NVIDIA Quadro M6000 vs NVIDIA RTX A2000 Comparison

The NVIDIA RTX A2000 and NVIDIA Quadro M6000 represent two distinct eras of workstation graphics, separated by six years of architectural evolution. The data shows a clear generational shift: the RTX A2000 wins both head-to-head benchmark comparisons, yet the Quadro M6000 retains advantages in raw memory capacity and pixel throughput that matter for specific professional workloads. This analysis breaks down where each card excels, what the underlying silicon differences mean, and who should choose which based strictly on the benchmark data.

Where Each One Wins

The RTX A2000 dominates the compute-oriented benchmarks. In Geekbench OpenCL, it scores 67,695 against the Quadro M6000’s 39,688 — a 70.6% advantage. In Geekbench Vulkan, the RTX A2000 posts 69,089 versus 46,913, a 47.3% lead. These results indicate the A2000 is the clear winner for general-purpose GPU compute, API-accelerated workloads, and modern rendering pipelines that leverage Vulkan or OpenCL.

The Quadro M6000 wins in areas not captured by the head-to-head benchmark list. Its 12 GB of GDDR5 memory is double the A2000’s 6 GB, and its 384-bit bus provides 317.4 GB/s bandwidth versus 288.0 GB/s. The M6000 also has higher pixel rate (106.9 GPixel/s vs 57.60 GPixel/s) and texture rate (213.9 GTexel/s vs 124.8 GTexel/s). For workloads that are fill-rate bound — such as high-resolution 2D compositing or multi-sample anti-aliasing at large frame buffers — the M6000’s data suggests it holds an edge, even if it loses in compute-focused synthetic tests.

The RTX A2000’s wins are consistent across both available benchmarks, while the M6000 has zero benchmark victories. However, the M6000’s 84th percentile versus the A2000’s 85th percentile shows they sit in the same performance tier overall, with an average benchmark score of 43,301 for the M6000 and 46,043 for the A2000. The A2000 is roughly 6.3% faster on average, but neither card is dramatically ahead in aggregate.

Architecture Differences

The architectural gap is enormous. The RTX A2000 uses the GA106 chip on an 8 nm Samsung process, while the Quadro M6000 uses the GM200 chip on a 28 nm TSMC process. This process shrink allows the A2000 to pack 12,000 million transistors into a 276 mm² die, achieving a transistor density of 43.5 million per mm². The M6000 has 8,000 million transistors spread across a massive 601 mm² die, with a density of just 13.3 million per mm² — less than one-third the density of the A2000.

The A2000 features 3,328 shading units, 104 texture mapping units, and 48 ROPs, along with 26 ray tracing cores and 104 tensor cores. The M6000 has 3,072 shading units, 192 TMUs, and 96 ROPs, but no ray tracing or tensor cores — those simply did not exist in the Maxwell 2.0 architecture. This means the A2000 supports DirectX 12 Ultimate (12_2), while the M6000 is limited to DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4.

Clock speeds tell a similar story of generational change. The M6000 runs at a 988 MHz base and 1114 MHz boost, while the A2000 has a much lower 562 MHz base but a 1200 MHz boost. The A2000’s FP32 throughput is 7.987 TFLOPS versus the M6000’s 6.844 TFLOPS. Notably, the A2000 achieves FP16 at 7.987 TFLOPS (1:1 ratio with FP32), while the M6000 has no FP16 support listed at all. The A2000 also uses GDDR6 memory at 12 Gbps effective, while the M6000 uses older GDDR5 at 6.6 Gbps effective.

The Verdict

The data points to a simple conclusion: choose the RTX A2000 for modern compute-heavy workflows, and choose the Quadro M6000 only if you specifically need its memory capacity or fill-rate advantages. The A2000 wins both head-to-head benchmarks by significant margins — 70.6% in OpenCL and 47.3% in Vulkan. It is also far more efficient, with a 70 W TDP versus the M6000’s 250 W, and it requires no power connectors compared to the M6000’s single 8-pin. The A2000 also has a smaller physical footprint at 167 mm length versus 267 mm.

However, the M6000’s 12 GB memory is a genuine advantage for large datasets that exceed 6 GB. The A2000’s 6 GB GDDR6 may hit capacity limits in scenarios like massive texture atlases or large simulation grids. The M6000’s higher bandwidth (317.4 GB/s) and superior pixel/texture rates also suggest it remains competitive for certain rasterization-heavy tasks. The M6000’s 96 ROPs are double the A2000’s 48 ROPs, which directly explains its higher pixel rate.

For most users, the RTX A2000 is the better choice: it is faster in compute, supports ray tracing and tensor cores, uses modern APIs, and consumes far less power. The Quadro M6000’s only compelling argument is memory size and fill-rate performance, which serves a narrow set of legacy or specialized workloads. If you need more than 6 GB of VRAM and do not rely on ray tracing or FP16 compute, the M6000 remains a viable option — but the data shows it is otherwise outclassed.

FAQ

Q: Which card is faster in OpenCL compute?

A: The RTX A2000 is 70.6% faster, scoring 67,695 versus the Quadro M6000’s 39,688 in Geekbench OpenCL.

Q: Does the Quadro M6000 support ray tracing?

