Welcome guest. Before posting on our computer help forum, you must register. Click here it's easy and free.

Author Topic: Video *almost* dead?  (Read 2877 times)

0 Members and 1 Guest are viewing this topic.

Rapid_Dissent

    Topic Starter


    Rookie
  • Thanked: 2
    • Yes
  • Computer: Specs
  • Experience: Experienced
  • OS: Windows 7
Video *almost* dead?
« on: February 11, 2011, 12:35:59 PM »
So, I fell asleep watching a movie on my computer, as I normally do, and when I woke up, my monitor was displaying nothing as though it was asleep, but it had the blue power light on (getting signal), and the monitor always goes into sleep mode under these circumstances. So, I tried waking the computer up, and nothing came on screen. When I inevitably went for the hardware reset switch, POST came up, but disappeared after just a second or two, leaving me in the dark about 3/4 of my POST screens, and keeping me from seeing my bootloader, GRUB. When Windows starts up, there is another part I can see, just as the mouse becomes available, right as the login screen is about to come up, and then a second later it disappears again, accompanied by the windows startup music. Starting into Ubuntu Linux, there is no video output during startup.

I have a theory, but I know too little about the BIOS to know if it's viable (but it's the only theory that makes sense atm). The idea is that the video card has crapped itself, and the splintered fragments of output I AM getting are when the BIOS / windows have the CPU handle video out, rather than the video card; when the video card is put to task, the ride is over. The one thing really holding this theory back is that I don't know if the BIOS would use the video card; seems like a waste, since it's (presumably) simpler to have the CPU handle video output rendering and pass it 'through' the video card... not exactly an intensive task.

What do you guys think?

Computer: Asus P6T, Intel i7 [email protected], 6gb Crucial Ballistix DDR3 @ 1066, nVidia BFG 275GTX OC, ViewSonic VA1930WM (almost 4 years old; seems to be fine), Win 7 Ultimate 64-bit, Ubuntu 10.10 32-bit.
All tyranny needs to gain a foothold is for people of good conscience to remain silent.
-Thomas Jefferson

Rapid_Dissent

    Topic Starter


    Rookie
  • Thanked: 2
    • Yes
  • Computer: Specs
  • Experience: Experienced
  • OS: Windows 7
Re: Video *almost* dead?
« Reply #1 on: February 11, 2011, 02:02:56 PM »
SOLVED:

Just got back from my friends place where we plugged in another video card; yahtzee. *sigh*, off to buy another card; hope this one lasts more than 8 freaking months...

This is why you follow the Socratic method 100%: it was not the video card after all. I was running the GPU test on my buddy's monitor as well; before rushing out to buy a new video card, I decided to test the monitor as well, just to be certain. It was the 4 year old monitor. I have no idea how or why, but it finally gave out.
« Last Edit: February 11, 2011, 02:39:04 PM by Rapid_Dissent »
All tyranny needs to gain a foothold is for people of good conscience to remain silent.
-Thomas Jefferson

patio

  • Moderator


  • Genius
  • Maud' Dib
  • Thanked: 1769
    • Yes
  • Experience: Beginner
  • OS: Windows 7
Re: Video *almost* dead?
« Reply #2 on: February 11, 2011, 02:03:41 PM »
I think the BIOS is not the issue here and should be eliminated as a possibility...
" Anyone who goes to a psychiatrist should have his head examined. "

Rapid_Dissent

    Topic Starter


    Rookie
  • Thanked: 2
    • Yes
  • Computer: Specs
  • Experience: Experienced
  • OS: Windows 7
Re: Video *almost* dead?
« Reply #3 on: February 11, 2011, 02:35:32 PM »
I never indicated the BIOS could be the problem; it was simply one of the two periods in which I got video output.
All tyranny needs to gain a foothold is for people of good conscience to remain silent.
-Thomas Jefferson

patio

  • Moderator


  • Genius
  • Maud' Dib
  • Thanked: 1769
    • Yes
  • Experience: Beginner
  • OS: Windows 7
Re: Video *almost* dead?
« Reply #4 on: February 11, 2011, 03:56:54 PM »
I never indicated the BIOS could be the problem; it was simply one of the two periods in which I got video output.

Then why this ? ?

Quote
I have a theory, but I know too little about the BIOS to know if it's viable (but it's the only theory that makes sense atm). The idea is that the video card has crapped itself, and the splintered fragments of output I AM getting are when the BIOS / windows have the CPU handle video out, rather than the video card; when the video card is put to task, the ride is over. The one thing really holding this theory back is that I don't know if the BIOS would use the video card; seems like a waste, since it's (presumably) simpler to have the CPU handle video output rendering and pass it 'through' the video card... not exactly an intensive task.
" Anyone who goes to a psychiatrist should have his head examined. "

Rapid_Dissent

    Topic Starter


    Rookie
  • Thanked: 2
    • Yes
  • Computer: Specs
  • Experience: Experienced
  • OS: Windows 7
Re: Video *almost* dead?
« Reply #5 on: February 11, 2011, 04:10:54 PM »
Didn't know if the BIOS would call the GPU to render output; it was the one aspect of my theory that didn't make sense, and now I know why. Interesting way for a monitor to die, though...

The idea is that the video card has crapped itself, and the splintered fragments of output I AM getting are when the BIOS / windows have the CPU handle video out, rather than the video card...
All tyranny needs to gain a foothold is for people of good conscience to remain silent.
-Thomas Jefferson

Salmon Trout

  • Guest
Re: Video *almost* dead?
« Reply #6 on: February 11, 2011, 04:56:03 PM »
I have a theory, but I know too little about the BIOS to know if it's viable

It's not viable

(but it's the only theory that makes sense atm).

