GPU Comparison

AMD
RADEON

AMD Radeon RX 6700S

CORE STATE Navi 23
VRAM 8 GB
CLOCK SPEED 2000 MHz
TDP 80 W
BUS WIDTH 128 bit
ARCHITECTURE RDNA 2.0
nm
PROCESS 7 nm
LAUNCH DATE 2022
VS
Intel
GPU

Arc B570

CORE STATE BMG-G21
VRAM 10 GB
CLOCK SPEED 2500 MHz
TDP 150 W
BUS WIDTH 160 bit
ARCHITECTURE Xe2-HPG
nm
PROCESS 5 nm
LAUNCH DATE 2025

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12
1,451
2,649
geekbench_metal
72,864
N/A
geekbench_opencl
70,965
83,514
geekbench_vulkan
76,150
96,844
passmark_directx_10
87
65
passmark_directx_11
134
118
passmark_directx_12
58
72
passmark_directx_9
195
164
passmark_g2d
739
661
passmark_g3d
15,066
14,195
passmark_gpu_compute
5,990
7,281

Analysis: AMD Radeon RX 6700S vs Intel Arc B570

The AMD Radeon RX 6700S and Intel Arc B570 are two graphics cards separated by three years of silicon design, yet they land within 7% of each other in average benchmark score. The RX 6700S, a mobile-first RDNA 2 part from the Navi Mobile generation, posts an average score of 22154, while the Arc B570, a desktop Battlemage GPU, averages 20556. This places the AMD card at the 67th percentile of all GPUs and the Intel card at the 65th, with the RX 6700S holding a 7.8% lead in aggregate performance. The head-to-head results, however, tell a more nuanced story: each card wins exactly five of the ten comparative tests, with the wins split along generational and architectural lines.

Where Each One Wins

The benchmark split reveals a clear thematic division. The Intel Arc B570 dominates in modern, API-heavy workloads, while the AMD Radeon RX 6700S excels in legacy and rasterization-focused tests. In the 3DMark Steel Nomad DX12 test, the Arc B570 scores 2649 against the RX 6700S’s 1451, a decisive 45.2% advantage. This is the largest delta in the entire comparison and signals that Intel’s newer architecture handles contemporary DirectX 12 rendering far more efficiently. The Arc B570 also wins in Geekbench Vulkan (96844 vs 76150, a 21.4% edge) and Geekbench OpenCL (83514 vs 70965, a 15% edge), reinforcing its strength in compute-heavy and cross-API scenarios.

Conversely, the RX 6700S wins in PassMark’s legacy DirectX tests. It takes DirectX 10 (87 vs 65, a 33.8% lead), DirectX 9 (195 vs 164, an 18.9% lead), and DirectX 11 (134 vs 118, a 13.6% lead). The AMD card also wins PassMark G2D (739 vs 661, an 11.8% margin) and PassMark G3D (15066 vs 14195, a 6.1% margin). The G3D result is particularly telling because it is a general 3D performance metric that does not favor a specific API generation, suggesting the RX 6700S retains a solid overall rasterization capability despite its older design.

The compute-oriented tests split the difference: the Arc B570 wins PassMark GPU Compute (7281 vs 5990, a 17.7% edge), while the RX 6700S wins PassMark DirectX 12 (58 vs 72, which actually favors Intel by 19.4% when read as the B570’s score being higher). Wait, the data shows the RX 6700S wins DirectX 10, 11, 9, G2D, and G3D, five wins, while the Arc B570 wins Steel Nomad, OpenCL, Vulkan, DirectX 12, and GPU Compute, also five wins. This parity in win count, despite the AMD card’s higher average score, indicates that the choice depends heavily on the software environment.

Architecture Differences

The architectural gap between these two is substantial. The RX 6700S uses the Navi 23 chip built on TSMC’s 7 nm process, containing 11,060 million transistors on a 237 mm² die. The Arc B570 uses the BMG-G21 chip on a 5 nm TSMC process, packing 19,600 million transistors into a 272 mm² die. This translates to a transistor density of 46.7M per mm² for AMD versus 72.1M per mm² for Intel, showing the newer node’s efficiency advantage. The RX 6700S belongs to the RDNA 2.0 architecture from the Navi Mobile generation, while the Arc B570 is Xe2-HPG from the Battlemage (Arc 5) generation.

Core counts differ significantly. The RX 6700S has 1792 shading units, 112 texture mapping units, and 64 ROPs, along with 28 ray tracing cores. The Arc B570 has 2304 shading units, 144 TMUs, and 80 ROPs, but only 18 ray tracing cores. Despite fewer RT cores, the Intel card’s higher overall core count and clock speeds drive its performance. The RX 6700S runs at a base clock of 1700 MHz with a boost of 2000 MHz, while the Arc B570 operates at a flat 2500 MHz base and boost. This 25% higher clock speed, combined with 28.6% more shading units, explains much of the Intel card’s lead in modern benchmarks.

