Stimulating question! Maybe some useful observations as we do client webl <> server cuda acceleration of vector data (our use case is big visual graph analytics

Stimulating question! Maybe some useful observations as we do client webl <> server cuda acceleration of vector data (our use case is big visual graph analytics) and have to handle large + real-time contexts: * SVG at those levels sounds quite amenable to serverside offloading for bulk rendering: not "easy", but many options for HPC types. The browser can handle pan/zoom. My main computational question is hitmaps & hover effects, but wouldn't bet against ingenuity here as the use case is fairly static. Arguably, even Figma should move off in-browser and to offloading, if it hasn't yet. * Browser acceleration of SVG has been improving over the last decade. Maybe more B round+, having 1-2 people work with Chrome will help knock out whatever is left in the browser. Working to optimize one site on both sides is WAY easier than optimizing for everyone, this is an open secret of Nvidia etc games & top websites for ~decades. Real-time SVG is less practical in our case b/c our users will run clustering & filters and expect real-time vector transforms of 2M nodes + edges (imagine Google Maps... but live), but for something like a site design.. a lot more room for perceptual tricks!