Radeon Instinct MI308X vs RTX PRO 6000 Blackwell
Contents
Primary details
GPU architecture, market segment, value for money and other general parameters compared.
| Place in the ranking | 1132 | not rated |
| Place by popularity | not in top-100 | not in top-100 |
| Cost-effectiveness evaluation | 0.03 | no data |
| Power efficiency | 0.15 | no data |
| Architecture | Blackwell 2.0 (2025−2026) | CDNA 3.0 (2023−2024) |
| GPU code name | GB202 | Aqua Vanjaram |
| Market segment | Workstation | Workstation |
| Release date | 18 March 2025 (1 year ago) | 6 December 2023 (2 years ago) |
| Launch price (MSRP) | $8,565 | 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 | 24064 | 19456 |
| Core clock speed | 1590 MHz | 1000 MHz |
| Boost clock speed | 2617 MHz | 2100 MHz |
| Number of transistors | 92,200 million | 153,000 million |
| Manufacturing process technology | 5 nm | 5 nm |
| Power consumption (TDP) | 600 Watt | 750 Watt |
| Texture fill rate | 1,968 | 2,553.6 |
| Floating-point processing power | 126 TFLOPS | 81.72 TFLOPS |
| ROPs | 192 | no data |
| TMUs | 752 | 1216 |
| Tensor Cores | 752 | 1216 |
| Ray Tracing Cores | 188 | no data |
| L1 Cache | 23.5 MB | 4.8 MB |
| L2 Cache | 128 MB | 16 MB |
| L3 Cache | no data | 256 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).
| Interface | PCIe 5.0 x16 | PCIe 5.0 x16 |
| Length | 304 mm | no data |
| Width | 2-slot | OAM Module |
| Supplementary power connectors | 1x 16-pin | None |
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 | GDDR7 | HBM3 |
| Maximum RAM amount | 96 GB | 192 GB |
| Memory bus width | 512 Bit | 8192 Bit |
| Memory clock speed | 1750 MHz | 1300 MHz |
| Memory bandwidth | 1.79 TB/s | 5.32 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 | 4x DisplayPort 2.1b | No outputs |
API and SDK support
List of supported 3D and general-purpose computing APIs, including their specific versions.
| DirectX | 12 Ultimate (12_2) | N/A |
| Shader Model | 6.8 | N/A |
| OpenGL | 4.6 | N/A |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.4 | N/A |
| CUDA | 12.0 | - |
| DLSS | + | + |
Pros & cons summary
| Recency | 18 March 2025 | 6 December 2023 |
| Maximum RAM amount | 96 GB | 192 GB |
| Power consumption (TDP) | 600 Watt | 750 Watt |
RTX PRO 6000 Blackwell has an age advantage of 1 year, and 25% lower power consumption.
Instinct MI308X, on the other hand, has a 100% higher maximum VRAM amount.
We couldn't decide between RTX PRO 6000 Blackwell and Radeon Instinct MI308X. 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.