Memory configurations also diverge. The RX 6700S uses 8 GB of GDDR6 on a 128-bit bus, yielding 224.0 GB/s bandwidth. The Arc B570 has 10 GB of GDDR6 on a 160-bit bus, providing 380.0 GB/s. The 69.6% bandwidth advantage for Intel is critical for high-resolution textures and compute workloads. The RX 6700S’s FP32 throughput is 7.168 TFLOPS, while the Arc B570 reaches 11.52 TFLOPS, a 60.7% difference. FP16 performance follows the same pattern: 14.34 TFLOPS for AMD versus 23.04 TFLOPS for Intel. Power consumption also differs sharply, with the RX 6700S rated at 80 W TDP and the Arc B570 at 150 W TDP, which explains the former’s IGP slot width and lack of power connectors versus the latter’s dual-slot design and single 8-pin connector.

The Verdict

The data points to distinct use cases. For gamers running modern DirectX 12 titles or leveraging Vulkan APIs, the Intel Arc B570 is the clear choice. Its 45.2% lead in 3DMark Steel Nomad and 21.4% lead in Geekbench Vulkan demonstrate a substantial advantage in current-generation rendering paths. The 10 GB memory capacity and 380.0 GB/s bandwidth also provide more headroom for texture-heavy games, and the 150 W TDP is appropriate for a desktop card with a 450 W suggested PSU.

For users dealing with legacy DirectX applications, the AMD Radeon RX 6700S is superior. Its wins in DirectX 9, 10, and 11 tests, with margins from 13.6% to 33.8%, show that RDNA 2 retains strong compatibility and optimization for older APIs. The PassMark G3D victory (15066 vs 14195) also suggests that the RX 6700S holds its own in general 3D workloads that do not specifically target the newest API features. The 80 W TDP and IGP form factor make it suitable for thin-and-light systems, though the Arc B570’s desktop orientation is evident in its physical dimensions of 272 mm length and 115 mm height.

The average benchmark scores favor the RX 6700S (22154 vs 20556), but this aggregate hides the fact that the Arc B570 wins in the most future-proof workloads. The Intel card’s nearest rival is the RTX 3070 Mobile (deltaPct of 0.1%), while the RX 6700S sits near the RX 5700 (deltaPct of -0.1%). This places both cards in similar performance tiers, but with different strengths. The Arc B570 is the better pick for new game releases and compute tasks; the RX 6700S is the safer bet for older software libraries and lower-power environments.

FAQ

Q: Which card has the higher average benchmark score?

A: The AMD Radeon RX 6700S has an average benchmark score of 22154, which is 7.8% higher than the Intel Arc B570’s average of 20556.

Q: How much faster is the Intel Arc B570 in 3DMark Steel Nomad DX12?

A: The Arc B570 scores 2649 in the 3DMark Steel Nomad DX12 test, while the RX 6700S scores 1451. This gives the Intel card a 45.2% lead, the largest single-test margin in the comparison.

Q: Which card wins in legacy DirectX 9 and DirectX 10 benchmarks?

A: The AMD Radeon RX 6700S wins both. It scores 195 in PassMark DirectX 9 (an 18.9% lead over the Arc B570’s 164) and 87 in PassMark DirectX 10 (a 33.8% lead over 65).

Q: What is the memory bandwidth difference between the two GPUs?

A: The Intel Arc B570 has 380.0 GB/s bandwidth from its 10 GB GDDR6 memory on a 160-bit bus, while the AMD Radeon RX 6700S has 224.0 GB/s from 8 GB GDDR6 on a 128-bit bus. The Intel card’s bandwidth is 69.6% higher.

Q: Which card has more shading units and a higher boost clock?

A: The Intel Arc B570 has 2304 shading units and a 2500 MHz boost clock. The AMD Radeon RX 6700S has 1792 shading units and a 2000 MHz boost clock.

Q: Does the AMD card win any DirectX 12 benchmark?

A: No, the Intel Arc B570 wins the PassMark DirectX 12 test with a score of 72 against the RX 6700S’s 58, a 19.4% margin for Intel.

Head-to-Head Benchmarks

The most striking result is in 3DMark Steel Nomad DX12, where the Arc B570’s 2649 score dwarfs the RX 6700S’s 1451. This 45.2% delta is nearly double the next-largest margin and indicates that Intel’s Xe2 architecture is far more adept at handling the geometry and shading workloads of modern DX12 titles. The RX 6700S, from the RDNA 2 era, simply cannot match this level of efficiency in the newest API.

The Geekbench results reinforce Intel’s lead in compute and modern APIs. In Vulkan, the Arc B570 scores 96844 versus 76150, a 21.4% advantage. In OpenCL, it scores 83514 versus 70965, a 15% edge. These tests stress general-purpose GPU compute and cross-vendor API performance, where the Arc B570’s higher core count (2304 vs 1792) and clock speed (2500 MHz vs 2000 MHz) provide a clear numerical advantage.

The RX 6700S counters in the PassMark suite. Its DirectX 10 score of 87 beats the Arc B570’s 65 by 33.8%, and its DirectX 9 score of 195 beats 164 by 18.9%. The DirectX 11 test shows a 13.6% lead (134 vs 118), and the G2D test shows an 11.8% lead (739 vs 661). The PassMark G3D result is the closest of the five AMD wins: 15066 vs 14195, a 6.1% margin that demonstrates the RX 6700S’s continued viability in general 3D rendering.

