Turn off the Ad Banner  

To print: Select File and then Print from your browser's menu.

    -----------------------------------------------
This story was printed from CdrInfo.com,
located at http://www.cdrinfo.com.
-----------------------------------------------

Appeared on: Thursday, March 10, 2016
Computex To Mark The Launch of AMD's New Polaris And Nvidia's Pascal GPUs

This summer, PC gamers should expect new 14nm GPU announcements from both Nvidia and AMD, as both companies move from the 4-year-old 28nm process.

AMD's new Radeons will be armed with the company's next-generation GPU architecture, Polaris. An official launch should be expected sometime in June, with a release date to follow in the third quarter.

The Polaris graphics architecture brings new fourth-gen GCN cores, a new memory controller, new multimedia cores, and a new geometry processor.

It is not clear whether AMD has chosen Globalfoundries or Samsung to manufacture its new 14nm FinFET GPUs. What we do know is that the Polaris chips will support both HDMI 2.0a and DisplayPort 1.3. The new chips are also capable of encoding and decoding H.265 video up to 4K resolution. AMD has promised "improved shader efficiency," "memory compression," "better instruction pre-fetch," and such.

Polaris chips may not leverage AMD’s involvement with High Bandwidth Memory, at lest for the some models. Instead relying on the more common GDDR5, or a variant of it called GDDR5+.

Currently, AMD has a share of around 20% in the discrete graphics card market for desktops.

Nvidia is also expected to launch its Pascal GPU architecture at about the same time. The company plans to offer a first glimpse of the new architecture at the GPU Technology Conference (GTC) 2016 in April, and the announcement of the new GTX 1080/1070 GPUs should be expected in June. The high-end parts should hit the market during Q3, with the mainstream Pascal-based graphics cards to follow in the fourth quarter.

It is not clear whether AIBs based on Pascal will have High Bandwidth Memory 2 (HMB2). Nvidia says that it will enhance the performance of Pascal GPUs with the NVLink interconnect technology, which is said to allow data sharing at rates 5 to 12 times faster than the traditional PCIe Gen3 interconnect.



Home | News | All News | Reviews | Articles | Guides | Download | Expert Area | Forum | Site Info
Site best viewed at 1024x768+ - CDRINFO.COM 1998-2017 - All rights reserved -
Privacy policy - Contact Us .