A: No. The M6000 has no ray tracing cores listed, while the RTX A2000 includes 26 RT cores. The M6000 is also limited to DirectX 12 (12_1), whereas the A2000 supports DirectX 12 Ultimate (12_2).

Q: How much memory does each card have?

A: The RTX A2000 has 6 GB of GDDR6 on a 192-bit bus, while the Quadro M6000 has 12 GB of GDDR5 on a 384-bit bus. The M6000 also has higher bandwidth at 317.4 GB/s versus 288.0 GB/s.

Q: What is the power consumption difference?

A: The RTX A2000 has a 70 W TDP and requires no power connectors, while the Quadro M6000 has a 250 W TDP and requires a single 8-pin connector. The recommended PSU is 250 W for the A2000 and 600 W for the M6000.

Q: Which card has better pixel fill rate?

A: The Quadro M6000 wins significantly, with 106.9 GPixel/s versus the RTX A2000’s 57.60 GPixel/s. This is due to the M6000’s 96 ROPs compared to the A2000’s 48 ROPs.

Q: Are these cards still in production?

A: Both are end-of-life. The RTX A2000 was released on August 9, 2021, with a launch MSRP of 449 USD, while the Quadro M6000 was released on March 20, 2015, with no listed launch MSRP.

Head-to-Head Benchmarks

The Geekbench OpenCL test reveals the starkest performance gap. The RTX A2000 scores 67,695, which is 70.6% higher than the Quadro M6000’s 39,688. This margin is massive — it suggests the A2000 handles general-purpose compute workloads nearly twice as fast. The A2000’s FP32 throughput of 7.987 TFLOPS versus 6.844 TFLOPS explains part of this, but the architectural advantages of Ampere (such as the 1:1 FP16 ratio) likely contribute more. The M6000’s 3072 shading units are actually close to the A2000’s 3328, yet the compute output is far lower.

In Geekbench Vulkan, the gap narrows but remains decisive. The A2000 scores 69,089 against the M6000’s 46,913, a 47.3% advantage. This smaller differential suggests the M6000’s Maxwell architecture handles Vulkan’s explicit multi-threading comparatively better than it handles OpenCL, but it still cannot match the A2000’s modern feature set. The A2000’s support for Vulkan 1.4 with hardware ray tracing and tensor cores gives it a structural advantage in any workload that leverages those features.

Looking at average benchmark scores across all tests, the A2000’s 46,043 average is 6.3% above the M6000’s 43,301. This smaller gap than the head-to-head margins suggests the M6000 performs relatively better in tests not covered by the two Geekbench runs. However, the A2000 wins 2 out of 2 head-to-head benchmarks, with zero wins for the M6000. The A2000’s 85th percentile versus the M6000’s 84th percentile confirms they are close in overall GPU rankings, but the A2000 consistently edges ahead in compute-oriented tests.

Specification Differences

The two cards differ in nearly every specification category. The RTX A2000 uses a GA106 chip on an 8 nm Samsung process, while the Quadro M6000 uses a GM200 chip on a 28 nm TSMC process. Transistor counts are 12,000 million for the A2000 versus 8,000 million for the M6000, with die sizes of 276 mm² and 601 mm² respectively. The A2000’s transistor density is 43.5 million per mm², more than triple the M6000’s 13.3 million per mm².

Memory configurations are a key differentiator. The A2000 has 6 GB of GDDR6 with a 192-bit bus and 288.0 GB/s bandwidth. The M6000 has 12 GB of GDDR5 with a 384-bit bus and 317.4 GB/s bandwidth. Clock speeds differ: the A2000 runs at 562 MHz base and 1200 MHz boost, while the M6000 runs at 988 MHz base and 1114 MHz boost. Memory clocks are 1500 MHz (12 Gbps effective) for the A2000 and 1653 MHz (6.6 Gbps effective) for the M6000.

Compute resources show the generational shift. The A2000 has 3,328 shading units, 104 TMUs, 48 ROPs, 26 RT cores, and 104 tensor cores. The M6000 has 3,072 shading units, 192 TMUs, and 96 ROPs, but no RT or tensor cores. Pixel rate is 57.60 GPixel/s for the A2000 versus 106.9 GPixel/s for the M6000. Texture rate is 124.8 GTexel/s versus 213.9 GTexel/s. FP32 performance is 7.987 TFLOPS for the A2000 and 6.844 TFLOPS for the M6000, with FP16 only listed for the A2000 at 7.987 TFLOPS.

Power and physical specifications differ dramatically. The A2000 has a 70 W TDP with no power connectors and a 250 W suggested PSU. The M6000 has a 250 W TDP with one 8-pin connector and a 600 W suggested PSU. The A2000 measures 167 mm in length and 69 mm in height, while the M6000 measures 267 mm by 111 mm. Both are dual-slot cards. The A2000 uses PCIe 4.0 x16, while the M6000 uses PCIe 3.0 x16. Display outputs are 4x mini-DisplayPort 1.4a for the A2000 and 1x DVI plus 4x DisplayPort 1.2 for the M6000. API support shows the A2000 with DirectX 12 Ultimate (12_2) versus the M6000’s DirectX 12 (12_1), with both supporting OpenGL 4.6 and Vulkan 1.4.