AMD Radeon RX 7700S vs NVIDIA P104-100 Comparison

AMD
RADEON

AMD Radeon RX 7700S

CORE STATE Navi 33
VRAM 8 GB
CLOCK SPEED 2500 MHz
TDP 100 W
BUS WIDTH 128 bit
ARCHITECTURE RDNA 3.0
nm
PROCESS 6 nm
LAUNCH DATE 2023
VS
NVIDIA
GEFORCE

P104-100

CORE STATE GP104
VRAM 4 GB
CLOCK SPEED 1733 MHz
TDP
BUS WIDTH 256 bit
ARCHITECTURE Pascal
nm
PROCESS 16 nm
LAUNCH DATE 2017

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12
2,185
1,413
geekbench_opencl
62,983
52,368
geekbench_vulkan
36,380
45,165

Analysis: AMD Radeon RX 7700S vs NVIDIA P104-100

AMD Radeon RX 7700S and NVIDIA P104-100 occupy adjacent percentile tiers — 78th versus 77th — yet their benchmark profiles diverge sharply. The RX 7700S wins two of three head-to-head tests by wide margins, while the P104-100 counters with a single decisive victory in Vulkan. These are not interchangeable parts; they are solutions to different problems, with the data revealing a generational gap masked by similar average scores.

Head-to-Head Benchmarks

The 3DMark Steel Nomad DX12 test delivers the most lopsided result. AMD's RX 7700S scores 2185 against NVIDIA's 1413, a delta of 54.6%. This is not a marginal edge; it is a 54.6% lead in a modern DirectX 12 workload, reflecting the RX 7700S's architectural advantage in contemporary rendering paths. The gap is large enough to classify the P104-100 as effectively non-competitive in this specific metric.

Geekbench OpenCL shows a similar, though less dramatic, AMD victory. The RX 7700S posts 62983 versus 52368, a 20.3% advantage. This test exercises general-purpose compute, and the RX 7700S's lead here is substantial. The 20.3% delta is roughly one-third the size of the Steel Nomad gap, suggesting the AMD part's advantage narrows in compute-heavy but less game-specific workloads.

The NVIDIA P104-100's lone win arrives in Geekbench Vulkan, where it scores 45165 against the RX 7700S's 36380. That is a 19.5% margin in NVIDIA's favor. This result is striking because Vulkan is a modern, low-level API, yet the older Pascal architecture outperforms RDNA 3.0 here. The data implies the P104-100 has a specific optimization or driver path that favors this workload, or that the RX 7700S's Vulkan implementation is comparatively unoptimized.

Across all three tests, the RX 7700S wins two and loses one, with an aggregate picture that favors AMD. However, the average benchmark scores tell a nuanced story: the RX 7700S averages 33849, while the P104-100 averages 32982. The difference is only 867 points, or roughly 2.6%. This near-parity in averages, despite wide swings in individual tests, indicates that the two GPUs are balanced in overall capability but achieve it through very different performance profiles. The RX 7700S is a peak-performance specialist; the P104-100 is a consistent, if lower-ceiling, performer.

The Verdict

The data points to the AMD Radeon RX 7700S as the superior choice for anyone prioritizing modern DirectX 12 gaming or OpenCL compute. A 54.6% lead in Steel Nomad is a definitive statement about future-proofing in DX12 titles. Its 20.3% OpenCL advantage further cements its position for general-purpose GPU compute tasks. The 78th percentile ranking versus 77th is a narrow overall gap, but the head-to-head deltas show the RX 7700S wins where it matters most for contemporary software.

The NVIDIA P104-100 is the pick only if Vulkan performance is the sole criterion. Its 19.5% Vulkan lead is significant, but it is an isolated victory. The P104-100's 77th percentile ranking, combined with its losses in the other two tests, suggests it is a one-trick pony. For a user whose workload is exclusively Vulkan-based, the P104-100 might be acceptable, but the RX 7700S's broader strengths make it the more rational default.

Considering the production status, the RX 7700S is Active, while the P104-100 is End-of-life. This alone tilts the verdict toward AMD for any new purchase. The P104-100's 2017 release date and mining-GPU lineage further reinforce that it is a legacy part. Benchmark data indicates a clear winner: the RX 7700S for overall performance, with the P104-100 reserved for niche Vulkan scenarios.

