AMD Radeon RX 9000 Series GPUs Debut: The RX 9070 and RX 9070 XT

Let’s take a quick look at the main technical specifications of these two new GPUs and compare them with their main competitors, the NVIDIA’s GeForce RTX 5070 Ti and RTX 5070.

GPUFrame Generation & UpscalingCUDA Cores/Stream processorsBoost Clock (GHz)Memory
(VRAM)
Memory BandwidthPrice*
RX 9070 XTFSR 440962.9716 GB GDDR6640 GB/s$599
RX 9070FSR 435842.5216 GB GDDR6640 GB/s$549
RTX 5070 TiDLSS 489602.4516 GB
GDDR7
896 GB/s$749
RTX 5070DLSS 461442.5112 GB
GDDR7
672 GB/s$549
Comparison between the AMD Radeon RX 9070 XT and RX 9070 and NVIDIA GeForce RTX 5070 Ti and RTX 5070. *Prices are in US dollars and subject to change, when a price range is available the lowest price is listed. Data from http://www.nvidia.com, marketplace.nvidia.com, and http://www.amd.com.

As shown in the table above, FSR 4 is AMD’s response to NVIDIA’s DLSS 4 technology for upscaling and frame generation. And while the number of main processors in the RX 9070 XT and RX 9070 isn’t as high as in their NVIDIA counterparts, the RX 9070 XT boasts a boost clock frequency about 20% higher.

In terms of dedicated memory, both the RX 9070 XT and RX 9070 feature 16 GB of VRAM, matching the RTX 5070 Ti and exceeding the RTX 5070 by 4 GB. However, AMD’s GPUs still use GDDR6 technology, whereas NVIDIA’s 50-Series GPUs utilize the faster and more efficient GDDR7, which offers higher bandwidth and improved performance in data-intensive tasks.

Regarding power consumption, we could not find directly comparable values from the manufacturers’ technical specifications. This is because NVIDIA lists the total graphics power (TGP, i.e., the maximum power the GPU can theoretically consume), while AMD states the typical board power (i.e., the expected average power draw under typical gaming workloads). However, power consumption comparisons between AMD and NVIDIA GPUs can be drawn from benchmark results (see below).

Interestingly, the recommended market price for the AMD Radeon RX 9070 matches that of its direct competitor, the NVIDIA GeForce RTX 5070, while the AMD Radeon RX 9070 XT undercuts the NVIDIA GeForce RTX 5070 Ti by $150.

So, how do these differences translate into actual gaming performance? Which GPU is worth your hard-earned money? Let’s find out below.

Benchmarks

RX 9070 XT vs RTX 5070 Ti

Results showed that at 1440p, there wasn’t a clear winner regarding average frames per second (FPS). Performance varied depending on the game, with each GPU excelling in different titles.

For example, the RX 9070 XT performed better in Ghost of Tsushima (+18% FPS) and Forza Horizon 5 (+11% FPS) but worse in Kingdom Come: Deliverance II (26% lower FPS) and Cyberpunk 2077 with ray tracing enabled (18% lower FPS).

Without frame generation enabled, both GPUs struggled with demanding games like Cyberpunk 2077 (with ray tracing on) and S.T.A.L.K.E.R. 2, producing average FPS values around 60 or lower.

At 4K resolution, the FPS results remained balanced and game-dependent. The RX 9070 XT performed worse in Black Myth: Wukong (46% lower FPS), Cyberpunk 2077 (25% lower FPS), and Kingdom Come: Deliverance II (18% lower FPS) but better in Ghost of Tsushima (+15% FPS) and Forza Horizon 5 (+14% FPS).

At 4K and without frame generation, neither GPU managed an average FPS above 60 in Black Myth: Wukong, Cyberpunk 2077, Kingdom Come: Deliverance II, Microsoft Flight Simulator 2024, and S.T.A.L.K.E.R. 2.

Keep in mind that enabling frame generation (AMD’s FSR 4 or NVIDIA’s DLSS 4) would significantly improve FPS, allowing for smoother 4K gameplay, but for testing purposes it wasn’t enabled.

Overall, at 1440p the RX 9070 XT had a small performance advantage over the RTX 5070 Ti (+2% FPS on average), but the situation was reversed at 4K, where the RTX 5070 Ti slightly outperformed the RX 9070 XT (+5% FPS on average).

Surprisingly, with the exception of Cyberpunk 2077 (5% lower FPS), the RX 9070 XT generally fared better with ray tracing enabled, especially in Star Wars Outlaws (+26% FPS). This result is surprising as NVIDIA has traditionally led in ray tracing performance, but the RX 9070 XT shows that AMD is closing the gap.

In summary, the performance of these two GPUs is very game-dependent, but in terms of average FPS they are closely matched. However, it’s the RTX 5070 Ti that gets the crown as the more power-efficient GPU, consuming about 35% less power at both 1440p and 4K.

RX 9070 vs RTX 5070

Overall, the RX 9070 held a slight edge over the RTX 5070 in non-ray-traced games (+7% average FPS). The margin was closer with ray tracing enabled (+2% FPS on average). It’s again noteworthy that the RX 9070 generally outperformed the RTX 5070 in ray tracing, even if by a small amount on average.

In terms of power consumption, NVIDIA again took the lead. The RTX 5070 consumed 21% less power without ray tracing and 35% less with ray tracing enabled.

Conclusions

Both the RX 9070 XT and RX 9070 deliver strong performance at competitive prices, with impressive improvements in ray tracing.

The RX 9070 XT performs similarly to its NVIDIA counterpart, the RTX 5070 Ti, but is $150 cheaper. While the RTX 5070 Ti is more power-efficient, the energy savings don’t justify the higher upfront cost. An average gamer playing 2 hours a day would save up to $2 monthly on energy by choosing the RTX 5070 Ti over the RX 9070 XT. It would take over six years to recover the $150 price difference, making the RX 9070 XT the better value.

Regarding the RX 9070, it performs slightly better than the RTX 5070, but with similar pricing and lower power efficiency. This makes it a tie, and the decision on which GPU to buy comes down to brand preference, favored games (given the performance variations), or available discounts.

back to top