AMD Radeon RX 6800 XT vs AMD Radeon RX 6900 XT 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
AMD
RADEON

Radeon RX 6900 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

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12
3,689
4,079
geekbench_metal
181,337
177,021
geekbench_opencl
171,304
187,673
geekbench_vulkan
136,875
148,525
passmark_directx_10
155
167
passmark_directx_11
264
279
passmark_directx_12
100
113
passmark_directx_9
265
268
passmark_g2d
1,030
1,055
passmark_g3d
24,980
26,732
passmark_gpu_compute
13,245
14,547

Analysis: AMD Radeon RX 6800 XT vs AMD Radeon RX 6900 XT

The AMD Radeon RX 6900 XT and RX 6800 XT are two very close relatives from the same RDNA 2.0 family, but the benchmark data reveals a clear performance hierarchy. The RX 6900 XT wins 10 out of 11 head-to-head tests, with an average benchmark score of 50,951 compared to the RX 6800 XT’s 48,477. Both cards occupy the same performance tier—the 6900 XT sits at the 86th percentile of all GPUs, while the 6800 XT is at the 85th—so the differences are about squeezing extra frames at the top end, not a generational leap.

Where Each One Wins

The RX 6900 XT is the default pick for raw compute and DirectX workloads. It leads in every DirectX test in the comparison, with its largest margin in 3DMark Steel Nomad DX12, where it scores 4,079 against 3,689—a 10.6% advantage. The gap persists across older API tests: 13% ahead in PassMark DirectX 12 (113 vs 100), 7.7% in DirectX 10 (167 vs 155), and 5.7% in DirectX 11 (279 vs 264). For compute-heavy tasks, the 6900 XT posts 18,7673 in Geekbench OpenCL against 171,304, a 9.6% win, and it also takes Geekbench Vulkan with 148,525 versus 136,875, an 8.5% edge. PassMark GPU Compute shows a similar story: 14,547 vs 13,245, a 9.8% delta.

The RX 6800 XT has exactly one win, and it is in Geekbench Metal. It scores 181,337 against the 6900 XT’s 177,021, a 2.4% advantage. This is a meaningful data point for macOS or Metal-based workloads, but it is an isolated result rather than a trend. In every other benchmark—including synthetic 2D tests where the 6900 XT leads PassMark G2D by 2.4% (1,055 vs 1,030)—the 6800 XT trails. The practical takeaway: if you are targeting DirectX gaming, Vulkan, OpenCL, or general GPU compute, the 6900 XT is the stronger choice. The 6800 XT only makes sense if Metal performance is your priority, and even then the lead is modest.

Architecture Differences

Both cards are built on the same Navi 21 chip, use TSMC’s 7 nm process, and pack 26,800 million transistors on a 520 mm² die with a transistor density of 51.5M per mm². They share identical clock speeds: 1825 MHz base, 2250 MHz boost, and 2015 MHz game clock. Memory configurations are also identical—16 GB of GDDR6 on a 256-bit bus with 512.0 GB/s bandwidth and 16 Gbps effective speed. The difference lies in the number of enabled compute units. The RX 6900 XT has 5,120 shading units, 320 texture mapping units (TMUs), and 80 ray tracing cores. The RX 6800 XT scales down to 4,608 shading units, 288 TMUs, and 72 ray tracing cores. Both retain 128 ROPs, so pixel fill rate is identical at 288.0 GPixel/s.

The shading unit and TMU deficit directly explains the performance gap. The 6900 XT’s texture rate is 720.0 GTexel/s versus 648.0 GTexel/s for the 6800 XT—an 11.1% difference that tracks closely with the observed benchmark deltas. Floating-point throughput follows the same pattern: 23.04 TFLOPS FP32 and 46.08 TFLOPS FP16 (2:1) for the 6900 XT, against 20.74 TFLOPS FP32 and 41.47 TFLOPS FP16 for the 6800 XT. The 6800 XT is essentially a cut-down version of the same silicon, with fewer active execution units but identical clocks, memory, and ROPs.

Both cards share the same feature set: PCIe 4.0 x16 interface, 1x HDMI 2.1, 2x DisplayPort 1.4a, and 1x USB Type-C outputs, plus DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4 support. They also share the same 300 W TDP, 2x 8-pin power connectors, and a suggested 700 W PSU. The only physical difference in specs is slot width: the 6900 XT is triple-slot, while the 6800 XT is dual-slot. Both are 267 mm long, 120 mm tall, and 50 mm wide. This means the 6900 XT demands more case clearance, but otherwise the cards are drop-in equivalents from a power and connectivity standpoint.

