AMD Radeon R9 370X vs NVIDIA GeForce RTX 5090 D V2 Comparison

AMD
RADEON

AMD Radeon R9 370X

CORE STATE Trinidad
VRAM 2 GB
CLOCK SPEED 1030 MHz
TDP 180 W
BUS WIDTH 256 bit
ARCHITECTURE GCN 1.0
nm
PROCESS 28 nm
LAUNCH DATE 2015
VS
NVIDIA
GEFORCE

GeForce RTX 5090 D V2

CORE STATE GB202
VRAM 24 GB
CLOCK SPEED 2407 MHz
TDP 575 W
BUS WIDTH 384 bit
ARCHITECTURE Blackwell 2.0
nm
PROCESS 5 nm
LAUNCH DATE 2025

PERFORMANCE BENCHMARKS

geekbench_metal
12,676
N/A
geekbench_opencl
25,893
N/A
geekbench_vulkan
9,018
N/A
3dmark_3dmark_steel_nomad_dx12
N/A
16,504

Analysis: AMD Radeon R9 370X vs NVIDIA GeForce RTX 5090 D V2

NVIDIA GeForce RTX 5090 D V2 and AMD Radeon R9 370X represent two vastly different eras of GPU design. The RTX 5090 D V2 is a modern Blackwell 2.0 architecture flagship built on a 5 nm process, while the R9 370X is a legacy GCN 1.0 part from the Pirate Islands generation on 28 nm. The data shows a generational chasm in raw compute, memory subsystem, and feature support, though the older card retains a niche in legacy API environments. Benchmark averages place the RTX 5090 D V2 at 16504 points (59th percentile of all GPUs) versus the R9 370X at 15862 points (58th percentile), a nominal 4% difference in aggregate scores that belies the architectural gulf.

FAQ

Q: How do the two cards compare in average benchmark scores?

A: The NVIDIA GeForce RTX 5090 D V2 averages 16504 points across its benchmark suite, while the AMD Radeon R9 370X averages 15862 points. This puts the RTX 5090 D V2 at the 59th percentile of all GPUs and the R9 370X at the 58th percentile.

Q: What are the closest rivals to each card according to benchmark data?

A: The RTX 5090 D V2's nearest rival is the NVIDIA T400 (16508 points, 0% delta), followed closely by the AMD Radeon PRO W7500 (16415 points, 0.5% delta) and NVIDIA RTX PRO 6000 Blackwell (16408 points, 0.6% delta). The R9 370X sits near the AMD Radeon RX 9060 (16014 points, -1% delta) and NVIDIA GeForce RTX 3060 Ti (16129 points, -1.7% delta).

Q: Which card has higher memory bandwidth?

A: The RTX 5090 D V2 offers 1.34 TB/s of bandwidth from 24 GB of GDDR7 on a 384-bit bus. The R9 370X provides 179.2 GB/s from 2 GB of GDDR5 on a 256-bit bus. This represents a roughly 7.5x advantage in raw bandwidth for the newer card.

Q: What is the transistor count difference between the two GPUs?

A: The RTX 5090 D V2 packs 92,200 million transistors on a 750 mm² die, yielding a density of 122.9M transistors per mm². The R9 370X contains 2,800 million transistors on a 212 mm² die, with a density of 13.2M per mm².

Q: Do both cards support the same DirectX version?

A: No. The RTX 5090 D V2 supports DirectX 12 Ultimate (12_2), while the R9 370X is limited to DirectX 12 (11_1). The newer card also supports Vulkan 1.4 versus Vulkan 1.2.170 on the older part.

Q: Which card has a higher pixel fill rate?

A: The RTX 5090 D V2 achieves 423.6 GPixel/s, whereas the R9 370X manages 32.96 GPixel/s. The RTX 5090 D V2 also delivers 1,636.8 GTexel/s of texture rate versus 82.40 GTexel/s on the R9 370X.

Architecture Differences

The NVIDIA GeForce RTX 5090 D V2 is built on the Blackwell 2.0 architecture using the GB202 chip, fabricated by TSMC on a 5 nm process. It contains 92,200 million transistors across a 750 mm² die, achieving a transistor density of 122.9M per mm². In contrast, the AMD Radeon R9 370X uses the Trinidad chip based on GCN 1.0 architecture, manufactured on TSMC's 28 nm process with just 2,800 million transistors on a 212 mm² die (13.2M per mm² density).

The compute configuration diverges dramatically. The RTX 5090 D V2 features 21,760 shading units, 680 texture mapping units, 176 ROPs, 170 RT cores, and 680 tensor cores. The R9 370X has 1,280 shading units, 80 TMUs, and 32 ROPs, with no dedicated RT or tensor cores at all. This means the newer card supports hardware-accelerated ray tracing and AI workloads, while the older card relies entirely on traditional rasterization.

