Radeon Instinct MI350P vs Tesla PG503-216
Contents
Primary details
GPU architecture, market segment, value for money and other general parameters compared.
| Place in the ranking | not rated | not rated |
| Place by popularity | not in top-100 | not in top-100 |
| Architecture | Volta (2017−2020) | CDNA 4.0 (2025−2026) |
| GPU code name | GV100 | MI350 128CU |
| Market segment | Workstation | Workstation |
| Release date | 26 November 2019 (6 years ago) | 7 May 2026 (less than a year ago) |
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 | 4608 | 8192 |
| Core clock speed | 1312 MHz | 1000 MHz |
| Boost clock speed | 1530 MHz | 2200 MHz |
| Number of transistors | 21,100 million | 73,000 million |
| Manufacturing process technology | 12 nm | 3 nm |
| Power consumption (TDP) | 250 Watt | 600 Watt |
| Texture fill rate | 440.6 | 1,126.4 |
| Floating-point processing power | 14.1 TFLOPS | 36.04 TFLOPS |
| ROPs | 128 | no data |
| TMUs | 288 | 512 |
| Tensor Cores | 640 | no data |
| L1 Cache | 10 MB | 2 MB |
| L2 Cache | 6 MB | 16 MB |
| L3 Cache | no data | 128 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 3.0 x16 | PCIe 5.0 x16 |
| Length | no data | 267 mm |
| Width | 2-slot | 2-slot |
| Supplementary power connectors | None | 1x 16-pin |
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 | HBM2 | HBM3e |
| Maximum RAM amount | 32 GB | 144 GB |
| Memory bus width | 4096 Bit | 8192 Bit |
| Memory clock speed | 1106 MHz | 2000 MHz |
| Memory bandwidth | 1.13 TB/s | 8.19 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 | No outputs | No outputs |
API and SDK support
List of supported 3D and general-purpose computing APIs, including their specific versions.
| DirectX | 12 (12_1) | N/A |
| Shader Model | 6.8 | N/A |
| OpenGL | 4.6 | N/A |
| OpenCL | 3.0 | 3.0 |
| Vulkan | 1.3 | N/A |
| CUDA | 7.0 | - |
| DLSS | + | - |
Pros & cons summary
| Recency | 26 November 2019 | 7 May 2026 |
| Maximum RAM amount | 32 GB | 144 GB |
| Chip lithography | 12 nm | 3 nm |
| Power consumption (TDP) | 250 Watt | 600 Watt |
Tesla PG503-216 has 140% lower power consumption.
Instinct MI350P, on the other hand, has an age advantage of 6 years, a 350% higher maximum VRAM amount, and a 300% more advanced lithography process.
We couldn't decide between Tesla PG503-216 and Radeon Instinct MI350P. 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.
