AMD Radeon RX 7800 XT & RX 7700 XT Official Benchmarks Leak: 7800 XT 1% Faster Than 4070, 7700 XT 9% Faster Than 4060 Ti

AMD is all set to launch its Radeon RX 7800 XT & RX 7700 XT RDNA 3 GPUs tomorrow & the official benchmarks have once again leaked out.

AMD's Official "Reviewers Guide" Benchmarks Show Single-Digit Gains For Radeon RX 7800 XT & RX 7700 XT GPUs Over Their Competition But Offer Better Value

The official benchmarks come from within the AMD "Reviewers Guide" which is seeded to press as a reference of what kind of performance they can expect with the graphics card that they are given for testing. In this case, we are looking at the performance benchmarks of the Radeon RX 7800 XT & the RX 7700 XT which are leaked by HDTechnologia and Videocardz, covering various games and synthetic performance tests.

Although the exact system confirmation isn't mentioned, we are likely looking at an AMD Ryzen 7000 CPU running DDR5 memory on an AM5 motherboard. The performance was measured in three different scenarios, first is the standard gaming rasterization, then we have the ray tracing and finally, we have the synthetic benchmarks. For comparison, the AMD Radeon RX 7800 XT has been compared to the RTX 4070 which costs $100 US more ($599 US vs. $499 US) and the 7700 XT has been compared to the RTX 4060 Ti which costs the same as the recent price cuts ($449 US).

Starting with the standard rasterization performance, we see that the AMD Radeon RX 7800 XT is on average 6,88% faster than the RTX 4070 Ti across the 17 titles tested while the RX 7700 XT is on average 15.9% faster than the RTX 4060 Ti. All games were tested out on 1440P and rasterization is usually a strong suit with the AMD Radeon cards however modern titles are relying heavily on an advanced set of graphical features such as ray tracing and that's evident by the growing use of the technology in consoles which are powered by AMD's RDNA architecture.

In ray tracing, the AMD Radeon RX 7800 XT is -11.55% slower than the RTX 4070 & takes a large hit in Cyberpunk 2077 where the performance falls by -69%. The RX 7700 XT ended up -5.4% slower than the RTX 4060 Ti 16 GB and while Cyberpunk 2077 is definitely an outlier here, most of these games tested are running Ray Tracing that is optimized around RDNA 2 and RDNA 3 hardware so adding more NVIDIA-optimized ray traced titles to the mix can result in even bigger losses in the RT department for camp red. Summing up the results:

AMD Radeon RX 7800 XT 16 GB ($499 US) vs NVIDIA RTX 4070 12 GB ($599US):

• Rasterization Performance: +6.88%
• Ray Tracing Performance: -11.55%
• Overall Performance: +0.5% Faster

AMD Radeon RX 7700 XT 12 GB ($449 US) vs NVIDIA RTX 4060 Ti 16 GB ($449US):

• Rasterization Performance: +15.9%
• Ray Tracing Performance: --5.44%
• Overall Performance: +8.5% Faster

Lastly, there are the synthetic performance benchmarks which once again show AMD taking the cake in rasterized benchmarks such as 3DMark Fire Strike, 3DMark 11, and Time Spy but things change as soon as Port Royal benchmark is conducted. The Speed Way benchmark is even more heavy when it comes to ray tracing so expect both the AMD Radeon RX 7800 XT & RX 7700 XT to be slower than the competition in those too.

Looking at the benchmarks, the AMD Radeon RX 7800 XT still offers much better value and comes with a Starfield bundle which sweetens the deal. You will also get advanced FSR 3 support but the Radeon RX 7700 XT might be a tough one. Sure it does offer faster performance than the 4060 Ti 16 GB but the recent price cuts and the availability of the older RX 6800 16 GB GPUs can hurt this card. Both cards are coming out this Wednesday and will have various models to select form such as the ones we rounded up here.

AMD Radeon RX 7000 RDNA 3 GPU Lineup:

Graphics Card	AMD Radeon RX 7900 XTX	AMD Radeon RX 7900 XT	AMD Radeon RX 7900 GRE	AMD Radeon RX 7800 XT	AMD Radeon RX 7700 XT	AMD Radeon RX 7700 (NEW)	AMD Radeon RX 7600 XT	AMD Radeon RX 7600
GPU	Navi 31 XTX	Navi 31 XT	Navi 31 LE	Navi 32	Navi 32	Navi 32	Navi 32	Navi 33 XL
Process Node	5nm+6nm	5nm+6nm	5nm+6nm	5nm+6nm	5nm+6nm	5nm+6nm	5nm+6nm	6nm
Die Size	300mm2 (Only GCD) 522mm2 (with MCDs)	300mm2 (Only GCD) 522mm2 (with MCDs)	300mm2 (Only GCD) 522mm2 (with MCDs)	~200mm2(Only GCD) 150mm2(with MCDs)	~200mm2(Only GCD) 150mm2(with MCDs)	~200mm2(Only GCD) 150mm2(with MCDs)	204mm2	204mm2
Transistors	58 Billion	58 Billion	58 Billion	28.1 Billion	28.1 Billion	28.1 Billion	13.3 Billion	13.3 Billion
GPU WGPs	48	42	40	30	27	20	16	16
Stream Processors	6144	5376	5120	3840	3456	2560	2048	2048
TMUs/ROPs	384 / 192	384 / 192	320 / 160	240 / 96	216 / 96	160/64	128 / 32	128 / 32
Game Clock	2.3 GHz	2.0 GHz	1.9 GHz	2.2 GHz	2.1 GHz	2.0 GHz?	2.47	2.2 GHz
Boost Clock	2.5 GHz	2.4 GHz	2.2 GHz	2.5 GHz	2.4 GHz	2.45 GHz?	2.76	2.6 GHz
FP32 TFLOPs	61 TFLOPs	52 TFLOPs	46 TFLOPs	37 TFLOPs	35 TFLOPs	25 TFLOPS	23 TFLOPs	22 TFLOPs
Memory Size	24 GB GDDR6	20 GB GDDR6	16 GB GDDR6	16 GB GDDR6	12 GB GDDR6	16 GB GDDR6	16 GB GDDR6	8 GB GDDR6
Infinity Cache	96 MB	80 MB	64 MB	64 MB	48 MB	40 MB	32 MB	32 MB
Memory Bus	384-bit	320-bit	256-bit	256-bit	192-bit	256-bit	128-bit	128-bit
Memory Clock	20 Gbps	20 Gbps	18 Gbps	19.5 Gbps	18 Gbps	19.5 Gbps	18 Gbps	18 Gbps
Bandwidth	960 GB/s	800 GB/s	576 GB/s	624 GB/s	432 GB/s	624 GB/s	288 GB/s	288 GB/s
Effective Bandwidth	3.5 TB/s	3.5 TB/s	2.2 TB/s	2.7 TB/s	2.0 TB/s	1.9 TB/s	476.9 GB/s	476.9 GB/s
TBP	355W	315W	260W	263W	245W	263W	190W	165W
PCIe Interface	PCIe 4.0 x16	PCIe 4.0 x16	PCIe 4.0 x16	PCIe 4.0 x16	PCIe 4.0 x16	PCIe 4.0 x16	PCIe 4.0 x8	PCIe 4.0 x8
Price	$999 US	$899 US	$549 US	$499 US	$419 US	TBD	$329 US	$269 US
Launch	Dec, 2022	Dec, 2022	July, 2023	September, 2023	September, 2023	September, 2025	January, 2024	May, 2023

Follow Wccftech on Google to get more of our news coverage in your feeds.