Clock speeds also differ fundamentally. The RTX 5090 D V2 runs at a 2017 MHz base and 2407 MHz boost clock, while the R9 370X operates at 980 MHz base and 1030 MHz boost. The memory subsystem shows similar divergence: the RTX 5090 D V2 uses 24 GB of GDDR7 at 1750 MHz (28 Gbps effective) on a 384-bit bus, whereas the R9 370X uses 2 GB of GDDR5 at 1400 MHz (5.6 Gbps effective) on a 256-bit bus.

Feature support reflects the generational gap. The RTX 5090 D V2 supports DirectX 12 Ultimate (12_2), Vulkan 1.4, and OpenGL 4.6, with display outputs including 1x HDMI 2.1b and 3x DisplayPort 2.1b. The R9 370X supports DirectX 12 (11_1), Vulkan 1.2.170, and OpenGL 4.6, with older outputs like 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2. The newer card also uses PCIe 5.0 x16 versus PCIe 3.0 x16 on the older part.

Head-to-Head Benchmarks

The benchmark data shows a closer aggregate picture than the architectural differences suggest. The RTX 5090 D V2 scores 16504 in the 3DMark Steel Nomad DX12 test, while the R9 370X achieves 12676 in Geekbench Metal, 25893 in Geekbench OpenCL, and 9018 in Geekbench Vulkan. Direct comparison is complicated because the tests differ, but the average scores indicate the RTX 5090 D V2 leads by roughly 4% overall.

Looking at nearest rivals provides context. The RTX 5090 D V2's 16504 score places it 0.5% ahead of the AMD Radeon PRO W7500 (16415) and 0.6% ahead of the NVIDIA RTX PRO 6000 Blackwell (16408). It trails the NVIDIA T400 by 0% (16508). The R9 370X's 15862 average puts it 0.1% ahead of the AMD Radeon RX 7700 (15852), 1.2% ahead of the AMD Radeon Pro W5500 (15679), but 1% behind the AMD Radeon RX 9060 (16014) and 1.7% behind the NVIDIA GeForce RTX 3060 Ti (16129).

The strongest advantage for the RTX 5090 D V2 appears in raw compute throughput. Its FP32 performance of 104.8 TFLOPS dwarfs the R9 370X's 2.637 TFLOPS, a roughly 40x difference. Pixel rate of 423.6 GPixel/s versus 32.96 GPixel/s and texture rate of 1,636.8 GTexel/s versus 82.40 GTexel/s show similar magnitudes of separation. Memory bandwidth of 1.34 TB/s versus 179.2 GB/s gives the newer card over 7x the data throughput.

The R9 370X shows no benchmark wins in the data, but its Geekbench OpenCL score of 25893 demonstrates that legacy GCN architecture still handles compute workloads respectably. Its Vulkan score of 9018 and Metal score of 12676 indicate functional, if dated, driver support across modern APIs.

Specification Differences

The two cards differ across nearly every specification category. Process node: 5 nm for the RTX 5090 D V2 versus 28 nm for the R9 370X. Transistors: 92,200 million versus 2,800 million. Die size: 750 mm² versus 212 mm². Transistor density: 122.9M per mm² versus 13.2M per mm².

Clock speeds: the RTX 5090 D V2 runs at 2017 MHz base and 2407 MHz boost, compared to 980 MHz base and 1030 MHz boost on the R9 370X. Memory clock is 1750 MHz (28 Gbps effective) versus 1400 MHz (5.6 Gbps effective). Memory configuration: 24 GB GDDR7 on a 384-bit bus versus 2 GB GDDR5 on a 256-bit bus. Bandwidth: 1.34 TB/s versus 179.2 GB/s.

Compute units: 21,760 shading units, 680 TMUs, 176 ROPs, 170 RT cores, and 680 tensor cores on the RTX 5090 D V2. The R9 370X has 1,280 shading units, 80 TMUs, and 32 ROPs, with no RT or tensor cores. Pixel rate: 423.6 GPixel/s versus 32.96 GPixel/s. Texture rate: 1,636.8 GTexel/s versus 82.40 GTexel/s. FP32: 104.8 TFLOPS versus 2.637 TFLOPS.

Power and connectivity: the RTX 5090 D V2 has a 575 W TDP with a single 16-pin connector and requires a 950 W PSU. The R9 370X has a 180 W TDP with dual 6-pin connectors and requires a 450 W PSU. Bus interface: PCIe 5.0 x16 versus PCIe 3.0 x16. Display outputs: 1x HDMI 2.1b and 3x DisplayPort 2.1b versus 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2. Dimensions: 304 mm length, 137 mm height, 48 mm width versus 221 mm length, 111 mm height, and no width specified.

