Xbox Series X Review – Specs

–
–
–

–
–

November 10 is the day, when the arrival of a new generation of consoles is a fact. Microsoft kicks off this time with the introduction of the fourth edition of the Xbox. The new console even comes in two flavors, the Series X and its smaller brother, the Series S. We will discuss the latter another time, now we focus on Microsoft’s standard bearer: the Xbox Series X which will cost 500 euros. We previously published an extensive preview of the new console and we recommend you that article especially to go through to get acquainted with the new console.
After three more weeks of experience with the Series X, it is now time for a balanced judgment, although we immediately reveal that we still do not have a really complete picture of the possibilities of the new console. Before we share our experiences with the Series X, let’s dive into the hardware first. Because what’s in that black fridge?

The Xbox Series X hardware

As with previous generations, the heart of the Xbox Series X and Series S comes from Advanced Micro Devices. AMD is currently the only manufacturer capable of delivering both high-performance x86 processors and GPUs, again in the form of a system-on-a-chip (soc).

What makes the launch of the new generations of consoles extra interesting is that not only the GPU part of the chip shows significant improvements compared to the previous generation, but also the CPU. For gaming, most of the calculations are handled by the GPU. That is where the focus lies when it comes to the question of how graphically complex a game can be made and at what resolution it is rendered. In the Xbox Series X, the GPU also takes up no less than three-quarters of the total surface of the soc. However, a fast GPU does not come into its own in games without a CPU that also has sufficient speed. After all, the processor must provide the GPU with instructions and handle in-game logic itself.

With the previous two generations, exactly that was a challenge: the processor used was on the slow side compared to the GPU. The chips for the Xbox One and One X also came from AMD, and at the time, the manufacturer’s processor offering was nowhere near as competitive as it is today. The Jaguar cores in previous generations of consoles were originally designed for laptops, but proved to be incapable of breaking even in that product category. As a result, it almost became the rule for games to run at 30fps; a higher frame rate was the exception. Although the processor of these generations already had eight cores with two clusters, this was not a complete solution due to the latency-sensitive nature of largely serial instructions. A large number of slow CPU cores is not necessarily a substitute for a smaller number of fast CPU cores in gaming.

Project Scarlett

The Zen 2 processor cores used and the RDNA 2 architecture for the GPU together form part of the SOC of the Series X, also known by the codename Project Scarlett. These parts are both competitive, at least at the time of writing this review. The hardware of the new Xbox generation seems to be a lot more balanced than the previous generations.

Finding the balance in the design of the chips for the new consoles is in any case of significant importance for the end product. If you keep the specifications of the new Xbox next to those of recent PC hardware, you will notice that video cards with many more compute units are available. A new generation of Zen processors has also recently been released and the clock speeds of the new consoles are lagging behind. However, the choice of hardware requires looking beyond raw specifications and accurately identifying what makes sense, not just what is possible. Viewed from the overall cost of a console, one chip with virtually all components integrated on it must be suitable for hundreds of contemporary and future games, have an acceptable energy consumption and heat production, and be cost-effective to mass-produce. All of this must be financially feasible on a more expensive production process than before.

–

Cpu, gpu en ssd

On the CPU side, it is striking that Microsoft is talking about custom Zen 2 cores with a total of 8MB L3 cache, where it is clear that there are no large amounts of L3 cache as on the Ryzen desktop processors. This has everything to do with the fact that the soc of the Series X is not built from chips but from one chip. This one chip is not used as a lower positioned product in the case of less successful or partially defective ones, such as the binning which we almost always see with video cards and processors. Instead, Microsoft has opted for overspecification, just as with the Xbox One X: 52 of the 56 GPU compute units are actually used in the final product.

–

The Zen 2 processor used is also custom in the sense that its memory controllers are linked to the GDDR6 shared memory along with the GPU. As on previous generations, this is therefore a uniform memory type, with a total of 16 GB via a 320-bit memory bus. The clock speeds are when using smt slightly lower than without. Microsoft expects developers to initially prefer the higher clock speeds over the additional processor threads. This will have everything to do with the fact that most games will be developed in such a way that they also work on existing generations of consoles (Xbox One and One X), and these are limited in hardware to 8 threads anyway.

The Project Scarlett GPU was developed on the basis of the RDNA 2 architecture, but we still know relatively little about exactly how various new features have been implemented. There is hardware support for ray tracing, as well as variable rate shading, mesh shaders and sampler feedback. These features are included in AMD’s RX 6000 series graphics cards and Microsoft has indicated that the feature set of this architecture is fully available to developers on both Series X and Series S consoles.

At least as innovative as the more powerful processor and GPU is the storage of the Series X. Under the name Velocity Architecture Microsoft, in collaboration with AMD, has linked fast storage in the form of an NVMe SSD to the soc. Although: we now call it ‘storage’, but the SSD partly fulfills the role of working memory in practice. Combined with the DirectStorage API, the hardware can exchange data at lightning speed between the GDDR6 memory and the SSD. As a gamer, this is particularly helpful in practice if you use the quick resume function: within seconds the exact state of your (paused) game is saved and that of another game is loaded, after which you can start exactly from the point where you can continue playing.

–

Related posts:

iPhone 14 Pro Will Come with 8GB RAM

What is "Metalenz" that achieves optical performance that surpasses that of conventional camera lens...

Samsung Galaxy Watch5 fully leaked ahead of expected August launch

This is the leaked release date of the Samsung Galaxy S22

Share this:

• Facebook
• X

Related