Three technologies of the gaming world

By Admin On Dec 6, 2022

The world of gaming is now dominated by three big companies. These three add up to hundreds of millions of graphics cards and are responsible for bringing revolutionary advances to the industry. Today we are going to see how their proposals differ.

2022 is a year for GPU history . After years of dominance by NVIDIA and attempts by AMD to compete on the same level, Intel made good on its promise to re-enter the graphics card market. Now we have NVIDIA, AMD and Intel competing for a pie that is very big.

If we compile what has happened so far this 2022, we have NVIDIA with its RTX 4000 series , how expensive and powerful they are; we have AMD with their RX 7000 graphics card due to make an appearance this month; and then we are left with Intel, who has launched the Intel Arc as a mid-range proposal.

Having seen this, let’s bring all the manufacturers to the table and dive deeper into their architectures , peeling back the marketing layers to see what’s new in graphics cards , what they have in common, and what it all means for the average user.

In this article you will find the most important of each company, which will allow you to differentiate between each brand, its architecture and the chip that makes it possible. Without further ado, we get into the matter that there is a lot of graphics card to cut:

AMD and RDNA 3: the Navi 31 chip

Navi 31 is the largest chip ever built under the RDNA 3 architecture so far. Compared to the Navi 21, the new SoC promises to compete head-to-head with NVIDIA and its RTX 4000 .

Shader Engines (SE) host fewer Compute Units (CUs), 16 vs. 200, but there are now 6 SEs in total, two more than before. This means that Navi 31 has up to 96 CUs, with a total of 6,144 Stream Processors (SP) from the Radeon 7900 XTX.

Each shader engine also contains a dedicated raster unit, a primitive engine for triangle configuration, 32 Render Output Units (ROPs), and two 256KB L1 caches.

AMD hasn’t changed much to the raster engines and primitives either: the claimed improvements of 50% are for the entire chip, as it has 50% more shaders than the Navi 21 chip.

The most obvious change is the one that sparked the most buzz and gossip before the November release: the chiplet approach to the GPU package . With several years of experience in this field, it’s only logical that AMD would have opted for this, but it’s entirely for cost/manufacturing reasons, rather than performance.

In Navi 31, the memory controllers and their associated end-level cache partitions are housed in separate chiplets (called MCD, or Memory Cache Dies) surrounding the primary processor (GCD, Graphics Compute Die).

With more SEs to feed, AMD has also increased the number of MCs by 50%, so the total bus width to global GDDR6 memory is now 384 bits. There’s less total Infinity Cache this time around (96MB vs. 128MB), but the higher bandwidth makes up for it.

Intel and its ACM-G10 chip

We move on to Intel and the ACM-G10 matrix (previously called DG2-512). While it’s not the largest GPU Intel makes, it’s their largest graphics matrix for customers like you and us (regular users and not servers).

The block diagram is a fairly standard layout, although it looks more like Nvidia’s than AMD’s. With a total of 8 Render Slices, each containing 4 Xe-Cores, for a total of 512 Vector Engines (Intel’s equivalent to AMD’s Stream Processors and Nvidia’s CUDA cores).

In addition, each Render Slice contains a primitive unit, a rasterizer, a depth buffer renderer, 32 texture units, and 16 ROPs. At first glance, this GPU seems quite large, with 256 TMUs and 128 ROPs being more than what is found on a Radeon RX 6800 or GeForce RTX 2080.

However, AMD’s RNDA 3 chip houses 96 compute units, each with 128 ALUs, while the ACM-G10 features a total of 32 Xe cores, with 128 ALUs per core. Thus, in terms of the number of ALUs alone, Intel’s Alchemist-powered GPU is one-third the size of AMD’s.

NVIDIA and its AD102 chip, the jewel of the RTX 4000

We end our review of the different designs with NVIDIA’s AD102, their first GPU to use the Ada Lovelace architecture . Compared to its predecessor, the Ampere GA102, it doesn’t look that different, just a lot bigger. And for all intents and purposes, it is.

NVIDIA uses a component hierarchy of a Graphics Processing Cluster (GPU) that contains 6 Texture Processing Clusters (TPCs), each hosting 2 Stream Multiprocessors (SMs).

In the full AD102 die, the number of GPCs has gone from 7 to 12, so there are now a total of 144 SMs, giving a total of 18,432 CUDA cores on the RTX 4090 and slightly less on the RTX 4080. This may seem like a ridiculous number when compared to Navi 31’s 6,144 SM, but AMD and NVIDIA count their components differently.

Although this is oversimplifying things, an NVIDIA SM is equivalent to an AMD CU: both contain 128 ALUs. So while the Navi 31 is twice the size of Intel’s ACM-G10, the AD102 is 3.5 times the size.

That’s why it’s unfair to compare the performance of chips when they’re so clearly different in terms of scale. However, once you’re inside a graphics card, with its price and marketing, things change.

With all this data on the table, you should now be able to differentiate how each brand approaches their approach to gaming and the technologies that make it possible to play games in 4K and with ray tracing. The world of GPUs is not easy, but it is very interesting.

technologies of the gaming world