Where Each One Wins

The RTX 5090 D V2 wins decisively in every modern workload category. Its 104.8 TFLOPS FP32 compute enables heavy simulation, AI inference, and content creation tasks that the R9 370X's 2.637 TFLOPS cannot approach. The 24 GB GDDR7 memory with 1.34 TB/s bandwidth supports large datasets and high-resolution textures, while the 2 GB GDDR5 on the older card limits it to low-resolution workloads. Hardware RT cores and tensor cores give the RTX 5090 D V2 dedicated acceleration for ray tracing and machine learning, features entirely absent from the R9 370X.

The RTX 5090 D V2's 423.6 GPixel/s pixel rate and 1,636.8 GTexel/s texture rate make it suitable for high-refresh-rate 4K gaming and professional visualization. Its DirectX 12 Ultimate support enables the latest rendering features, and PCIe 5.0 x16 bandwidth ensures no bottleneck with modern CPUs. The card's 59th percentile ranking among all GPUs, with rival scores within 0.6%, places it in a competitive tier for its generation.

The R9 370X retains value only in specific legacy contexts. Its GCN 1.0 architecture with 1,280 shading units still handles older DirectX 11 titles and OpenGL workloads at modest settings. The 180 W TDP and dual 6-pin connectors make it easier to install in older systems with smaller power supplies (450 W suggested versus 950 W). Its 2 GB GDDR5 memory is sufficient for 1080p gaming from its era, and the 2x DVI outputs support legacy monitors. The card's 58th percentile ranking, with rivals like the RX 7700 within 0.1%, shows it remains functional in aggregate benchmarks despite its age.

For modern gaming, ray tracing, AI workloads, or professional content creation, the RTX 5090 D V2 is the clear choice based on every measurable specification. The R9 370X is only relevant for retro builds, budget systems with strict power limits, or testing legacy API compatibility. The data shows no scenario where the older card outperforms the newer one in raw performance, but its lower power requirements and PCIe 3.0 compatibility may suit specific older platforms.

DETAILED SPECIFICATIONS

SPECIFICATION
R9 370X
RTX 5090 D V2
Core Specs
Shading Units
1,280
21,760 +1600.0%
Shaders
1,280
21,760 +1600.0%
TMUs
80
680 +750.0%
ROPs
32
176 +450.0%
Compute Units
24
—
SM Count
—
170
Clocks
Base Clock
980 MHz
2017 MHz
Boost Clock
1030 MHz
2407 MHz
Memory Clock
1400 MHz 5.6 Gbps effective
1750 MHz 28 Gbps effective
Memory
Memory Size
2 GB
24 GB
VRAM (MB)
2,048
24,576 +1100.0%
Memory Type
GDDR5
GDDR7
Memory Bus
256 bit
384 bit
Bandwidth
179.2 GB/s
1.34 TB/s
Cache
L1 Cache
16 KB (per CU)
128 KB (per SM)
L2 Cache
512 KB
96 MB
Performance
Pixel Rate
32.96 GPixel/s
423.6 GPixel/s
Texture Rate
82.40 GTexel/s
1,636.8 GTexel/s
FP32 (TFLOPS)
2.637 TFLOPS
104.8 TFLOPS
FP64 (TFLOPS)
164.8 GFLOPS (1:16)
1.637 TFLOPS (1:64)
FP16 (TFLOPS)
—
104.8 TFLOPS (1:1)
AI/RT
RT Cores
—
170
Tensor Cores
—
680
Power
TDP
180 W
575 W
TDP (W)
180
575 +219.4%
Suggested PSU
450 W
950 W
Power Connectors
2x 6-pin
1x 16-pin
Architecture
Architecture
GCN 1.0
Blackwell 2.0
GPU Name
Trinidad
GB202
Generation
Pirate Islands (R9 300)
GeForce 50
Process Size
28 nm
5 nm
Transistors
2,800 million
92,200 million
Die Size
212 mm²
750 mm²
Foundry
TSMC
TSMC
Density
13.2M / mm²
122.9M / mm²
API Support
DirectX
12 (11_1)
12 Ultimate (12_2)
OpenGL
4.6
4.6
Vulkan
1.2.170
1.4
OpenCL
2.1 (1.2)
3.0
CUDA
—
12.0
Shader Model
6.5 (5.1)
6.9
Physical
Slot Width
Dual-slot
Dual-slot
Length
221 mm 8.7 inches
304 mm 12 inches
Height
111 mm 4.4 inches
137 mm 5.4 inches
Outputs
2x DVI1x HDMI 1.4a1x DisplayPort 1.2
1x HDMI 2.1b3x DisplayPort 2.1b
Bus Interface
PCIe 3.0 x16
PCIe 5.0 x16
Other
Launch Price
199 USD
2,299 USD
Production
End-of-life
Active
Predecessor
Volcanic Islands
GeForce 40
Successor
Arctic Islands
GeForce 60
View Radeon R9 370X Details View GeForce RTX 5090 D V2 Details