Introduction

AMD today launched its flagship desktop processor, the Ryzen 9 7950X3D, enhanced with the phenomenal 3D Vertical Cache technology that worked like magic on 5800X3D, providing a generational gaming performance uplift. You can read all about the 7950X3D "Zen 4" processor in our main review. AMD did something interesting with the way it released 3D Vertical Cache to the market this time around—the company launches the 16-core 7950X3D and 12-core 7900X3D now, which go on sale from tomorrow, February 28, while the 8-core 7800X3D joins the lineup on April 8.

The 7950X3D and 7900X3D are high core-count processors designed to take the fight to Intel's "Raptor Lake" Core i9-13900K and Core i7-13700K, respectively, but they're not exactly cheap. The 7950X3D is priced at $700, while the 7900X3D goes for $600. The Socket AM4 5800X3D continues to grab attention among people still on the old platform, as it offers gaming performance that's easily in the league of 12th Gen "Alder Lake" processors—good enough to avoid a total system upgrade. This also means that gamers are happy to buy an 8-core/16-thread processor, and 8 cores is really all you need for gaming. Even Intel agrees, and packs its desktop processors with no more than 8 P-cores that typically execute gaming workloads.



AMD's decision to defer the launch of the 8-core 7800X3D appears to be more of a business-minded one: Why not push a few 12-core and 16-core 7000X3D processors before bringing in the $450 8-core part that could potentially cannibalize them? The 7950X3D and 7900X3D achieve their core-counts using two 8-core CPU complex dies (CCDs), or chiplets. Only one of the two pack 3D Vertical Cache, while the other is a regular "Zen 4" CCD. The one with 3D Vertical Cache has a total last-level cache size of 96 MB; while the other CCD has 32 MB. This way, gaming workloads that are content with 8 cores, are localized to the CCD with the 3D Vertical Cache, while the other CCD only steps in as needed. For non-gaming multi-threaded workloads, you get the raw compute muscle of both CCDs, and the 24-thread or 32-thread parallelism they bring.



With launch of the 7800X3D months away, and a 7950X3D sample on hand, we felt the itch to play with the chip a little. The UEFI setup program of our motherboard lets us disable CPU cores one by one, so we shut off the entire second CCD—the one without the 3D Vertical Cache—essentially turning this into a 7800X3D. The resulting processor would be completely disencumbered from the need for software-level optimization, such as that from the 3D V-cache Optimization driver or PPM driver, which collaboratively optimize gaming workloads, such that gaming workloads are localized to the CCD with the added cache.

In this review, we present performance numbers of a Ryzen 9 7950X3D processor with its second CCD disabled, leaving it with just the one with the 96 MB L3 cache. This contraption isn't a perfect simulation of a 7800X3D—the clock-speeds of the 7800X3D will be lower according to AMD's announcement, and boosting behavior may be slightly different. These numbers will still give you a fairly good idea of what to expect from the 7800X3D, and whether it's worth waiting a month to save some money.

Test Setup

• All applications, games, and processors are tested with the drivers and hardware listed below—no performance results were recycled between test systems.
• All games and applications are tested using the same version.
• All games are set to their highest quality setting unless indicated otherwise.

Super Pi

SuperPi is one of the most popular benchmarks with overclockers and tweakers. It has been used in world-record competitions practically forever. It is a purely single-threaded CPU test that calculates Pi to a large number of digits—32 million for our testing. Released in 1995, it only supports x86 floating-point instructions and thus makes for a good test for single-threaded legacy application performance.

Y-Cruncher

Y-Cruncher is a highly optimized piece of software that can calculate Pi and other constants to a huge number of digits. It is fully multi-threaded, uses a modern code design and is optimized for all major processor architectures. This ability has made it a popular application, used by the enthusiast community to determine and compare how powerful their overclocked systems are.

Chess Simulation

Creating a worthwhile opponent for the game of chess has been one of the most fascinating tasks for computer programmers for decades. In this test we're using the highly popular Stockfish chess engine to calculate optimal moves in a chess scenario.

Rendering — Cinebench

Cinebench is one of the most popular modern CPU benchmarks because it is built around the renderer of Maxon's Cinema 4D software. Both AMD and Intel have been showing this performance test at various public events, making it almost an industry standard. Using Cinebench R23, we test both single-threaded and multi-threaded performance.

Rendering — Blender

Blender is one of the few professional-grade rendering programs out there that is both free and open source. That fact alone helped build a strong community around the software, making it a highly popular benchmark program due to its ease of use as well. For our testing, we're using the Blender "BMW 27" benchmark scene with Blender 3.2.