Architecture Differences

The fundamental split is process node and architecture generation. The RX 7700S uses TSMC's 6 nm process, while the P104-100 uses 16 nm. This is a two-generation leap in manufacturing, directly explaining the RX 7700S's efficiency and transistor density. The RX 7700S packs 13,300 million transistors into 204 mm², yielding a density of 65.2M per mm². The P104-100 has 7,200 million transistors on a larger 314 mm² die, at just 22.9M per mm². The RX 7700S achieves more than 2.8 times the transistor density.

Architecturally, the RX 7700S is RDNA 3.0, codenamed Hotpink Bonefish, belonging to the Navi Mobile (RX 7000M) generation. The P104-100 is Pascal, on the GP104 chip, categorized under Mining GPUs. This is not just a clock-speed difference; it is a fundamental design philosophy shift. RDNA 3.0 introduces dedicated ray-tracing cores — the RX 7700S has 32 RT cores — while Pascal has none. The RX 7700S also features 2048 shading units, 128 TMUs, and 64 ROPs; the P104-100 has 1920 shading units, 120 TMUs, and 64 ROPs. The AMD part has more of everything except ROPs, which are equal.

Clock speeds further separate them. The RX 7700S has a 1500 MHz base and 2500 MHz boost, with a 2200 MHz game clock. The P104-100 runs at 1607 MHz base and 1733 MHz boost. The RX 7700S's boost clock is 44% higher, a consequence of the smaller 6 nm node. Memory configurations differ too: the RX 7700S uses 8 GB GDDR6 on a 128-bit bus for 288.0 GB/s bandwidth, while the P104-100 has 4 GB GDDR5X on a 256-bit bus for 320.3 GB/s. The P104-100 has more bandwidth but half the capacity.

API support reveals the generational split. The RX 7700S supports DirectX 12 Ultimate (12_2), while the P104-100 is limited to DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4, but the Vulkan benchmark results suggest implementation quality differs wildly. The P104-100's FP16 rate is 104.0 GFLOPS at a 1:64 ratio, while the RX 7700S achieves 40.96 TFLOPS at 2:1. The FP32 rates are 20.48 TFLOPS versus 6.655 TFLOPS, a 3.1x difference favoring AMD.

FAQ

Q: Which GPU has a higher transistor density?

A: The AMD Radeon RX 7700S, at 65.2M transistors per mm², versus the NVIDIA P104-100's 22.9M per mm².

Q: Does the P104-100 support hardware ray tracing?

A: No. The data lists null for RT cores on the P104-100, while the RX 7700S has 32 RT cores.

Q: Which card has more memory bandwidth?

A: The NVIDIA P104-100, with 320.3 GB/s, edges out the RX 7700S's 288.0 GB/s.

Q: What is the average benchmark score difference?

A: The RX 7700S averages 33849, and the P104-100 averages 32982, a difference of 867 points.

Q: Which GPU is in production now?

A: The AMD Radeon RX 7700S is Active; the NVIDIA P104-100 is End-of-life.

Q: Why does the P104-100 win the Vulkan test?

A: The data shows a 19.5% victory for the P104-100 in Geekbench Vulkan, but the pack does not specify the cause; it only records the score difference.

Where Each One Wins

The RX 7700S is the clear winner in DirectX 12 workloads, as evidenced by its 54.6% lead in 3DMark Steel Nomad. For modern game development targeting DX12 Ultimate features, this is the only rational choice. Its 20.3% OpenCL advantage also makes it superior for compute tasks like machine learning inference or physics simulations that leverage OpenCL. The 8 GB memory capacity doubles the P104-100's 4 GB, allowing larger textures and datasets.

The P104-100 wins exclusively in Vulkan-based applications. The 45,165 Geekbench Vulkan score, 19.5% ahead of the RX 7700S, suggests that Vulkan renderers — common in certain emulators or Linux gaming — would perform better on this older card. Its 256-bit memory bus provides 320.3 GB/s bandwidth, which is 11% higher than the RX 7700S, potentially benefiting bandwidth-sensitive workloads. The dual-slot design and 1x 8-pin power connector also indicate it is a standalone add-in card, unlike the RX 7700S's IGP form factor.

For pixel and texture throughput, the RX 7700S leads: 160.0 GPixel/s versus 110.9 GPixel/s, and 320.0 GTexel/s versus 208.0 GTexel/s. These are 44% and 54% advantages, respectively, reinforcing the RX 7700S's dominance in fill-rate-bound scenarios. The P104-100's only spec-sheet victories are memory bandwidth and a smaller die area — the latter being a moot point given its lower transistor count.

