The Nvidia RTX 4080 and 4090 GPUs deliver amazing performance. They’re also surprisingly expensive, starting at $1,200 and $1,500 and up for cards from partners like MSI, Gigabyte, and Asus. The 4080 is almost twice as expensive as the original MSRP of $699 for the RTX 3080.
These price spikes are driven in part by pandemic-era fears such as supply chain disruptions and inflation, and in part by the crypto-driven boom (now over, thankfully) that prompted a network of speculators to snap up every single high-end GPU they could. The game was also affected by the lack of competition and the increasing cost and complexity of building giant monolithic chips in advanced manufacturing processes. Today, AMD is trying to address the last two issues by releasing its Radeon RX 7900 series GPUs.
At $899 and $999, the RX 7900 XT and RX 7900 XTX are still objectively expensive, but since they’re not a further increase in the RX 6900 XT’s starting price, both cards are considered worthwhile in today’s GPU market. If you’re looking for cards that can consistently handle 4K games at 60fps and beyond, these GPUs do it for less than Nvidia’s latest, and they’re good enough and fast enough to hopefully start slashing Nvidia’s prices a bit., too much.
But Nvidia still retains some of the key strengths that make the easy David and Goliath storytelling difficult. These GPUs don’t quite look like Ryzen’s moment for AMD’s graphics division – a pivotal moment when the rogue AMD manages to pull hard from the market share of an entrenched, complacent competitor. But if you can actually find them at their starting prices, that’s the first sign that we’ve got some relief for high-end but price-sensitive PC gamers.
Big Navi
The RX 7000 series is the third version of the RDNA GPU architecture, also sometimes referred to as “Navi”after the code names for the GPU chips themselves. RDNA 3 doesn’t add anything as significant as RDNA 2’s ray tracing support, but AMD has added a lot of additional hardware and made important internal changes.
Most important is the new chipset-based approach, similar in concept to what AMD is using for its Ryzen processors. Instead of creating the entire GPU die in a single manufacturing process—increasing the die size and thus the chance that some or all of them could be defective—AMD is creating the main Navi 3 GPU die based on a 5nm manufacturing process. TSMC process and a number of smaller memory controller (MCD) dies in the 6nm process. All of these chips are linked together by a high-speed connection that AMD says can transfer data at up to 5.3 terabytes per second.
The main graphics compute die (GCD) contains most of the hardware that you think of when you think of a GPU – compute units, shaders, ray tracing hardware, media encoding and decoding unit, and display output. Both the 7900 XTX and XT use the same Navi 31 GCD, but the XTX runs at higher clock speeds and has more CUs and stream processors included. XTX has 96 CUs and 6144 stream processors, while XT has 84 CUs and 5376 stream processors. Both cards represent a jump up from the Navi 21 die used in the RX 6900 series, which peaked at 80 CUs and 5120 stream processors (and that’s not counting other performance-enhancing improvements).
All MCDs include a single 64-bit memory controller and 16MB of AMD Infinity Cache and demonstrate the benefits of a chip-based approach. The 7900 XTX has a 384-bit memory bus and 96MB of Infinity Cache, while the 7900 XT has a 320-bit bus and 80MB of cache; to do this, all AMD has to do is remove the MCD. The same MCDs can be reused up and down the stack with all the different RDNA 3 GCDs that AMD chooses to release, from low-end products with a single MCD to mid-range GPUs that use two to four. Defects in MCDs will not require larger, more complex GCDs to be discarded or removed, and vice versa.
In the area of brand new features for RDNA 3, there are three things to note. First, GPUs include new AI accelerators, which can be useful for both creating massive amounts of content with AI that came out last year, and scaling with AI (if AMD decides to implement it in which or a future version of its software). FSR scaling algorithm; both DLSS and XeSS use AI for scaling, while FSR 2.0 does not).
Secondly, the video encoder/decoder supports hardware accelerated encoding for the AV1 video codec, as does the RTX 4000 series and Intel Arc GPUs. This should be useful for both content creators and streamers who want to stream higher resolution videos or the same resolution videos using less bandwidth.
And third, “Radiance Display Engine”adds DisplayPort 2.1 support for GPUs. At the time of this writing, monitors that take full advantage of DisplayPort 2.1’s extra bandwidth don’t really exist, but when they do, RDNA 3 GPUs will be able to drive 4K displays at up to 480Hz and 8K displays at up to 165Hz. .