AMD Radeon Instinct MI60 vs NVIDIA RTX PRO 5000 Blackwell Comparison

Head-to-Head Benchmarks

The recorded data shows a decisive performance gap between the NVIDIA RTX PRO 5000 Blackwell and the AMD Radeon Instinct MI60 across the two shared benchmark workloads. In the Geekbench OpenCL test, the NVIDIA card scores 254,116 points, while the AMD card scores 92,488 points. This translates to a delta of 174.8% in favor of the NVIDIA product. In the Geekbench Vulkan test, the gap widens further: the RTX PRO 5000 Blackwell reaches 282,631 points, versus 92,444 for the MI60, a 205.7% advantage. Both benchmarks are unambiguous wins for the NVIDIA part, and the margin is substantial in each case.

The average benchmark score across all recorded tests tells the same story. The RTX PRO 5000 Blackwell holds an average score of 182,109, while the MI60 averages 92,466. The NVIDIA product nearly doubles the AMD card's average output. Looking at the percentile rankings, the RTX PRO 5000 Blackwell sits in the 98th percentile of all GPUs in the database, while the MI60 lands in the 93rd percentile. That 5-point percentile gap understates the raw score difference, because the upper reaches of the database are densely packed with high-performing parts.

For context, the RTX PRO 5000 Blackwell's nearest rivals in the database include the NVIDIA A100 SXM4 80 GB, which scores 183,725 on average (0.9% above the RTX PRO 5000), and the NVIDIA RTX 5000 Ada Generation, which scores 184,664 (1.4% above). The RTX PRO 5000 also leads the GeForce RTX 4090 D, which averages 178,050 (2.3% below). These are tight margins, suggesting the RTX PRO 5000 sits comfortably in the top tier of workstation and data center GPUs.

The MI60's nearest rivals, by contrast, include the NVIDIA RTX A4500 at 91,671 (0.9% above the MI60), the RTX A4500 Mobile at 91,134 (1.5% above), the AMD Radeon Pro VII at 97,131 (4.8% above), and the AMD Radeon RX 7900M at 97,487 (5.2% above). The MI60 is outperformed by all four of these parts, but by relatively small margins. This suggests the MI60, while not competitive with the RTX PRO 5000 Blackwell, is still within striking distance of its own peer group.

Where Each One Wins

The NVIDIA RTX PRO 5000 Blackwell wins every recorded benchmark in the head-to-head comparison. It takes both the OpenCL and Vulkan tests, and it does so with margins that are not close. The Vulkan score of 282,631 is particularly strong, representing a 205.7% improvement over the MI60's 92,444. The OpenCL score of 254,116 is similarly dominant at 174.8% over the AMD part.

The AMD Radeon Instinct MI60 wins no benchmarks in this comparison. Its best showing is in OpenCL, where it scores 92,488, but that is still less than half of the RTX PRO 5000's result. The Vulkan score of 92,444 is nearly identical to its OpenCL score, indicating a balanced but limited compute profile. The MI60 does not appear to have any workload in the recorded data where it can close the gap.

From a use-case perspective, the RTX PRO 5000 Blackwell is the clear choice for any application that relies on OpenCL or Vulkan compute. The data shows it is not merely faster, but dramatically faster, with margins exceeding 170% in both tests. For users running GPU-accelerated workloads in these APIs, the NVIDIA part offers a level of performance that the MI60 cannot match.

The MI60, meanwhile, shows its age in the data. Its scores are comparable to mid-range workstation parts like the RTX A4500, but far behind the current generation of high-end accelerators. The MI60's 93rd percentile ranking means it outperforms most GPUs in the database, but the RTX PRO 5000's 98th percentile places it in a different class entirely.

The Verdict

The data is clear: the NVIDIA RTX PRO 5000 Blackwell is the superior product in this comparison. It wins both head-to-head benchmarks, has a higher average score, and ranks higher in the overall percentile distribution. For anyone choosing between these two cards for OpenCL or Vulkan workloads, the RTX PRO 5000 Blackwell is the only rational selection.

The MI60, however, is not without its place. It is an end-of-life product with a 93rd percentile ranking, and its scores align closely with those of the RTX A4500 and the AMD Radeon Pro VII. If the requirement is a capable accelerator for legacy workloads, the MI60 may still be serviceable, but it cannot compete with the RTX PRO 5000 on any metric recorded in the database.

The RTX PRO 5000 Blackwell also holds an advantage in its nearest rival comparisons. It is within 1.4% of the RTX 5000 Ada Generation and within 2.3% of the GeForce RTX 4090 D, both of which are strong performers in their own right. The MI60, by contrast, trails its nearest rivals, with the Radeon Pro VII and RX 7900M both exceeding its average score by nearly 5%.

In summary, the RTX PRO 5000 Blackwell is the decisive winner. The MI60 is a capable but outdated part that cannot match the NVIDIA offering in any measured workload.

