Unsolved
This post is more than 5 years old
22 Posts
0
7173
July 17th, 2014 15:00
Switchable graphic questions
I'm buying a new laptop, but wondering whether I have to settle one with switchable graphics or IGP alone.
The main reason I ask : My previous bad experience with 2 laptops dying within 2-3 years because the GPU got unstable (artifacts, no video output) due to high temperature inside the laptop, even when I only play games occasionally. After some research on the net, I found this is a common problem with discrete GPU in a laptop. The constant exposure of high temperature tends to break the solder joints of the video chip. . That's the reason why reflowing/reballing service repair to the GPU became popular in the first place. Such repair has been proven to be a temporary fix though.
My question is related to how the switchable graphic works.
1. Is the IGP always the primary device connected directly to display output port in all notebooks with switchable graphics ?
As far as I understand, when discrete GPU (dGPU) is used for rendering, it still has to copy frames to the IGP to display its output. Is this a common design in all notebooks with switchable graphics regardless of models, laptop manufacturer or dGPU type (Nvidia/ATI) ?
2. During the initial booting process (or in Bios setup), does the notebook use only the IGP to display output ? What happens to the dGPU in this state ? In Windows, it is clearly understood that dGPU is either non-active or in sleep mode because the driver application is managing both GPUs.
3. In a situation where the dGPU is malfunctioning (unstable/artifacts/dead), can the notebook still boot reliably ? Can a user simply disable dGPU completely (assuming the boot process relies completely on the IGP) ?
4. How much battery life is affected by dGPU in sleep/non-active state vs a notebook with only IGP ?
It's nice to be able to play some games occasionally, but I really hate it when dGPU starts to crack.
ejn63
11 Legend
•
87.5K Posts
•
321.3K Points
0
July 17th, 2014 16:00
1. No. There are two ways to design switching graphics - software (muxless) or hardware (muxed). In general, most consumer-grade systems are muxless, while gaming systems and many business-grade systems have a multiplexer for the dGPU.
2. The iGPU is used most of the time, and before Windows loads.
3. In cases where there IS a multiplexer, the dGPU can be turned off at the BIOS level. In muxless systems, you can usually disable the dGPU by software.
4. Shouldn't be affected at all in a muxless system - the multiplexer in a muxed system does consumer some power. One of the two reasons for the muxless design is improved battery life (the other of course is lower cost).
Patroclus
22 Posts
0
July 18th, 2014 13:00
Thank you for the answer.
After carefully thinking pro & cons, I think I will avoid dGPU this time.
It's just becoming too expensive for me to spend over $1000 for a laptop that only survives for 2-3 years after the warranty expired, not mentioning the sudden disruption for looking for a new laptop & reinstallation,
It looks like the technology has been pushed too far sacrificing reliability/durability for the sake of a powerful laptop. Even before they kept dying on me, I've been questioning myself how durable a laptop could be when the GPU & CPU reaching 85-90C or higher inside such a closed space considered normal. No more dGPU for me in a laptop. I will just rely on desktop computer for powerful applications/gaming.
Thank you.