AMD Levels Up Gaming GPUs, Providing Mid

Jul 19, 2023

AMD has revealed two new members of its RX 7000 series of graphics cards that may compete with Nvidia's 4000 series cards. These new members, the RX 7700 XT and RX 7800 XT, fill a large gap in the GPU ecosystem and give consumers new options between entry-level and enthusiast-grade graphics cards.

AMD claims the new cards may benefit users and designers at a potentially lower price per performance. However, depending on the specific application, the Nvidia 4000 series cards may still provide performance gains over their AMD counterparts.

This article dives into the performance specs of the new AMD RX 7000 cards to give designers a sense of how the cards stack up against competitors. In addition, we'll discuss other uses outside gaming to demonstrate the broad applicability of the latest GPUs.

The RX 7700 XT and RX 7800 XT fill a major hole in AMD's RX 7000 lineup. Announced in May 2023, the RX 7600 was designed as a budget-friendly solution, while the RX 7900 XT and XTX, both announced in November 2022, were designed as top-of-the-line graphics cards.

The step-up from 7600 to 7900 XT is certainly considerable, both in price and performance, leaving consumers and designers fewer options for a middle-of-the-road solution. With the release of the 7700 XT and 7800 XT, however, there are now many more options when shopping for a parallel processor.

In certain circumstances, the new AMD cards may also outperform Nvidia’s 4070 and 4060 Ti. This ultimately provides designers with more options when selecting a GPU based on their specific needs.

Both the 7700 XT and 7800 XT are built on AMD's RDNA 3 architecture, providing critical new features, including chiplet-based designs, AI acceleration, and the AMD Radiance Display Engine for improved color accuracy.

From in-house tests, AMD reports that this architecture yields a 20% performance-per-dollar upgrade compared to direct competition. And while the raw performance of the devices may only be a modest improvement compared to previous generations, the new technologies could certainly open new applications for the cards.

In the case of AI workloads, for example, AMD has reported increased instruction efficiency and data throughput, giving the RDNA 3 cards over double the performance compared to RDNA 2 and offering new methods for increasing frame rates. For designers working in machine learning or AI, these improvements are compounded by the increased AI workload efficiency.

While it's common to associate consumer GPUs exclusively with gaming, AMD’s RDNA 3-based graphics cards can improve AI performance in several other use cases as well. As parallel processing becomes important to more applications, the availability of a cost-effective GPU could open doors to other innovative use cases.