The compute split goes to Intel, with the Arc B570 taking PassMark GPU Compute at 7281 versus 5990, a 17.7% lead. The DirectX 12 PassMark result also favors Intel (72 vs 58, a 19.4% margin), which aligns with the Steel Nomad outcome. Both cards win five tests each, but the magnitude of Intel’s wins in modern workloads (45.2%, 21.4%, 19.4%, 17.7%, 15%) versus AMD’s wins in legacy tests (33.8%, 18.9%, 13.6%, 11.8%, 6.1%) suggests that the Arc B570 is better positioned for future software development.

Specification Differences

The process node is a major differentiator: the RX 6700S uses TSMC’s 7 nm process, while the Arc B570 uses TSMC’s 5 nm process. This allows Intel to fit 19,600 million transistors on a 272 mm² die (72.1M per mm²), versus AMD’s 11,060 million on a 237 mm² die (46.7M per mm²). Clock speeds also differ, with the RX 6700S running at 1700 MHz base and 2000 MHz boost, while the Arc B570 runs at a constant 2500 MHz for both base and boost.

Memory capacity and bandwidth favor Intel: 10 GB GDDR6 on a 160-bit bus with 380.0 GB/s versus 8 GB GDDR6 on a 128-bit bus with 224.0 GB/s. The core configuration shows Intel with 2304 shading units, 144 TMUs, and 80 ROPs, compared to AMD’s 1792 shading units, 112 TMUs, and 64 ROPs. Ray tracing cores are the exception, with AMD having 28 RT cores versus Intel’s 18, though the performance data does not show an AMD advantage in RT-specific tests.

The RX 6700S has a TDP of 80 W and is an IGP slot width with no power connectors, while the Arc B570 has a 150 W TDP, dual-slot design, and one 8-pin power connector. The Intel card also lists a 450 W suggested PSU, while the AMD card has none. Display outputs differ: the RX 6700S is portable-device dependent, whereas the Arc B570 offers 1x HDMI 2.1a and 3x DisplayPort 2.1. The RX 6700S is end-of-life with a release date of 2022-01-03, while the Arc B570 is active with a 2025-01-15 release and a launch MSRP of 219 USD. Both support DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, and both use a PCIe 4.0 x8 bus interface.

DETAILED SPECIFICATIONS

SPECIFICATION
RX 6700S
B570
Core Specs
Shading Units
1,792
2,304 +28.6%
Shaders
1,792
2,304 +28.6%
TMUs
112
144 +28.6%
ROPs
64
80 +25.0%
Compute Units
28
Execution Units
18
Clocks
Base Clock
1700 MHz
2500 MHz
Boost Clock
2000 MHz
2500 MHz
Game Clock
1890 MHz
Memory Clock
1750 MHz 14 Gbps effective
2375 MHz 19 Gbps effective
Memory
Memory Size
8 GB
10 GB
VRAM (MB)
8,192
10,240 +25.0%
Memory Type
GDDR6
GDDR6
Memory Bus
128 bit
160 bit
Bandwidth
224.0 GB/s
380.0 GB/s
Cache
L1 Cache
128 KB per Array
256 KB (per EU)
L2 Cache
2 MB
13.5 MB
L3 Cache
32 MB
L0 Cache
32 KB per WGP
Performance
Pixel Rate
128.0 GPixel/s
200.0 GPixel/s
Texture Rate
224.0 GTexel/s
360.0 GTexel/s
FP32 (TFLOPS)
7.168 TFLOPS
11.52 TFLOPS
FP64 (TFLOPS)
448.0 GFLOPS (1:16)
720.0 GFLOPS (1:16)
FP16 (TFLOPS)
14.34 TFLOPS (2:1)
23.04 TFLOPS (2:1)
AI/RT
RT Cores
28
18 -35.7%
XMX Cores
144
Power
TDP
80 W
150 W
TDP (W)
80
150 +87.5%
Suggested PSU
450 W
Power Connectors
None
1x 8-pin
Architecture
Architecture
RDNA 2.0
Xe2-HPG
GPU Name
Navi 23
BMG-G21
Generation
Navi Mobile (RX 6000M)
Battlemage (Arc 5)
Process Size
7 nm
5 nm
Transistors
11,060 million
19,600 million
Die Size
237 mm²
272 mm²
Foundry
TSMC
TSMC
Density
46.7M / mm²
72.1M / 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
Shader Model
6.8
6.6
Physical
Slot Width
IGP
Dual-slot
Length
272 mm 10.7 inches
Height
115 mm 4.5 inches
Outputs
Portable Device Dependent
1x HDMI 2.1a3x DisplayPort 2.1
Bus Interface
PCIe 4.0 x8
PCIe 4.0 x8
Other
Launch Price
219 USD
Production
End-of-life
Active
Predecessor
Polaris Mobile
Alchemist
View Radeon RX 6700S Details View Arc B570 Details