Visualizing Large Datasets on the GPU with Vega and MapD

&gt; MapD uses Vega to drive the rendering engine directly on the result set of a SQL query without ever requiring the data to leave the GPU<p>How big is the dataset? If it can&#x27;t ever leave the GPU then it is at most a few GB? Unless there are several GPUs at play then it&#x27;s N * (a few GB). If there are a few GPUs at play then this dataset would fit into DDR3 RAM on a single mainstream Xeon node, or entirely into MCDRAM on a Xeon Phi node.<p>Please correct me if I&#x27;m wrong.

At first, I thought it was referring to AMD&#x27;s new Vega GPU family. I was hoping they found a particularly good use case for it.

@tmostak large dataset visualisation like this looks great, but one of the most appealing parts of Vega for me is interaction. It&#x27;s just as easy with Vega to create composite interactions for filtering and navigating data as it is to visualise it. Is there any scope for this type of architecture to support more than just serving rendered PNGs? (Can it do that at 60fps? :P)

A lot of these visualisations would benefit greatly by using 2d or 3d kernel density estimates instead of a simple scatter plot. See for example: <a href="https:&#x2F;&#x2F;youtu.be&#x2F;Xz_7Ej6JsMY" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;Xz_7Ej6JsMY</a>

Can I use it on a MacBook Pro? My concern is that it doesn&#x27;t have a dedicated GPU.