GPU Comparison
GEFORCE
NVIDIA GeForce RTX 4090 D
CORE STATE
AD102
VRAM
24 GB
CLOCK SPEED
2520 MHz
TDP
425 W
BUS WIDTH
384 bit
ARCHITECTURE
Ada Lovelace
PROCESS
5 nm
LAUNCH DATE
2023
VS
GEFORCE
L20
CORE STATE
AD102
VRAM
48 GB
CLOCK SPEED
2520 MHz
TDP
275 W
BUS WIDTH
384 bit
ARCHITECTURE
Ada Lovelace
PROCESS
5 nm
LAUNCH DATE
2023
PERFORMANCE BENCHMARKS
3dmark_3dmark_steel_nomad_dx12
geekbench_opencl
geekbench_vulkan
DETAILED SPECIFICATIONS
SPECIFICATION
RTX 4090 D
L20
Core Specs
Shading Units
14,592
11,776
-19.3%
Shaders
14,592
11,776
-19.3%
TMUs
456
368
-19.3%
ROPs
176
128
-27.3%
SM Count
114
92
-19.3%
Clocks
Base Clock
2280 MHz
1440 MHz
Boost Clock
2520 MHz
2520 MHz
Memory Clock
1313 MHz
21 Gbps effective
2250 MHz
18 Gbps effective
Memory
Memory Size
24 GB
48 GB
VRAM (MB)
24,576
49,152
+100.0%
Memory Type
GDDR6X
GDDR6
Memory Bus
384 bit
384 bit
Bandwidth
1.01 TB/s
864.0 GB/s
Cache
L1 Cache
128 KB (per SM)
128 KB (per SM)
L2 Cache
72 MB
96 MB
Performance
Pixel Rate
443.5 GPixel/s
322.6 GPixel/s
Texture Rate
1,149.1 GTexel/s
927.4 GTexel/s
FP32 (TFLOPS)
73.54 TFLOPS
59.35 TFLOPS
FP64 (TFLOPS)
1,149.1 GFLOPS (1:64)
927.4 GFLOPS (1:64)
FP16 (TFLOPS)
73.54 TFLOPS (1:1)
59.35 TFLOPS (1:1)
AI/RT
RT Cores
114
92
-19.3%
Tensor Cores
456
368
-19.3%
Power
TDP
425 W
275 W
TDP (W)
425
275
-35.3%
Suggested PSU
800 W
600 W
Power Connectors
1x 16-pin
1x 16-pin
Architecture
Architecture
Ada Lovelace
Ada Lovelace
GPU Name
AD102
AD102
Generation
GeForce 40
Server Ada
(Lxx)
Process Size
5 nm
5 nm
Transistors
76,300 million
76,300 million
Die Size
609 mm²
609 mm²
Foundry
TSMC
TSMC
Density
125.3M / mm²
125.3M / mm²
API Support
DirectX
12 Ultimate (12_2)
12 Ultimate (12_2)
OpenGL
4.6
4.6
Vulkan
1.4
1.4
OpenCL
3.0
3.0
CUDA
8.9
8.9
Shader Model
6.8
6.8
Physical
Slot Width
Triple-slot
Dual-slot
Length
304 mm
12 inches
267 mm
10.5 inches
Height
137 mm
5.4 inches
111 mm
4.4 inches
Outputs
1x HDMI 2.13x DisplayPort 1.4a
4x DisplayPort 1.4a
Bus Interface
PCIe 4.0 x16
PCIe 4.0 x16
Other
Launch Price
1,599 USD
—
Production
End-of-life
Active
Predecessor
GeForce 30
Server Ampere
Successor
GeForce 50
Server Hopper