge failure,
b) unexpected transition of the laptop from power supply to battery power and its discharge,
c) changes in backlight brightness and sound signal.
In addition to the possible lack of power for the video card, there is another problem:

✅

CAUSE: Incorrect monitoring of the GPU power source. In essence, this means that the power management system receives incorrect data about the power consumed by the video chip. REQUIRES:Qualified inspection of the video card power supply monitoring and control unit (soldering of components and their ratings).

In simple terms, a controller that incorrectly receives information about the power consumption of the video chip simply decides that too much power is being transmitted to its power supply circuits.
It adjusts the transmitted power, which automatically puts the system into the state described above with a drop in performance.

❗	When loading a discrete video card, the system crashes with a blue screen or shuts down abruptly.

to the beginning of the article
This behavior is also common in problematic gaming laptops.
The most common issue is the appearance of a “blue screen.”
The system usually starts to “crash with blue screens” either immediately after installing drivers for a discrete graphics card or when the system activates a GPU to run graphics-intensive programs.

✅

CAUSE: “Blue screen” during driver installation on the GPU or when loading it. The nature of this phenomenon may vary, but essentially it is a graphics driver failure due to a hardware problem in the video system. This could be a faulty video chip, errors when working with video memory, or even instability in one of the GPU’s power sources. REQUIRES: checking all power sources (presence/level/stability), video memory tests, replacing the video chip in the laptop and/or video memory, or repairing the power source.

Although the problem is very common, as you can see, there can be many different causes for it.

☝	So, as in the case of “No video card in the list of devices,” there is no need to rush to the service center shouting, “I need a GPU replacement in my laptop.”!

This will only create a rush, provoke an increase in repair costs, and potentially unnecessary repairs.
There is a more dangerous modification of this symptom:

✅

CAUSE: Sudden shutdown of the laptop during load on the video card. In this case, the shutdown is complete, as if it had been unplugged from the outlet without a battery. This is a very dangerous symptom, because it can occur not only due to overheating (so-called thermal shutdown), but also due to the activation of the charging controller and/or power supply protection when the transistor that forms secondary voltages from 19V is shorted. Roughly speaking, such a shutdown can occur for the first time at the first symptoms of a transistor “breakdown” (its uncontrolled opening) and save the load. With repeated turn-ons, the transistor may already be completely “broken through” and uncontrolled, so the possibility of sad consequences increases significantly. And in the case of a video memory source, the entire video memory and video chip may be burnt out! REQUIRES: immediately stop using the laptop, contact the service center for further disassembly of the laptop and diagnosis of power sources for “breakdown.”

☝	I consider this symptom to be the most potentially dangerous and one that requires serious attention at the first signs of its appearance. The difference in repair costs if this symptom is ignored can be 2-3 times higher, or even more.

❗	Artifacts occur when the discrete graphics card is used.

to the beginning of the article
This symptom is also common and can occur even without video load, but simply when the operating system is booting up.
It can look different, for example like this:

(photo pending)

I would like to mention that not all artifacts should be attributed to the video card.
Problems with incorrect image display on the screen may be due to the display panel as a result of damage to the decoders or poor contact in the connector itself.
A grid of small dots or a “floating” image may be due to problems with the display panel cable.
Even poor contact in the display matrix connector (or in the connector) on the motherboard can cause image defects to appear on your laptop screen.
So, on average, about half of these issues are caused by the video chip and video memory.
But if they do occur, then:

✅	CAUSE: This could indeed be a problem with the GPU. This can happen, for example, due to systematic overheating, which causes degradation of the chip structure and one of the blocks in the crystal begins to malfunction. REQUIRES: in case of a video chip malfunction, of course, GPU replacement.

But in life, things are often not so clear-cut.
There is another reason for the appearance of artifacts:

✅

CAUSE: Problems with video memory (VRAM). However, this doesn’t necessarily mean that the video memory itself is to blame. As with Code 43, the issue may lie in the contact between the video memory and the motherboard. Or, for example, there may be a problem with the contacts of the data transfer interface with the video memory on the GPU side. REQUIRES: One more situation where experience, intuition, and the steady hands of the engineer working on your laptop take center stage. The key is to accurately identify where the problem really lies. And quite often, that means picking up the soldering iron. Soldering carefully and precisely to methodically go through the necessary components and find the exact spot with a bad connection, for example.

Even though an experienced engineer eventually develops an intuitive sense of which type of artifacts point to a chip issue and which suggest a problem with video memory, the process always starts with a series of tests and diagnostic procedures.
The goal is to pinpoint the most likely location of the fault with the highest possible accuracy.

☝	Only after gathering all the test results and diagnostic data — and even then, often without 100% certainty about the root cause — do we proceed with the repair. And when we do, every step is taken with extreme care to avoid introducing any new issues in the process.

You wouldn’t believe how much work sometimes goes into this kind of approach before we finally reach a successful result.

And yet, we only charge for the work that was actually necessary to fix the issue — or which we know from experience is required.

To make it clearer, let us tell you a real story involving a Lenovo laptop and its infamous compound (by the way, there’s a video about this issue above, in the section titled “Code 43”).

🔧 The beginning:
A laptop came in showing a “Code 43” error on its GTX 1060 discrete GPU.
Diagnostics pointed to one of the video memory banks.
We replaced it — and the laptop booted. Everything worked. Great! We were happy and put it on testing.

🕐 Day two:
After two days of stress testing — the “Code 43” error returned.
Diagnostics showed the same memory bank acting up.

❗ Second replacement:
We replaced the chip again.
Everything worked again.
Back to testing.
And again — two days later, same error.

🤯 Hmm, now it’s clear:
It’s not the memory itself. And not the contact underneath.
Most likely — either a faulty GPU chip, or a bad connection under the chip on the line to that memory bank.

🧯 Time for a serious step:
We carefully desoldered the GPU — knowing full well that with Lenovo’s compound, this can easily lead to having to reball several (or even all) memory chips.
Why? Because when the solder under the GPU melts, it also melts under the memory chips.
And when that happens, the solder balls — under pressure from heat expansion — can “pop” and short together with neighboring ones.

🔍 Diagnostics:
We checked the memory power rail — short circuit.
Started desoldering memory chips, one by one.
Only after removing the last one did resistance return.
Frustrating, but expected.

🛠️ Reballing:
We reballed the memory chips and the GPU.
Soldered everything back.

🤨 Surprise:
The laptop didn’t boot. No initialization at all.
We checked everything — including the CPU.
And there it was: solder balls had shorted under the processor too, cutting off its power.

😬 Going even deeper:
We had to desolder the CPU as well.
Lots of care (and prayers) to avoid tearing pads off the board (thanks, compound!) or cracking the die.
Reballed the CPU, soldered it back in.

✅ Happy end:
Laptop booted. Display came up. Driver installed successfully.
Ran tests for a full week — everything was rock solid.

📦 The bottom line:
We handed the laptop back to the client.
And on the invoice? Just the GPU and memory soldering.
Everything else — on the house.
Because that’s how we do things.