AMD Radeon RX 6800 XT vs NVIDIA GeForce RTX 4070 Comparison

AMD
RADEON

AMD Radeon RX 6800 XT

CORE STATE Navi 21
VRAM 16 GB
CLOCK SPEED 2250 MHz
TDP 300 W
BUS WIDTH 256 bit
ARCHITECTURE RDNA 2.0
nm
PROCESS 7 nm
LAUNCH DATE 2020
VS
NVIDIA
GEFORCE

GeForce RTX 4070

CORE STATE AD104
VRAM 12 GB
CLOCK SPEED 2475 MHz
TDP 200 W
BUS WIDTH 192 bit
ARCHITECTURE Ada Lovelace
nm
PROCESS 5 nm
LAUNCH DATE 2023

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12
3,689
3,854
geekbench_metal
181,337
N/A
geekbench_opencl
171,304
154,858
geekbench_vulkan
136,875
174,152
passmark_directx_10
155
139
passmark_directx_11
264
244
passmark_directx_12
100
103
passmark_directx_9
265
320
passmark_g2d
1,030
1,164
passmark_g3d
24,980
26,927
passmark_gpu_compute
13,245
14,720

Analysis: AMD Radeon RX 6800 XT vs NVIDIA GeForce RTX 4070

Head-to-Head Benchmarks

The head-to-head comparison between the AMD Radeon RX 6800 XT and the NVIDIA GeForce RTX 4070 shows a clear split across different workloads. The RTX 4070 takes seven of the ten recorded tests, while the RX 6800 XT secures three victories. The most decisive win for either card belongs to NVIDIA in the Geekbench Vulkan test, where the RTX 4070 scores 174,152 against 136,875 for the RX 6800 XT, a 21.4% advantage. This is a substantial gap and indicates a strong lead in Vulkan-based applications.

The RX 6800 XT's largest win comes in Geekbench OpenCL, where it scores 171,304 versus 154,858 for the RTX 4070, a 10.6% margin. This shows that AMD's older architecture still holds a notable edge in OpenCL compute workloads. In the DirectX 10 test, the RX 6800 XT leads with 155 points to 139, an 11.5% advantage. The DirectX 11 test also favors AMD, with the RX 6800 XT scoring 264 against 244, an 8.2% lead.

The RTX 4070 dominates the newer DirectX 12 and 3DMark tests. In 3DMark Steel Nomad DX12, the RTX 4070 scores 3854 versus 3689 for the RX 6800 XT, a 4.3% lead. The Passmark DirectX 12 test is closer, with the RTX 4070 at 103 and the RX 6800 XT at 100, a 2.9% margin. The largest NVIDIA win outside of Vulkan is in the Passmark DirectX 9 test, where the RTX 4070 scores 320 against 265, a 17.2% gap. The RTX 4070 also wins the Passmark G3D test with 26,927 points versus 24,980, a 7.2% lead, and the Passmark GPU Compute test with 14,720 versus 13,245, a 10% margin. Finally, in the Passmark G2D test, the RTX 4070 scores 1164 against 1030, an 11.5% lead.

Overall, the average benchmark scores reflect this split. The RX 6800 XT has an average benchmark score of 48,477, while the RTX 4070 sits at 37,648. However, the RX 6800 XT's percentile rank among all GPUs is 85, higher than the RTX 4070's 81. This suggests that while the RTX 4070 wins more direct head-to-head tests, the RX 6800 XT sits in a slightly higher percentile overall, likely due to its strong showing in specific compute-heavy workloads.

Where Each One Wins

The RX 6800 XT is the clear choice for OpenCL compute tasks. Its 10.6% lead in Geekbench OpenCL and its wins in DirectX 10 and DirectX 11 indicate that it handles older API workloads and general-purpose compute with relative ease. The Passmark G2D score of 1030 is lower than the RTX 4070's 1164, but the RX 6800 XT's compute-oriented strengths are evident in its 13,245 Passmark GPU Compute score, which is only 10% behind NVIDIA despite the latter's newer architecture.

The RTX 4070 is the winner for modern API workloads and Vulkan-specific applications. Its 21.4% lead in Geekbench Vulkan is the single largest margin in the comparison. It also takes the 3DMark Steel Nomad DX12 test, which is a modern DX12 benchmark, and the Passmark DirectX 12 test, though the latter is a narrow 2.9% win. For users focused on DirectX 9 legacy titles or 2D performance, the RTX 4070's 17.2% and 11.5% leads respectively make it the better option.

