It’s been quite some time since AMD refreshed its GPU architecture. To be exact, the current architecture, GCN – Graphics Core Next – was first introduced by AMD back in 2011. Now, it seems like AMD is finally going to update its already ageing GPU architecture to a newer design codenamed Polaris.
It’s rumoured that the Polaris GPU would unsurprisingly bring better efficiency and “viewing experience” for AMD users compared to to the older GCN architecture. In addition to that, the Polaris architecture is also said to provide twice the performance per watt compared to GCN. Interestingly, it was reported that due to the architectural update, upcoming GPUs from AMD will feature the second generation HBM – High-Bandwith Memory – technology. This feature will allow future GPUs by AMD to operate at an eye watering 1TB/s (yes, terabyte) of bandwidth speed. Talk about a major upgrade.
One might wonder why the new architecture is called Polaris. By referring to the leaked Polaris architecture photo by AMD, it could probably mean that AMD wants to mimic what Polaris and stars alike are – the most efficient photon generators in the universe – into its upcoming architectural update. This could mean that AMD’s upcoming GPU architecture might bring an increase in efficiency in terms of pixels generated on-screen; probably beyond what existing GPU technologies on earth is currently able to generate.
If this is the case, it means that AMD is changing its GPU naming scheme from branding GPUs according to locations on earth to intergalactic stars in galaxies far far away. It seems that AMD fans have a lot to look forward to in 2016, especially now that Radeon Technologies Group has taken over.