Questions tagged [vram]

shortcut for Video Random Access Memory, It is memory specially dedicated to store and represent visual data.

Video Random Access Memory

  • It is memory specially dedicated to store and represent visual data.
  • In the past was the same as screen buffer (Video Display)
  • and accessing it means directly access video output ...

With new technologies in Graphics is VRAM more structured usually to:

  1. Buffers

    • direct screen,back,secondary back,depth,... buffers
    • Alpha,stencil and effect buffers
  2. Texture memory

    • store the 1/2/3D image data
    • usually with special priorities because this memory is more often accessed in 3D graphics by gfx card during rasterization
  3. Geometry

    • this part holds the geometry representation of the scene
    • usually store VBO/VAO or display lists
    • store the vertexes,colors,texture coordinates and much much more
    • in vector form
  4. Shaders

    • with Shading language the gfx cards need to store also shader programs
    • these are usually stored in physically separate memory so the gfx processors can execute it

Modern memory controllers incorporated into graphic chips can use single memory space for all of these structures.

some VRAM memories also have more then single interface like (DUAL interface memories)

  • these allow to more then single access to memory per cycle
  • so host computer can access VRAM and gfx chip too without contentions or dellays
  • these are also used for storing streams from Hi-FPS camera systems where usually 3 and more streams of data are at all times to widen the bandwidth of the CCD chips ...
26 questions
45
votes
4 answers

Use shared GPU memory with TensorFlow?

So I installed the GPU version of TensorFlow on a Windows 10 machine with a GeForce GTX 980 graphics card on it. Admittedly, I know very little about graphics cards, but according to dxdiag it does have: 4060MB of dedicated memory (VRAM) and; 8163MB…
User1291
  • 7,664
  • 8
  • 51
  • 108
23
votes
9 answers

Tensorflow OOM on GPU

i'm training some Music Data on a LSTM-RNN in Tensorflow and encountered some Problem with GPU-Memory-Allocation which i don't understand: I encounter an OOM when there actually seems to be just about enough VRAM still available. Some background:…
LJKS
  • 899
  • 2
  • 9
  • 17
22
votes
2 answers

TensorFlow: how to log GPU memory (VRAM) utilization?

TensorFlow always (pre-)allocates all free memory (VRAM) on my graphics card, which is ok since I want my simulations to run as fast as possible on my workstation. However, I would like to log how much memory (in sum) TensorFlow really uses.…
daniel451
  • 10,626
  • 19
  • 67
  • 125
7
votes
1 answer

What happens when VRAM is full?

I want to know the current nvidia/AMD implementation of handling VRAM resource allocation. We already know that operating systems use swap/virtual memory when system RAM is full, then what is the equivalent of swap when it comes to VRAM? Do they…
manatails
  • 136
  • 3
  • 11
6
votes
0 answers

Given the number of parameters, how to estimate the VRAM needed by a pytorch model?

I am trying to estimate the VRAM needed for a fully connected model without having to build/train the model in pytorch. I got pretty close with this formula: # params = number of parameters # 1 MiB = 1048576 bytes estimate = params * 24 /…
RDlady
  • 378
  • 2
  • 16
5
votes
2 answers

Can i clear up gpu vram in colab

I'm trying to use aitextgen to finetune 774M gpt 2 on a dataset. unfortunately, no matter what i do, training fails because there are only 80 mb of vram available. how can i clear the vram without restarting the runtime and maybe prevent the vram…
Blazeolmo 343
  • 51
  • 1
  • 2
3
votes
1 answer

Three.JS VRAM memory leak when adding removing THREE.Geometry to scene

I have encountered a VRAM memory leak in my app. The app adds and removes THREE.Geometry very often to create a volumetric animation. If instead of a THREE.Geometry with it's own populated vertices, I used instead THREE.SphereBufferGeometry, the…
2
votes
1 answer

Set custom base address of video memory VGA/VESA in assembly

I started programming in 32-bit protected mode. Im using it for high graph resolutions like 1280x1024 256 colors: mov ax,0x4F02 mov bx,0x107 int 0x10 but I have a problem with size of video memory (VRAM) because after switching to this resolution…
Segy
  • 213
  • 2
  • 12
2
votes
1 answer

How to calculate vram usage size from a image?

I am learning WebGL and I want to know the formula of calculating vram usage size from a image(jpg/png). Thanks.
user2331095
  • 6,577
  • 8
  • 36
  • 56
1
vote
1 answer

nvidia-smi vs torch.cuda.memory_allocated

I am checking the gpu memory usage in the training step. To start with the main question, checking the gpu memory using the torch.cuda.memory_allocated method is different from checking with nvidia-smi. And I want to know why. Actually, I measured…
core_not_dumped
  • 759
  • 2
  • 22
1
vote
0 answers

How to see VRAM of google tpu?

Is it possible to see the VRAM of google TPU being used? In case of using gpu, we can identify how much memory is being used using nvidia-smi or gpustat, but I don't know how to identity it in tpu. I need help :) Thank you, in advance.
Seungwoo Ryu
  • 121
  • 2
  • 8
1
vote
1 answer

How to get Taskmanager's GPU memory usage from C# code

I would like to grab the GPU memory usage (dedicated and shared) from my C# code. Are there any chances to get these infos via WDDM or performance counter? How does the Taskmanager get these infos? Since I also need the usage of AMD cards in…
mFDev
  • 83
  • 7
1
vote
1 answer

TensorFlow GPU and CPU offloaded ops segregation

Assuming TensorFlow GPU library being used in computation, which operations are offloaded to GPU (and how often)? What is the performance impact of: CPU Core count (because it is now not actively involved in computation) RAM size. GPU VRAM (What…
Karan Shah
  • 417
  • 6
  • 21
1
vote
0 answers

x86 Assembly : How to move code to video memory and execute the code?

I’m working on simple program to change memory bank configuration of old x86 16bit PC with MASM6. Currently, when I execute the code from the main memory it hang up. It seems that it’s because the code itself is in the main memory. When the code is…
ychh0
  • 27
  • 3
1
vote
1 answer

DirectX RenderContext RAM/VRAM

I have 8GB or Vram (Gpu) & 16GB of Normal Ram when allocating (creating) many lets say large 4096x4096 textures i eventual run out of Vram.. however from what i can see it then create it on ram instead.. When ever you need to render (with or to) it…
1
2