Specification Differences

The two cards differ in nearly every measurable category. Process node: 6 nm versus 16 nm. Transistors: 13,300 million versus 7,200 million. Die size: 204 mm² versus 314 mm². Transistor density: 65.2M/mm² versus 22.9M/mm². Base clock: 1500 MHz versus 1607 MHz. Boost clock: 2500 MHz versus 1733 MHz. The RX 7700S has a game clock of 2200 MHz; the P104-100 has none listed.

Memory is a major differentiator. The RX 7700S has 8 GB GDDR6 at 18 Gbps effective on a 128-bit bus, yielding 288.0 GB/s. The P104-100 has 4 GB GDDR5X at 10 Gbps effective on a 256-bit bus, yielding 320.3 GB/s. Shading units: 2048 versus 1920. TMUs: 128 versus 120. ROPs are tied at 64. The RX 7700S has 32 RT cores; the P104-100 has none.

Compute rates diverge sharply: FP32 is 20.48 TFLOPS versus 6.655 TFLOPS; FP16 is 40.96 TFLOPS versus 104.0 GFLOPS. Pixel rate is 160.0 GPixel/s versus 110.9 GPixel/s; texture rate is 320.0 GTexel/s versus 208.0 GTexel/s. The RX 7700S has a 100 W TDP; the P104-100's TDP is not listed, but it requires a 200 W suggested PSU. Form factor: the RX 7700S is IGP with no power connectors, while the P104-100 is dual-slot with 1x 8-pin and a 267 mm length. Bus interface: PCIe 4.0 x16 versus PCIe 1.0 x4. The RX 7700S supports DirectX 12 Ultimate; the P104-100 only DirectX 12 (12_1).

DETAILED SPECIFICATIONS

SPECIFICATION
RX 7700S
P104-100
Core Specs
Shading Units
2,048
1,920 -6.3%
Shaders
2,048
1,920 -6.3%
TMUs
128
120 -6.3%
ROPs
64
64 0.0%
Compute Units
32
SM Count
15
Clocks
Base Clock
1500 MHz
1607 MHz
Boost Clock
2500 MHz
1733 MHz
Game Clock
2200 MHz
Memory Clock
2250 MHz 18 Gbps effective
1251 MHz 10 Gbps effective
Memory
Memory Size
8 GB
4 GB
VRAM (MB)
8,192
4,096 -50.0%
Memory Type
GDDR6
GDDR5X
Memory Bus
128 bit
256 bit
Bandwidth
288.0 GB/s
320.3 GB/s
Cache
L1 Cache
128 KB per Array
48 KB (per SM)
L2 Cache
2 MB
2 MB
L3 Cache
32 MB
L0 Cache
32 KB per WGP
Performance
Pixel Rate
160.0 GPixel/s
110.9 GPixel/s
Texture Rate
320.0 GTexel/s
208.0 GTexel/s
FP32 (TFLOPS)
20.48 TFLOPS
6.655 TFLOPS
FP64 (TFLOPS)
640.0 GFLOPS (1:32)
208.0 GFLOPS (1:32)
FP16 (TFLOPS)
40.96 TFLOPS (2:1)
104.0 GFLOPS (1:64)
AI/RT
RT Cores
32
Power
TDP
100 W
TDP (W)
100
Suggested PSU
200 W
Power Connectors
None
1x 8-pin
Architecture
Architecture
RDNA 3.0
Pascal
GPU Name
Navi 33
GP104
Codename
Hotpink Bonefish
Generation
Navi Mobile (RX 7000M)
Mining GPUs
Process Size
6 nm
16 nm
Transistors
13,300 million
7,200 million
Die Size
204 mm²
314 mm²
Foundry
TSMC
TSMC
Density
65.2M / mm²
22.9M / mm²
API Support
DirectX
12 Ultimate (12_2)
12 (12_1)
OpenGL
4.6
4.6
Vulkan
1.4
1.4
OpenCL
2.2
3.0
CUDA
6.1
Shader Model
6.8
6.8
Physical
Slot Width
IGP
Dual-slot
Length
267 mm 10.5 inches
Outputs
Portable Device Dependent
No outputs
Bus Interface
PCIe 4.0 x16
PCIe 1.0 x4
Other
Production
Active
End-of-life
Predecessor
Polaris Mobile
View Radeon RX 7700S Details View P104-100 Details