H200 NVL vs Radeon Pro Duo
Contents
Primary details
GPU architecture, market segment, value for money and other general parameters compared.
| Place in the ranking | 314 | not rated |
| Place by popularity | not in top-100 | not in top-100 |
| Cost-effectiveness evaluation | 2.07 | no data |
| Power efficiency | 4.49 | no data |
| Architecture | GCN 3.0 (2014−2019) | Hopper (2022−2024) |
| GPU code name | Capsaicin | GH100 |
| Market segment | Workstation | Workstation |
| Design | reference | no data |
| Release date | 26 April 2016 (10 years ago) | 18 November 2024 (1 year ago) |
| Launch price (MSRP) | $1,499 | no data |
Cost-effectiveness evaluation
The higher the ratio, the better. We use the manufacturer's recommended prices.
Performance to price scatter graph
Currently popular graphics cards are shown for comparison.
Detailed specifications
General parameters such as number of shaders, GPU core base clock and boost clock speeds, manufacturing process, texturing and calculation speed. Real power consumption of some graphics cards can well exceed their nominal TDP, especially if overclocked.
| Pipelines / CUDA cores | 4096 ×2 | 16896 |
| Compute units | 128 | no data |
| Core clock speed | no data | 1365 MHz |
| Boost clock speed | 1000 MHz | 1785 MHz |
| Number of transistors | 8,900 million | 80,000 million |
| Manufacturing process technology | 28 nm | 5 nm |
| Power consumption (TDP) | 350 Watt | 600 Watt |
| Texture fill rate | 256.0 ×2 | 942.5 |
| Floating-point processing power | 8.192 TFLOPS ×2 | 60.32 TFLOPS |
| ROPs | 64 ×2 | 24 |
| TMUs | 256 ×2 | 528 |
| Tensor Cores | no data | 528 |
| L1 Cache | 1 MB | 33 MB |
| L2 Cache | 2 MB | 50 MB |
Form factor & compatibility
Information on compatibility with other computer components. Useful when choosing a future computer configuration or upgrading an existing one. For desktop graphics cards it's interface and bus (motherboard compatibility), additional power connectors (power supply compatibility).
| Bus support | PCIe 3.0 | no data |
| Interface | PCIe 3.0 x16 | PCIe 5.0 x16 |
| Length | 277 mm | 267 mm |
| Width | 2-slot | 2-slot |
| Supplementary power connectors | 3x 8-pin | 8-pin EPS |
VRAM capacity and type
Parameters of VRAM installed: its type, size, bus, clock and resulting bandwidth. Integrated GPUs have no dedicated video RAM and use a shared part of system RAM.
| Memory type | High Bandwidth Memory (HBM) | HBM3e |
| Maximum RAM amount | 8 GB ×2 | 141 GB |
| Memory bus width | 4096 Bit ×2 | 6144 Bit |
| Memory clock speed | 500 MHz | 1593 MHz |
| Memory bandwidth | 512 GB/s ×2 | 4.89 TB/s |
| Resizable BAR | - | + |
Connectivity and outputs
This section shows the types and number of video connectors on each GPU. The data applies specifically to desktop reference models (for example, NVIDIA’s Founders Edition). OEM partners often modify both the number and types of ports. On notebook GPUs, video‐output options are determined by the laptop’s design rather than the graphics chip itself.
| Display Connectors | 1x HDMI, 3x DisplayPort | No outputs |
| Eyefinity | + | - |
| Number of Eyefinity displays | 6 | no data |
| HDMI | + | - |
| DisplayPort support | + | - |
Supported technologies
Supported technological solutions. This information will prove useful if you need some particular technology for your purposes.
| AppAcceleration | + | - |
| CrossFire | + | - |
| Enduro | + | - |
| FRTC | + | - |
| FreeSync | + | - |
| HD3D | + | - |
| LiquidVR | + | - |
| PowerTune | + | - |
| TressFX | + | - |
| TrueAudio | + | - |
| ZeroCore | + | - |
| UVD | + | - |
| VCE | + | - |
API and SDK support
List of supported 3D and general-purpose computing APIs, including their specific versions.
| DirectX | DirectX® 12 | N/A |
| Shader Model | 6.0 | N/A |
| OpenGL | 4.5 | N/A |
| OpenCL | 2.0 | 3.0 |
| Vulkan | + | N/A |
| Mantle | + | - |
| CUDA | - | 9.0 |
| DLSS | - | + |
Pros & cons summary
| Recency | 26 April 2016 | 18 November 2024 |
| Maximum RAM amount | 8 GB | 141 GB |
| Chip lithography | 28 nm | 5 nm |
| Power consumption (TDP) | 350 Watt | 600 Watt |
Pro Duo has 71% lower power consumption.
H200 NVL, on the other hand, has an age advantage of 8 years, a 1663% higher maximum VRAM amount, and a 460% more advanced lithography process.
We couldn't decide between Radeon Pro Duo and H200 NVL. We've got no test results to judge.
Other comparisons
We selected several comparisons of graphics cards with performance close to those reviewed, providing you with more options to consider.