FAQ

Q: Which card has better overall benchmark performance?

A: The RX 6900 XT. It wins 10 of 11 head-to-head tests and has an average benchmark score of 50,951 versus 48,477 for the RX 6800 XT, a 5.1% advantage.

Q: Is the RX 6800 XT ever faster?

A: Yes, in Geekbench Metal. The RX 6800 XT scores 181,337 against the 6900 XT’s 177,021, a 2.4% lead. This is its only benchmark win in this comparison.

Q: Do the cards have the same memory configuration?

A: Yes. Both have 16 GB of GDDR6 on a 256-bit bus with 512.0 GB/s bandwidth, running at 16 Gbps effective speed.

Q: What is the core count difference?

A: The RX 6900 XT has 5,120 shading units, 320 TMUs, and 80 ray tracing cores. The RX 6800 XT has 4,608 shading units, 288 TMUs, and 72 ray tracing cores. Both have 128 ROPs.

Q: Are the power requirements different?

A: No. Both cards have a 300 W TDP, require 2x 8-pin power connectors, and list a 700 W suggested PSU.

Q: Which card is better for DirectX 12 gaming?

A: The RX 6900 XT. It leads by 13% in PassMark DirectX 12 (113 vs 100) and by 10.6% in 3DMark Steel Nomad DX12 (4,079 vs 3,689).

Specification Differences

The two cards differ in only a handful of fields, all stemming from the 6900 XT having more enabled silicon on the same Navi 21 die.

  • Shading Units: 5,120 (6900 XT) vs 4,608 (6800 XT)
  • TMUs: 320 (6900 XT) vs 288 (6800 XT)
  • Ray Tracing Cores: 80 (6900 XT) vs 72 (6800 XT)
  • Texture Rate: 720.0 GTexel/s (6900 XT) vs 648.0 GTexel/s (6800 XT)
  • FP32 Performance: 23.04 TFLOPS (6900 XT) vs 20.74 TFLOPS (6800 XT)
  • FP16 Performance: 46.08 TFLOPS (6900 XT) vs 41.47 TFLOPS (6800 XT)
  • Slot Width: Triple-slot (6900 XT) vs Dual-slot (6800 XT)
  • Launch MSRP: 999 USD (6900 XT) vs 649 USD (6800 XT)

All other specifications are identical: 7 nm process, 26,800 million transistors, 520 mm² die, 1825 MHz base clock, 2250 MHz boost clock, 2015 MHz game clock, 16 GB GDDR6, 256-bit bus, 512.0 GB/s bandwidth, 128 ROPs, 288.0 GPixel/s pixel rate, 300 W TDP, 2x 8-pin connectors, 700 W suggested PSU, PCIe 4.0 x16, and the same display outputs and API support.

Head-to-Head Benchmarks

The largest single win for the RX 6900 XT comes in PassMark DirectX 12, where it scores 113 versus 100—a 13% delta. This is followed closely by 3DMark Steel Nomad DX12, where the 6900 XT’s 4,079 beats the 6800 XT’s 3,689 by 10.6%. These two tests represent the modern gaming workload, and the 6900 XT’s consistent double-digit advantage suggests it scales better with the extra shader and TMU resources when the API is fully utilized.

Compute-oriented benchmarks reinforce the same pattern. Geekbench OpenCL shows a 9.6% lead for the 6900 XT (187,673 vs 171,304), and PassMark GPU Compute shows a 9.8% lead (14,547 vs 13,245). Geekbench Vulkan is slightly narrower at 8.5% (148,525 vs 136,875). The older DirectX APIs are closer: DirectX 10 gives the 6900 XT a 7.7% edge (167 vs 155), DirectX 11 a 5.7% edge (279 vs 264), and DirectX 9 a marginal 1.1% win (268 vs 265). The 2D test, PassMark G2D, is also close at 2.4% (1,055 vs 1,030), which makes sense since 2D performance depends more on memory and ROPs, which are identical between the cards.

The RX 6800 XT’s only victory is Geekbench Metal, where it scores 181,337 versus 177,021—a 2.4% advantage. Notably, the 6900 XT’s OpenCL score (187,673) is higher than its own Metal score, while the 6800 XT’s Metal score exceeds its OpenCL score (171,304). This suggests the 6800 XT has a specific optimization or driver path for Metal that the 6900 XT lacks, but it is an isolated result with no impact on the overall benchmark average.

