# Available GPU Models

The following GPU models are available:

<table><thead><tr><th width="256">Item</th><th>GPU Basic</th><th data-hidden></th></tr></thead><tbody><tr><td>GPU type</td><td>NVIDIA GeForce RTX 4090</td><td></td></tr><tr><td>GPU count</td><td>1</td><td></td></tr><tr><td>GPU memory</td><td>24 GiB</td><td></td></tr><tr><td>GPU compute power</td><td>82+ TFLOPS (FP32)</td><td></td></tr><tr><td>GPU CUDA cores</td><td>16,384</td><td></td></tr><tr><td>GPU memory bandwidth</td><td>8,064 Gbps</td><td></td></tr><tr><td>CPU size &#x26; type</td><td>z2a (z2a.std.8)</td><td></td></tr><tr><td>vCPU count</td><td>8 cores</td><td></td></tr><tr><td>CPU RAM</td><td>32 GiB</td><td></td></tr></tbody></table>

{% hint style="info" %} <mark style="color:blue;">**Note**</mark>

<mark style="color:blue;">GPU memory refers to the memory available on a GPU device used for temporary data storage. It is distinct from the CPU memory and is specifically used to manage the high bandwidth requirements of graphics-intensive workloads.</mark>
{% endhint %}