Nov 14, 2018 | Updated: 03:14 AM EDT

AMD Radeon RX Vega Leaks: Features, Performance, Potentiality Compared With Nvidia GTX 1080 Ti

May 04, 2017 04:17 PM EDT

Close

AMD has long been under high speculations among tech lovers for its Radeon RX Vega GPU. It has encountered huge round of rumors throughout the timeline. Now tech fans can breathe some fresh air as AMD has itself revealed some potential specs and features which are yet to be seen on this tremendous graphics unit by the later part of this year. The leaks came directly from a new Direct Rendering Manager (DRM) update which was rolled out by AMD for Linux.

According to Game Debate, AMD Radeon RX Vega GPU will boast major advancements which are enough to kill its Radeon RX 480 graphics unit. The graphics unit. AMD's next-gen graphics magnet will come with high-definition performance qualities driven by the second generation High-Bandwidth Memory (HBM 2). The graphical giant will be built on the 14nm Vega 10 XT architecture. As for technical upgradations, AMD Radeon RX Vega would be complimented with four Shader Engines alongside 4096 GCN Stream Processors. These units will supposedly hold 64 next-gen compute units each made with 64 GCN stream processors, collectively making up the whole.

Tweaktown further noted that AMD Radeon RX Vega graphics masterpiece would sport 256 Texture Mapping Units (TMUs) divided into 16 texture blocks, each having 16 texture mapping units. The GPU would also have 64 Render Output Units (ROPs) coupled with 8 GB of HBM2 all placed on a 2048 Bit memory interface. Each 1024 bit stream processor shader engine will be packed with two Asynchronous Compute Units. The memory interface will come with High Bandwidth Cache (HBC).

Apart from all these, the AMD Radeon RX Vega may also beat Nvidia GTX 1080 Ti in terms of performance. At a potential benchmarking, AMD Radeon RX Vega GPU demonstrated 12.5 TFLOPs of FP32 computing performance, which is already greater than GTX 1080 Ti's limit. The later can hardly secure 11.3 TFLOPS performance rate while running on natural state.

©2017 ScienceTimes.com All rights reserved. Do not reproduce without permission. The window to the world of science times.
Real Time Analytics