For gaming, the Passmark G3D score is often seen as a general indicator of 3D performance. Here the RTX 4070 leads by 7.2%, scoring 26,927 versus 24,980. The RTX 4070 also wins the Passmark GPU Compute test by 10%, which could benefit productivity workloads that leverage GPU compute. In summary, AMD wins on raw OpenCL compute and older DirectX APIs, while NVIDIA wins on Vulkan, modern DX12, and overall 3D rendering performance.

Architecture Differences

The two cards are built on fundamentally different architectures and process nodes. The AMD Radeon RX 6800 XT uses the Navi 21 chip based on RDNA 2.0 architecture, fabricated on a 7 nm process at TSMC. The NVIDIA GeForce RTX 4070 uses the AD104 chip based on Ada Lovelace architecture, fabricated on a 5 nm process, also at TSMC. The process node difference is significant, with the RTX 4070's 5 nm process allowing for a much higher transistor density of 121.8M per mm², compared to the RX 6800 XT's 51.5M per mm².

The RTX 4070 packs 35,800 million transistors on a 294 mm² die, while the RX 6800 XT has 26,800 million transistors on a larger 520 mm² die. This makes the RTX 4070 far more transistor-dense. The RX 6800 XT compensates with more memory: 16 GB of GDDR6 on a 256-bit bus with 512.0 GB/s bandwidth, versus 12 GB of GDDR6X on a 192-bit bus with 504.2 GB/s bandwidth. The memory types differ, with GDDR6X being a newer technology, but the RX 6800 XT's wider bus gives it a slight bandwidth edge.

In terms of compute units, the RX 6800 XT has 4,608 shading units, 288 TMUs, and 128 ROPs, along with 72 ray tracing cores. The RTX 4070 has 5,888 shading units, 184 TMUs, and 64 ROPs, with 46 ray tracing cores and 184 tensor cores. The RX 6800 XT has more TMUs and ROPs, which contributes to its higher pixel rate of 288.0 GPixel/s versus 158.4 GPixel/s, and higher texture rate of 648.0 GTexel/s versus 455.4 GTexel/s. However, the RTX 4070 has a higher FP32 throughput at 29.15 TFLOPS versus 20.74 TFLOPS, and its FP16 performance is 1:1 at 29.15 TFLOPS, while the RX 6800 XT's FP16 is 2:1 at 41.47 TFLOPS.

The RTX 4070 also features tensor cores, which the RX 6800 XT lacks entirely. This is a key architectural difference, as tensor cores are used for AI and deep learning workloads. The clock speeds differ as well: the RTX 4070 has a base clock of 1920 MHz and a boost clock of 2475 MHz, while the RX 6800 XT has a base clock of 1825 MHz and a boost clock of 2250 MHz, with a game clock of 2015 MHz. The RTX 4070's memory runs at 1313 MHz with 21 Gbps effective, while the RX 6800 XT's memory runs at 2000 MHz with 16 Gbps effective.

FAQ

Q: Which card has a higher average benchmark score?

A: The AMD Radeon RX 6800 XT has an average benchmark score of 48,477, which is higher than the NVIDIA GeForce RTX 4070's 37,648.

Q: Which card wins in Vulkan performance?

A: The NVIDIA GeForce RTX 4070 wins decisively in Geekbench Vulkan, scoring 174,152 versus 136,875 for the RX 6800 XT, a 21.4% lead.

Q: What is the memory configuration difference?

A: The RX 6800 XT has 16 GB of GDDR6 on a 256-bit bus with 512.0 GB/s bandwidth, while the RTX 4070 has 12 GB of GDDR6X on a 192-bit bus with 504.2 GB/s bandwidth.

Q: How does the transistor density compare?

A: The RTX 4070 has a transistor density of 121.8M per mm² on a 294 mm² die, while the RX 6800 XT has 51.5M per mm² on a 520 mm² die.

Q: Which card has more shading units?

A: The RTX 4070 has 5,888 shading units, while the RX 6800 XT has 4,608 shading units.

Q: What is the power connector requirement?

A: The RX 6800 XT requires two 8-pin power connectors, while the RTX 4070 requires a single 16-pin connector.

Specification Differences

The two cards differ in nearly every major specification. The RX 6800 XT uses the Navi 21 chip on RDNA 2.0 architecture, while the RTX 4070 uses the AD104 chip on Ada Lovelace architecture. The process node is 7 nm for AMD and 5 nm for NVIDIA. Transistor counts are 26,800 million for AMD and 35,800 million for NVIDIA. Die size is 520 mm² for AMD and 294 mm² for NVIDIA.

