AMD Unveils GPU Architecture Roadmap, “Polaris” no HBM2 Memory?

Alongside its big Radeon Pro Duo flagship card, AMD unveiled its GPU architecture roadmap that appears as far in to the future as early-2018. At that time, AMD shall have introduced as much as three innovative GPU architectures. It commences with the launch of its 4th generation Graphics CoreNext architecture, codenamed “Polaris,” in mid-2016. Built on the 14 nm FinFET procedure, “Polaris” is likely to give a whopping 2.5x upsurge in performance-per-Watt for AMD, in comparison to its current GCN 1.2 architecture on 28 nm.

AMD GPU Architecture Roadmap

In early on-2017, AMD packages to release the “Vega” GPU architecture. While this seems to provide a 50% upsurge in performance-per-Watt over Polaris, its highlight is usually HBM2 recollection. Does this imply that AMD projects to skip HBM2 on Polaris, and adhere to GDDR5X? Could AMD come to be deciding on a similar method of NVIDIA, by launching its performance-segment GPU as an enthusiast product first, giving it a free of charge operate on the markets till early on-2017, and launching a Vega-centered big-chip with HBM2 memory then, overtaking as the enthusiast-segment merchandise? Some correct time in early-2018, AMD will start the “Navi” architecture, which seems to give a 2.5x performance-per-Watt lead over Polaris, taking advantage of an even newer memory standard.

source

Advertisement