It makes no sense.

The idea is that the video card has crapped itself, and the splintered fragments of output I AM getting are when the BIOS / windows have the CPU handle video out, rather than the video card

If the video card had crapped itself, you would see nothing at all, certainly no POST. Your understanding of what goes on inside a computer is sketchy indeed if you think that the "BIOS" or the "CPU" can somehow magically send video signals down the lead (which is connected to the video card) to the monitor.

Quote
The one thing really holding this theory back is that I don't know if the BIOS would use the video card; seems like a waste, since it's (presumably) simpler to have the CPU handle video output rendering and pass it 'through' the video card... not exactly an intensive task.

The BIOS is a kind of flash memory which contains binary code. Nothing else. It has no video hardware at all. Neither does the CPU. Your monitor is connected directly to a display card which is where any video signals must come from, and nowhere else.


Rapid_Dissent

    Topic Starter


    Rookie
  • Thanked: 2
    • Yes
  • Computer: Specs
  • Experience: Experienced
  • OS: Windows 7
Re: Video *almost* dead?
« Reply #7 on: February 13, 2011, 09:01:57 PM »
Hrm, I suppose that's one way of answering a question.

You're right that I am unfamiliar with the specifics of how the hardware works, but I don't think you understood where I was going with that. The output obviously goes through the video card, but whether or not it is rendered on the video card was the bit I was angling toward with my previous assumption. I assumed that if the CPU calculated output and handed it to the video card, most of the video card's facilities (including damaged ones), would be bypassed and the video card would likely hand the output over to the monitor verbatim with minimal calculations. It's not as though computers aren't used to doing this; even as 'recent' as Unreal Tournament, you could elect to have the CPU render video out.

Also, given how my 8800gts died, video cards don't always just stop working.

tl;dr Quit belittling.
All tyranny needs to gain a foothold is for people of good conscience to remain silent.
-Thomas Jefferson

BC_Programmer


    Mastermind
  • Typing is no substitute for thinking.
  • Thanked: 1140
    • Yes
    • Yes
    • BC-Programming.com
  • Certifications: List
  • Computer: Specs
  • Experience: Beginner
  • OS: Windows 11
Re: Video *almost* dead?
« Reply #8 on: February 13, 2011, 10:37:09 PM »
You're right that I am unfamiliar with the specifics of how the hardware works, but I don't think you understood where I was going with that. The output obviously goes through the video card, but whether or not it is rendered on the video card was the bit I was angling toward with my previous assumption.
I assumed that if the CPU calculated output and handed it to the video card, most of the video card's facilities (including damaged ones), would be bypassed and the video card would likely hand the output over to the monitor verbatim with minimal calculations.
One way of "thought-experimenting" to prove-disprove things like this, is to consider- what if this was the case.

if I understand your thought process, you are thinking that the Video card merely acts as an "adapter" so that the BIOS can direct output to the monitor; however, if this was the case, then there would have been no such thing as Video card failures until rather recently; for example, older Video cards (ISA) being merely "pass-through" things, would not require video memory; however, they do require video memory. If that memory get's corrupted/damaged then you will see artifacts on the screen, regardless of the screen mode.



Quote
It's not as though computers aren't used to doing this; even as 'recent' as Unreal Tournament, you could elect to have the CPU render video out.
Ahh, here is the confusion. You are thinking that the video card only "does work" when it is being made to render things such as using it's 3-D accelerator features. To be fair, 3-D acceleration is more likely to reveal corruption (particularly with Video memory) merely because it uses it for storing textures; if you (as you noted) use software mode in a game, all textures,vertex, and Lighting data are stored in system memory, and the CPU is essentially used to build a "bitmap" of each frame; the Video card, however, still needs to store that video data in video memory and if it happens to decide to store it in a area that is corrupted you will see the results of that (in fact, you will see it everywhere). Also, the windows desktop uses basic 2-D features; while many of these are accelerated, the acceleration is done by the drive in many cases, and if the driver says "draw this line" either the video card draws the line or it doesn't; I don't think I'd expect that the video card would continue to say "ok I draw it" for every single request the driver makes (thus leaving the screen blank). Usually if the "higher" functions of the video card are dead you simply get a hard freeze during boot or when you try to switch to a 3-D mode.



Thing is, although your theory makes some sense since using higher resolutions uses more memory, such as text-mode 80x25 (what the BIOS uses.. although newer computers switch the card to a graphics mode(640x480) so they can draw the EPA or their manufacturer logo/splash), windows switches to 640x480 to display it's splash, and then switches to your desktop resolution. Thing is, though, if the video card was having problems, then it wouldn't simply accept the various mode switching VGA commands and later accept the control of the windows driver; at some point during boot it would either freeze solid (as the video card freezes the entire machine) or, you would see a garbled display the entire way along. As you have discovered, however, it was the monitor; at some point it stopped "supporting" the video modes the video card wanted to change to. I've had monitors mysteriously stop supporting modes, or in some cases, they start screaming in agony (very high-pitched squeal) when switching to that mode; I had a old monitor that supported 1024x768 out of the box that at some point decided to start screaming like a banshee when in that mode, to the point where I had to switch to 800x600; Eventually the entire unit gave up the ghost as the vertical coils died, leaving me with a very bright horizontal line in the center of the screen, making things a tad difficult.
I was trying to dereference Null Pointers before it was cool.