FAQ

Q: Which GPU has the higher Geekbench OpenCL score?

A: The NVIDIA RTX PRO 5000 Blackwell scores 254,116 in OpenCL, while the AMD Radeon Instinct MI60 scores 92,488, a 174.8% difference in favor of NVIDIA.

Q: What is the average benchmark score for each card?

A: The RTX PRO 5000 Blackwell averages 182,109 across all recorded tests, while the MI60 averages 92,466.

Q: How does the RTX PRO 5000 Blackwell compare to its nearest rivals?

A: It is 0.9% below the NVIDIA A100 SXM4 80 GB, 1.4% below the RTX 5000 Ada Generation, and 2.3% above the GeForce RTX 4090 D in average score.

Q: Does the AMD Radeon Instinct MI60 win any benchmark against the RTX PRO 5000?

A: No. The MI60 loses both the OpenCL and Vulkan tests, with no recorded wins in the head-to-head data.

Q: What are the percentile rankings of the two GPUs?

A: The RTX PRO 5000 Blackwell ranks in the 98th percentile of all GPUs, while the MI60 ranks in the 93rd percentile.

Q: Which card has the larger Vulkan score advantage?

A: The RTX PRO 5000 Blackwell scores 282,631 in Vulkan, which is 205.7% higher than the MI60's 92,444.

Architecture Differences

The NVIDIA RTX PRO 5000 Blackwell is built on the GB202 chip using the Blackwell 2.0 architecture, fabricated on a 5 nm process at TSMC. It contains 92,200 million transistors on a die size of 750 mm², yielding a transistor density of 122.9 million per square millimeter. The AMD Radeon Instinct MI60 uses the Vega 20 chip with the GCN 5.1 architecture, fabricated on a 7 nm process at TSMC. It contains 13,230 million transistors on a 331 mm² die, giving a density of 40.0 million per square millimeter. The NVIDIA chip is significantly larger and more transistor-dense, reflecting a more modern design.

The RTX PRO 5000 Blackwell includes 110 ray tracing cores and 440 tensor cores, features that are entirely absent from the MI60, which has no dedicated ray tracing or tensor hardware. The NVIDIA card also has 14,080 shading units, 440 texture mapping units, and 160 raster output units. The MI60 has 4,096 shading units, 256 TMUs, and 64 ROPs. The NVIDIA part leads in every category of shader and fixed-function hardware.

The MI60 supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. The RTX PRO 5000 Blackwell supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The NVIDIA card has a newer DirectX feature level and a more recent Vulkan version.

The memory subsystems also differ fundamentally. The RTX PRO 5000 Blackwell uses 48 GB of GDDR7 memory on a 384-bit bus, achieving 1.34 TB/s of bandwidth. The MI60 uses 32 GB of HBM2 memory on a 4096-bit bus, achieving 1.02 TB/s. Despite the MI60's wider bus, the RTX PRO 5000 still delivers higher total bandwidth.

Specification Differences

The NVIDIA RTX PRO 5000 Blackwell has a base clock of 1740 MHz and a boost clock of 2377 MHz, while the MI60 operates at 1200 MHz base and 1800 MHz boost. The NVIDIA card's memory runs at 1750 MHz (28 Gbps effective), whereas the MI60's memory runs at 1000 MHz (2 Gbps effective). The RTX PRO 5000 delivers 66.94 TFLOPS of FP32 performance and 66.94 TFLOPS of FP16 (1:1 ratio). The MI60 delivers 14.75 TFLOPS of FP32 and 29.49 TFLOPS of FP16 (2:1 ratio). The NVIDIA card is over 4.5 times faster in FP32 and over 2 times faster in FP16.

The pixel rate for the RTX PRO 5000 Blackwell is 380.3 GPixel/s, and its texture rate is 1,045.9 GTexel/s. The MI60's pixel rate is 115.2 GPixel/s, and its texture rate is 460.8 GTexel/s. The NVIDIA card is roughly 3.3 times faster in pixel throughput and 2.3 times faster in texture throughput.

Both cards have a 300 W TDP and are dual-slot designs. The RTX PRO 5000 uses a single 16-pin power connector, while the MI60 uses one 6-pin and one 8-pin connector. Both have a suggested PSU of 700 W. The RTX PRO 5000 uses PCIe 5.0 x16, while the MI60 uses PCIe 4.0 x16. The NVIDIA card has 4x DisplayPort 2.1b outputs; the MI60 has a single mini-DisplayPort 1.4a.

The RTX PRO 5000 Blackwell has a launch MSRP of 5,099 USD. The MI60 has no recorded launch MSRP. The NVIDIA card is still in active production, while the MI60 is end-of-life. The RTX PRO 5000 was released in March 2025, whereas the MI60 was released in November 2018. Both cards share the same physical length (267 mm) and height (111 mm), but the RTX PRO 5000 has a recorded width of 40 mm, while the MI60's width is not listed.