The Verdict

The data is unambiguous: the RX 6900 XT is the faster card in nearly every measurable way. It wins 10 of 11 benchmarks, posts a higher average score (50,951 vs 48,477), and holds a percentile ranking of 86 versus 85 for the 6800 XT. The performance delta is consistent—roughly 5-10% in most tests, with peaks of 13% in DirectX 12 and 10.6% in 3DMark Steel Nomad. If you are building a system for DirectX gaming, Vulkan titles, OpenCL compute, or any general GPU workload, the 6900 XT is the clear choice, provided your case can accommodate its triple-slot cooler.

The RX 6800 XT is not a bad card; it is simply a slightly less enabled version of the same chip. Its single win in Geekbench Metal (181,337 vs 177,021) makes it relevant only for users who prioritize Metal performance, and even then the margin is just 2.4%. For everything else, the 6800 XT trails by margins that grow larger as the workload gets more demanding. The 6800 XT does offer the advantage of a dual-slot design versus the 6900 XT’s triple-slot, which matters for compact builds, and it shares the identical 300 W TDP, 700 W PSU recommendation, and 2x 8-pin power requirements. But on pure benchmark performance, the 6900 XT justifies its higher position in the lineup. Buy the 6900 XT for maximum performance; buy the 6800 XT only if you specifically need Metal acceleration or a slimmer card.

DETAILED SPECIFICATIONS

SPECIFICATION
RX 6800 XT
RX 6900 XT
Core Specs
Shading Units
4,608
5,120 +11.1%
Shaders
4,608
5,120 +11.1%
TMUs
288
320 +11.1%
ROPs
128
128 0.0%
Compute Units
72
80 +11.1%
Clocks
Base Clock
1825 MHz
1825 MHz
Boost Clock
2250 MHz
2250 MHz
Game Clock
2015 MHz
2015 MHz
Memory Clock
2000 MHz 16 Gbps effective
2000 MHz 16 Gbps effective
Memory
Memory Size
16 GB
16 GB
VRAM (MB)
16,384
16,384 0.0%
Memory Type
GDDR6
GDDR6
Memory Bus
256 bit
256 bit
Bandwidth
512.0 GB/s
512.0 GB/s
Cache
L1 Cache
128 KB per Array
128 KB per Array
L2 Cache
4 MB
4 MB
L3 Cache
128 MB
128 MB
L0 Cache
32 KB per WGP
32 KB per WGP
Performance
Pixel Rate
288.0 GPixel/s
288.0 GPixel/s
Texture Rate
648.0 GTexel/s
720.0 GTexel/s
FP32 (TFLOPS)
20.74 TFLOPS
23.04 TFLOPS
FP64 (TFLOPS)
1,296.0 GFLOPS (1:16)
1,440.0 GFLOPS (1:16)
FP16 (TFLOPS)
41.47 TFLOPS (2:1)
46.08 TFLOPS (2:1)
AI/RT
RT Cores
72
80 +11.1%
Power
TDP
300 W
300 W
TDP (W)
300
300 0.0%
Suggested PSU
700 W
700 W
Power Connectors
2x 8-pin
2x 8-pin
Architecture
Architecture
RDNA 2.0
RDNA 2.0
GPU Name
Navi 21
Navi 21
Generation
Navi II (RX 6000)
Navi II (RX 6000)
Process Size
7 nm
7 nm
Transistors
26,800 million
26,800 million
Die Size
520 mm²
520 mm²
Foundry
TSMC
TSMC
Density
51.5M / mm²
51.5M / 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
2.1
Shader Model
6.8
6.8
Physical
Slot Width
Dual-slot
Triple-slot
Length
267 mm 10.5 inches
267 mm 10.5 inches
Height
120 mm 4.7 inches
120 mm 4.7 inches
Outputs
1x HDMI 2.12x DisplayPort 1.4a1x USB Type-C
1x HDMI 2.12x DisplayPort 1.4a1x USB Type-C
Bus Interface
PCIe 4.0 x16
PCIe 4.0 x16
Other
Launch Price
649 USD
999 USD
Production
End-of-life
End-of-life
Predecessor
Navi
Navi
Successor
Navi III
Navi III
View Radeon RX 6800 XT Details View Radeon RX 6900 XT Details