AMD 2016-2019 GPU Roadmap: Vega Launching Next Year with HBM2 followed by Navi in 2018
At its Capsaicin GDC 2016 event, AMD has announced that its next generation Polaris architecture will be followed by Vega and Navi which are set to arrive in 2017 and 2018, respectively.
According to a slide pulled up by Raja Koduri, head of Radeon Technologies Group (RTG) at AMD, Vega will be the company’s first architecture to feature the next iteration of High Bandwidth Memory, or HBM2. Which means Polaris GPUs will only be based on HBM and GDDR5(X), though it’s not confirmed yet.
Navi on the other hand will benefit from scalability and ‘Nextgen memory’.
When AMD announced the 14nm FinFET based Polaris architecture back in January at CES 2016, the company said their guiding principle for Polaris was “to power every pixel on every device efficiently,” and that they are looking to stars as their inspiration.
At the time, Koduri revealed that the new naming scheme for their future GPU architectures will be based on galaxies, star systems, and stars. He further said Polaris is just “the beginning our journey through space” and more will come in the future.
As part of their strategy of giving astronomical code names, AMD has allegedly been calling its flagship Greenland GPU “Vega 10” internally for some time. This new GPU roadmap for 2016-2019 now confirms that report.
Vega was the first star other than the Sun to be photographed and the first to have its spectrum recorded. But perhaps the most significant fact of all about Vega is that it served as the baseline for calibrating the photometric brightness scale, and was one of the stars used to define the mean values for the UBV photometric system.
Vega graphics architecture, which is coming next year, will benefit from AMD’s second gen of the the vertically stacked HBM standard. HBM2 operates at twice the speed of HBM1 for double the memory bandwidth, as well as scales to capacities 8 times larger than the first generation.