Just one month after the launch of its first Pascal-based graphics cards Tesla P100 which is targeted at HPC server accelerators, NVIDIA (NSDQ:NVDA) has now revealed high-end Pascal desktop GPUs, GeForce GTX 1080 and 1070, that will retail at $599 and $379 respectively. Those price points make the new cards slightly more expensive than the ”Founders Edition” that Nvidia plans to launch a little down the line.

Nvidia says that the two GPUs will outperform the company’s Titan X cards which were built on the older Maxwell architecture and retailed at $999. The company says the 1080 will be available on May 27 while the 1070 will be available on June 10. Both GPUs rely on Taiwan Semiconductor’s (NYSE:TSM) 16nm manufacturing process compared to 28nm for older Maxwell chips. 1080 contains 7.2B transistors and will deliver 3 times the power efficiency of Titan X. Both 1080 and 1070 cards rely on GDDR5X and GDDR5 memory by Micron (NSDQ:MU)rather than Fiji’s HBM (High Bandwidth Memory) which accompanies AMD’s (NSDQ:AMD) most powerful cards. HBM is a relatively new technology and was in short supply in 2015 which led to a severe shortage for high-end Nvidia GPUs.

Meanwhile, Nvidia has said that it will use simultaneous multi-projection, a new technology that will help improve VR and multi-monitor gaming performance, in its new cards. Other features include spatial hand controller feedback for VR headsets and Ansel, a solution that allows users to capture screenshots on ongoing games from different scenes at resolutions of as high as 4.5 Gigapixels. Nvidia says Ansel will support VR-friendly 360° panoramic screenshots as well.

Nvidia’s latest GPUs represent a serious step-up in specs and performance compared to the previous generation cards but at a much lower price making the two cards a real bargain. Also notable is the fact that Nvidia unveiled the two cards well ahead of AMD’s launch of its next-gen Polaris GPUs expected to hit the market sometime in mid-2016. Polaris GPUs will use Samsung Electronics (OTC:SSNLF) and GlobalFoundries’ 14nm manufacturing process and support HBM and GDDR5 which AMD promises will deliver big improvements in performance-per-watt. One notable difference between the two and one that might give Nvidia a decisive edge is that Polaris GPUs won’t target high-end desktops but will instead target notebooks and mainstream desktops. AMD says that its Vega architecture, due for launch in 2017, will target high-end desktops.