A40 PCIe vs Radeon HD 8670D
Contents
Primary details
GPU architecture, market segment, value for money and other general parameters compared.
| Place in the ranking | 1057 | not rated |
| Place by popularity | not in top-100 | not in top-100 |
| Power efficiency | 0.97 | no data |
| Architecture | TeraScale 3 (2010−2013) | Ampere (2020−2025) |
| GPU code name | Devastator | GA102 |
| Market segment | Desktop | Workstation |
| Release date | 12 March 2013 (12 years ago) | 5 October 2020 (4 years ago) |
Detailed specifications
General parameters such as number of shaders, GPU core base clock and boost clock speeds, manufacturing process, texturing and calculation speed. Note that power consumption of some graphics cards can well exceed their nominal TDP, especially when overclocked.
| Pipelines / CUDA cores | 384 | 10752 |
| Core clock speed | 844 MHz | 1305 MHz |
| Boost clock speed | 950 MHz | 1755 MHz |
| Number of transistors | 1,303 million | 28,300 million |
| Manufacturing process technology | 32 nm | 8 nm |
| Power consumption (TDP) | 100 Watt | 300 Watt |
| Texture fill rate | 22.80 | 589.7 |
| Floating-point processing power | 0.7296 TFLOPS | 37.74 TFLOPS |
| ROPs | 8 | 112 |
| TMUs | 24 | 336 |
| Tensor Cores | no data | 336 |
| Ray Tracing Cores | no data | 84 |
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).
| Interface | IGP | PCIe 4.0 x16 |
| Length | no data | 267 mm |
| Width | IGP | 2-slot |
| Supplementary power connectors | no data | 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 | System Shared | GDDR6 |
| Maximum RAM amount | System Shared | 48 GB |
| Memory bus width | System Shared | 384 Bit |
| Memory clock speed | System Shared | 1812 MHz |
| Memory bandwidth | no data | 695.8 GB/s |
| Shared memory | + | - |
| 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 | No outputs | 3x DisplayPort |
API and SDK compatibility
List of supported 3D and general-purpose computing APIs, including their specific versions.
| DirectX | 11.2 (11_0) | 12 Ultimate (12_2) |
| Shader Model | 5.0 | 6.5 |
| OpenGL | 4.4 | 4.6 |
| OpenCL | 1.2 | 2.0 |
| Vulkan | N/A | 1.2 |
| CUDA | - | 8.6 |
| DLSS | - | + |
Pros & cons summary
| Recency | 12 March 2013 | 5 October 2020 |
| Chip lithography | 32 nm | 8 nm |
| Power consumption (TDP) | 100 Watt | 300 Watt |
HD 8670D has 200% lower power consumption.
A40 PCIe, on the other hand, has an age advantage of 7 years, and a 300% more advanced lithography process.
We couldn't decide between Radeon HD 8670D and A40 PCIe. We've got no test results to judge.
Be aware that Radeon HD 8670D is a desktop graphics card while A40 PCIe is a workstation one.
Other comparisons
We selected several comparisons of graphics cards with performance close to those reviewed, providing you with more options to consider.
