Luckily, during its Financial Analysts Day call this afternoon, AMD spilled some more information on Vega, at least from a pro graphics standpoint. The aim with Vega is to deliver 4K gaming performance at 60 FPS, and the Radeon RX Vega, the consumer GPU variant, is claimed to deliver on that front when it arrives in market. Although Raja Koduri, Senior Vice President and Chief Architect for Radeon Technologies Group, was light on specifics for the upcoming graphics card, he did provide the following teaser for Radeon Vega Frontier Edition, again the professional workstation version:
These figures are compared to the Radeon R9 Fury X, which is based on the 28nm Fiji core. That previous-generation high-end GPU delivered 8.6 TFLOPs of compute performance compared to 13 TFLOPs for the Radeon Vega Frontier Edition. FP16 performance is also through the roof, coming in at 25 TFLOPs.
AMD has taken memory all the way to 16GB of HBM2, making this a new high-watermark for single-GPU workstation boards. For comparison, the NVIDIA GeForce GTX 1080 Ti delivers 11.4 TFLOPs of compute performance and packs in 11GB of GDDR5X memory. However, we should clearly underscore that Radeon Vega Frontier Edition is targeted fr pro graphics workloads and machine learning tasks according to AMD, including:
- Machine Learning: Together with AMD’s ROCm open software platform, Radeon Vega Frontier Edition enables developers to tap into the power of Vega for machine learning algorithm development. Frontier Edition delivers more than 50 percent more performance than today’s most powerful machine learning GPUs.
- Advanced Visualization: Radon Vega Frontier Edition provides the performance required to drive increasingly large and complex models for real-time visualization, physically-based rendering and virtual reality through the design phase as well as rendering phase of product development.
- VR Workloads: Radeon Vega Frontier Edition is ideal for VR content creation supporting AMD’s LiquidVR technology to deliver the gripping content, advanced visual comfort and compatibility needed for next-generation VR experiences.
- Revolutionized Game Design Workflows: Radeon Vega Frontier Edition simplifies and accelerates game creation by providing a single GPU optimized for every stage of a game developer’s workflow, from asset production to playtesting and performance optimization.
- Compute Units: 64
- Single Precision compute performance (FP32): ~13 TFLOPS
- Half Precision compute performance (FP16): ~25 TFLOPS
- Pixel Fillrate: ~90 Gpixels/sec
- Memory Capacity: 16 GBs of High Bandwidth Cache
- Memory Bandwidth: ~480 GB/s
AMD Radon Vega Frontier Edition graphics cards will begin shipping in late June. Unfortunately, this was just a small taste of what we can expect with regards to Vega RX gaming cards when they appear in market. AMD will reportedly unleash complete details on Vega on March 31st at Computex in Taipei, Taiwan.
Editor's note, 5/16/17 8:35PM: This article has been updated to correct messaging regarding the Radeon Vega Frontier Edition positioning in market, which is specifically for workstation pro graphics, VR, visualization and machine learning workloads.