GPU Comparison
RADEON
AMD Radeon Instinct MI50
CORE STATE
Vega 20
VRAM
16 GB
CLOCK SPEED
1746 MHz
TDP
300 W
BUS WIDTH
4096 bit
ARCHITECTURE
GCN 5.1
PROCESS
7 nm
LAUNCH DATE
2018
VS
GEFORCE
GeForce RTX 2080 Ti
CORE STATE
TU102
VRAM
11 GB
CLOCK SPEED
1545 MHz
TDP
250 W
BUS WIDTH
352 bit
ARCHITECTURE
Turing
PROCESS
12 nm
LAUNCH DATE
2018
PERFORMANCE BENCHMARKS
3dmark_3dmark_steel_nomad_dx12
geekbench_opencl
geekbench_vulkan
passmark_directx_10
passmark_directx_11
passmark_directx_12
passmark_directx_9
passmark_g2d
passmark_g3d
passmark_gpu_compute
DETAILED SPECIFICATIONS
SPECIFICATION
Instinct MI50
RTX 2080 Ti
Core Specs
Shading Units
3,840
4,352
+13.3%
Shaders
3,840
4,352
+13.3%
TMUs
240
272
+13.3%
ROPs
64
88
+37.5%
Compute Units
60
—
SM Count
—
68
Clocks
Base Clock
1200 MHz
1350 MHz
Boost Clock
1746 MHz
1545 MHz
Memory Clock
1000 MHz
2 Gbps effective
1750 MHz
14 Gbps effective
Memory
Memory Size
16 GB
11 GB
VRAM (MB)
16,384
11,264
-31.3%
Memory Type
HBM2
GDDR6
Memory Bus
4096 bit
352 bit
Bandwidth
1.02 TB/s
616.0 GB/s
Cache
L1 Cache
16 KB (per CU)
64 KB (per SM)
L2 Cache
4 MB
5.5 MB
Performance
Pixel Rate
111.7 GPixel/s
136.0 GPixel/s
Texture Rate
419.0 GTexel/s
420.2 GTexel/s
FP32 (TFLOPS)
13.41 TFLOPS
13.45 TFLOPS
FP64 (TFLOPS)
6.705 TFLOPS (1:2)
420.2 GFLOPS (1:32)
FP16 (TFLOPS)
26.82 TFLOPS (2:1)
26.90 TFLOPS (2:1)
AI/RT
RT Cores
—
68
Tensor Cores
—
544
Power
TDP
300 W
250 W
TDP (W)
300
250
-16.7%
Suggested PSU
700 W
600 W
Power Connectors
2x 8-pin
2x 8-pin
Architecture
Architecture
GCN 5.1
Turing
GPU Name
Vega 20
TU102
Generation
Radeon Instinct
(MIx)
GeForce 20
Process Size
7 nm
12 nm
Transistors
13,230 million
18,600 million
Die Size
331 mm²
754 mm²
Foundry
TSMC
TSMC
Density
40.0M / mm²
24.7M / mm²
API Support
DirectX
12 (12_1)
12 Ultimate (12_2)
OpenGL
4.6
4.6
Vulkan
1.3
1.4
OpenCL
2.1
3.0
CUDA
—
7.5
Shader Model
6.7
6.8
Physical
Slot Width
Dual-slot
Dual-slot
Length
267 mm
10.5 inches
267 mm
10.5 inches
Height
—
116 mm
4.6 inches
Outputs
1x mini-DisplayPort 1.4a
1x HDMI 2.03x DisplayPort 1.4a1x USB Type-C
Bus Interface
PCIe 4.0 x16
PCIe 3.0 x16
Other
Launch Price
—
999 USD
Production
End-of-life
End-of-life
Predecessor
FirePro Data Center
GeForce 10
Successor
—
GeForce 30