Radeon RX 9070 XT: Real Furmark leak or just a marketing stunt?

AMD seems to be preparing for a new surprise – or at least a cleverly staged leak. The Radeon RX 9070 XT, which is supposedly due on the market in two weeks, has been revealed in a leaked Furmark benchmark. Particularly interesting: the performance is said to be almost on a par with the RX 7900 XTX. Anyone raising their eyebrows now is not entirely wrong, as Furmark is not exactly the benchmark of choice when it comes to realistic gaming performance. Nevertheless, the result raises questions.

Furmark as the new power meter?

Traditionally, Furmark is used as a stress test to push GPUs to the limit – and not as a reliable indicator of gaming performance. Nevertheless, the result that has emerged gives a first impression of the raw performance of the RX 9070 XT. The card is said to achieve up to 7979 points, which puts it clearly above the RX 7900 XT, RX 7900 GRE and over 50 % faster than the RX 7800 XT.

Particularly interesting: According to current rumors, AMD itself is promising a performance increase of over 40% in 4K gaming compared to the RX 7900 GRE. If these figures are confirmed, the RX 9070 XT could become a real price-performance candidate – provided AMD does not exaggerate in the official presentation slides.

https://x.com/GawroskiT/status/1892927711242232071

Camouflaged hardware or targeted manipulation?

According to the leaker Tomasz Gawroński, the card is currently running with a manipulated driver ID that displays it as the RX 7800 XT. In reality, however, it has a unique hardware ID of the RX 9070 series. Why would someone do this? Either it’s a deliberate cover-up by AMD to deliberately cause confusion, or there are pre-production models in circulation that are being tested with customized drivers. This would not be the first time that unofficial drivers have been optimized for benchmarks in order to build up a certain expectation at an early stage. Another possibility would be that AMD is already running more aggressive performance optimization with the latest drivers – although this could also indicate potential problems with efficiency or power consumption.

Classification: Marketing, reality or wishful thinking?

The question remains as to how much of this leak will actually hold up in practice. Furmark is not a gaming benchmark, and real games could tell a completely different story. The RX 7900 XTX has a significantly higher memory bandwidth and more compute units – so can an RX 9070 XT with a smaller memory interface and lower raw performance really get that close to the top? AMD will officially unveil the architecture in a week’s time, and by then at the latest it will be clear how much truth there is behind the current figures. If the RX 9070 XT really is that strong, it could position itself as an attractive alternative in the high-end segment – if AMD doesn’t nip all hopes in the bud with exaggerated price expectations.

Until then, stay calm and wait for real gaming benchmarks. Furmark is nice, but in the end it’s the performance in real games that counts – and not how well a GPU performs in a stress test.

Source: Furmark, GawroskiT, Geeks3d VideoCardz