AMD’s best graphics cards of this generation all fall under the RDNA 4 umbrella. Also known as the RX 9000 series, these GPUs bring notable improvements over the previous generation, but they also don’t quite manage to beat the top performance offered by AMD’s last-gen halo card, the RX 7900 XTX.
Here’s everything you need to know about the AMD Radeon RX 9000 series graphics cards.
AMD’s RDNA 4 lineup is still fairly small, especially when compared to Nvidia’s RTX 50-series. Right now, only four graphics cards are widely available, with two more models having limited availability.
The RX 9070 XT and non-XT launched on March 6, 2025. The XT model carried a recommended list price (MSRP) of $599, and the non-XT variant sold at $549. Unfortunately, the pricing didn’t last long, and the cards were selling above MSRP for a long time.
Meanwhile, the RX 9060 XT (available in two variants — one with 8GB of VRAM and one with 16GB) launched on June 5, 2025. The 8GB version starts at $299, and the 16GB model starts at $349.
Beyond those two, AMD has also launched the RX 9070 GRE and the RX 9060 non-XT. The RX 9070 is, as of right now, a China-only product, and the RX 9060 is only available to system builders.
Leading up to RDNA 4, it was widely known that AMD would be sticking to the midrange with this generation. Instead of making a flagship card, AMD decided to keep its specs rather mainstream, and this shows.
The RX 9070 GRE and the RX 9060 aren’t really available, so we’ll skip them. However, one important thing to note across the entire lineup is that AMD chose to stick to GDDR6 memory in this generation. Meanwhile, Nvidia uses the faster, newer GDDR7 VRAM modules, increasing bandwidth.
AMD’s top two GPUs offer a decent amount of video memory for the price (although, unfortunately, they don’t often sell at MSRP). When it comes to the RX 9060 XT, we’re encountering the same problem as the one we’ve seen in Nvidia’s RTX 5060 Ti — the version with 8GB VRAM is much less popular and an overall worse deal.
The RX 9070 XT and non-xt are midrange GPUs, with specs that allow them to rival Nvidia’s RTX 5070 and RTX 5070 Ti. The cheaper GPUs are targeting Nvidia equivalents lower down the stack, such as the RTX 5060 Ti and the RTX 5060.
We haven’t had the chance to test these GPUs ourselves just yet, but based on reputable publications, it’s safe to say that the RX 9070 XT and non-XT were praised upon launch. In a way, they appear to have done better than the more powerful GPUs from the RDNA 3 lineup did back when they were first released.
The RX 9070 XT, priced at $549, was an especially great deal when matched up against Nvidia’s RTX 5070 Ti. The two cards are seen trading blows on occasion, and even when the Nvidia GPU pulls ahead, that’s still not enough to justify a $200 price gap.
Unfortunately, AMD’s lead didn’t last long, as once the initial stock of RX 9070 XTs sold out, it came back at a higher price. With the two cards both hovering in the $700 to $800 range, picking Nvidia made more sense for many, if only due to the better performance of DLSS 4 when compared to FSR 4 — more on that later.
When we look at raw performance, though, the RX 9070 XT is a solid graphics card. TechRadar’s John Loeffler, who called the card “exactly what the market needs,” benchmarked the card against Nvidia options. At 1440p, the RX 9070 XT turned out to be around 8.4% faster than the RX 7900 XTX, and around 7% slower than the RTX 5070 Ti. Assuming the MSRP holds, that is a good result.
AMD also took the lead at 4K, beating the RTX 4070 Ti by 15.5%, but losing to the RTX 5070 Ti by 8%.
Meanwhile, the 8GB RX 9060 XT was around 4% slower than the RTX 5060 at 1080p. The RX 9060 XT 16GB is slightly slower than the RTX 5060 Ti 16GB, falling around 9% behind in TechRadar’s review of the two cards.
AMD’s FidelityFX Super Resolution 4 (FSR 4) is the latest generation of its proprietary upscaling and frame generation tech. For the first time in AMD’s history, the tech is not vendor-agnostic and is locked to a single generation of graphics cards. It cannot be enabled outside of RDNA 4 right now, although AMD is reportedly working on adding support for FSR 4 on older GPUs.
FSR 4 is a machine learning-based upscaler. AMD promises to deliver both higher frame rates thanks to frame generation, but also an improvement in image quality when compared to FSR 3.1. Outside of that, expect to see less ghosting, better details, and more stability.
FSR 4 is available in various quality modes, including Native AA, Quality, Balanced, Performance, and Ultra Performance. Which of these will be available to you depends on the game, though.
The tech isn’t just locked to a single generation of GPUs; it’s also only available in around 85 games at the time of writing. However, that list covers some of the biggest hits right now, and AMD is steadily expanding support in an effort to keep up with Nvidia’s rival technology, DLSS 4.
AMD’s RDNA 4 architecture powers the Radeon RX 9000-series and represents a significant leap in efficiency and per-core performance over RDNA 3.
At its core, RDNA 4 introduces a redesigned compute unit structure with up to 64 CUs, each benefiting from enhanced AI and ray tracing accelerators. The 3rd-generation ray accelerators double ray-tracing throughput compared to RDNA 3, while the 2nd-generation AI accelerators add support for FP8 and INT4 operations, delivering up to eight times the throughput in certain sparse workloads. These upgrades directly fuel new features like FSR 4, giving you more of those sweet, sweet, AI-generated frames for your enjoyment.
Alongside raw compute improvements, RDNA 4 integrates updated display and media engines. The Radiance Display Engine supports DisplayPort 2.1a and HDMI 2.1b, while the Enhanced Media Engine improves AV1, HEVC, and H.264 encode and decode efficiency for streamers and creators.
Manufactured on TSMC’s 4nm N4P process, Navi 48 and Navi 44 chips balance performance with power efficiency, allowing AMD to deliver competitive GPUs across the midrange even without releasing a so-called halo GPU. As a result of these architectural changes, AMD is able to keep up in this generation even without delivering much of an upgrade in terms of raw specifications.
Nearly a month after announcing the GPT-6.1 Sol model, OpenAI unveiled GPT-6.1 Terra today. The latest addition to the family focuses on computer use, visual reasoning, coding, and professional work, while costing 50% less through the API than GPT-5.6 Terra.
OpenAI says Terra can handle everything from filling out browser forms and working inside spreadsheets to interpreting technical drawings and editing videos. The company says it can also complete computer-based tasks more than three times faster than GPT-6 Astra and GPT-6.1 Sol.
ASUS has started selling its UGen Intel Arc Pro B70 graphics card in Europe, bringing its take on Intel’s workstation GPU to customers with a hefty €1,799 recommended price. The card is available in Germany and Austria, while ASUS lists it at CHF 1,799 in Switzerland.
Intel first introduced the Arc Pro B70 in March 2026, with its own branded model starting at $949. ASUS’s version arrives several months later with the same core GPU architecture but a different cooling design and a higher power rating. The price also places it well above Intel’s original starting point, although the two prices are not directly comparable because ASUS’s European figure includes VAT.
Anthropic has officially filed to go public, and buried inside its S-1 is a blunt warning about the technology at the center of its business. The maker of Claude AI says advanced AI could cause everything from large-scale cyberattacks and geopolitical instability to catastrophic or even existential risks for humanity.
The warnings take up a surprisingly large part of the filing. Axios notes that more than 90 pages, or roughly a third of the prospectus, focus on risk factors. Many of them warn about what could happen as AI systems become more capable and autonomous.