Clock speeds: the RX 6800 XT has a base clock of 1825 MHz, boost of 2250 MHz, and game clock of 2015 MHz, while the RTX 4070 has a base clock of 1920 MHz and boost of 2475 MHz. Memory: AMD uses 16 GB GDDR6 on a 256-bit bus with 512.0 GB/s bandwidth, while NVIDIA uses 12 GB GDDR6X on a 192-bit bus with 504.2 GB/s bandwidth.

Compute units: AMD has 4,608 shading units, 288 TMUs, 128 ROPs, and 72 ray tracing cores, while NVIDIA has 5,888 shading units, 184 TMUs, 64 ROPs, 46 ray tracing cores, and 184 tensor cores. Pixel rate is 288.0 GPixel/s for AMD versus 158.4 GPixel/s for NVIDIA. Texture rate is 648.0 GTexel/s for AMD versus 455.4 GTexel/s for NVIDIA. FP32 is 20.74 TFLOPS for AMD versus 29.15 TFLOPS for NVIDIA. FP16 is 41.47 TFLOPS (2:1) for AMD versus 29.15 TFLOPS (1:1) for NVIDIA.

Power and dimensions: AMD has a TDP of 300 W with a suggested PSU of 700 W, while NVIDIA has a TDP of 200 W with a suggested PSU of 550 W. AMD uses two 8-pin connectors, NVIDIA uses one 16-pin connector. AMD is 267 mm long, 120 mm tall, and 50 mm wide, while NVIDIA is 240 mm long, 110 mm tall, and 40 mm wide. Display outputs differ: AMD has 1x HDMI 2.1, 2x DisplayPort 1.4a, and 1x USB Type-C, while NVIDIA has 1x HDMI 2.1 and 3x DisplayPort 1.4a.

DETAILED SPECIFICATIONS

SPECIFICATION
RX 6800 XT
RTX 4070
Core Specs
Shading Units
4,608
5,888 +27.8%
Shaders
4,608
5,888 +27.8%
TMUs
288
184 -36.1%
ROPs
128
64 -50.0%
Compute Units
72
SM Count
46
Clocks
Base Clock
1825 MHz
1920 MHz
Boost Clock
2250 MHz
2475 MHz
Game Clock
2015 MHz
Memory Clock
2000 MHz 16 Gbps effective
1313 MHz 21 Gbps effective
Memory
Memory Size
16 GB
12 GB
VRAM (MB)
16,384
12,288 -25.0%
Memory Type
GDDR6
GDDR6X
Memory Bus
256 bit
192 bit
Bandwidth
512.0 GB/s
504.2 GB/s
Cache
L1 Cache
128 KB per Array
128 KB (per SM)
L2 Cache
4 MB
36 MB
L3 Cache
128 MB
L0 Cache
32 KB per WGP
Performance
Pixel Rate
288.0 GPixel/s
158.4 GPixel/s
Texture Rate
648.0 GTexel/s
455.4 GTexel/s
FP32 (TFLOPS)
20.74 TFLOPS
29.15 TFLOPS
FP64 (TFLOPS)
1,296.0 GFLOPS (1:16)
455.4 GFLOPS (1:64)
FP16 (TFLOPS)
41.47 TFLOPS (2:1)
29.15 TFLOPS (1:1)
AI/RT
RT Cores
72
46 -36.1%
Tensor Cores
184
Power
TDP
300 W
200 W
TDP (W)
300
200 -33.3%
Suggested PSU
700 W
550 W
Power Connectors
2x 8-pin
1x 16-pin
Architecture
Architecture
RDNA 2.0
Ada Lovelace
GPU Name
Navi 21
AD104
Generation
Navi II (RX 6000)
GeForce 40
Process Size
7 nm
5 nm
Transistors
26,800 million
35,800 million
Die Size
520 mm²
294 mm²
Foundry
TSMC
TSMC
Density
51.5M / mm²
121.8M / mm²
API Support
DirectX
12 Ultimate (12_2)
12 Ultimate (12_2)
OpenGL
4.6
4.6
Vulkan
1.4
1.4
OpenCL
2.1
3.0
CUDA
8.9
Shader Model
6.8
6.8
Physical
Slot Width
Dual-slot
Dual-slot
Length
267 mm 10.5 inches
240 mm 9.4 inches
Height
120 mm 4.7 inches
110 mm 4.3 inches
Outputs
1x HDMI 2.12x DisplayPort 1.4a1x USB Type-C
1x HDMI 2.13x DisplayPort 1.4a
Bus Interface
PCIe 4.0 x16
PCIe 4.0 x16
Other
Launch Price
649 USD
599 USD
Production
End-of-life
End-of-life
Predecessor
Navi
GeForce 30
Successor
Navi III
GeForce 50
View Radeon RX 6800 XT Details View GeForce RTX 4070 Details