NVIDIA GeForce GTX 1060 6 GB vs NVIDIA Tesla K20m Comparison

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 1060 6 GB records an average benchmark score of 21,856, while the NVIDIA Tesla K20m records 19,089. That puts the GTX 1060 about 14.5% higher overall.

Q: How do the two compare in the Geekbench OpenCL test?

A: The GTX 1060 scores 32,322 in Geekbench OpenCL, versus 16,241 for the Tesla K20m. The GTX 1060 wins by 99%, a near-doubling of the compute score.

Q: Does the Tesla K20m win any benchmark?

A: Yes. In Geekbench Vulkan, the Tesla K20m scores 21,936 against 18,659 for the GTX 1060, a 14.9% advantage for the Tesla part.

Q: What are the percentile rankings of these GPUs among all GPUs?

A: The GTX 1060 sits at the 67th percentile, while the Tesla K20m is at the 64th percentile. Both are mid-pack performers, with the GTX 1060 slightly higher.

Q: Which card has more shading units?

A: The Tesla K20m has 2,496 shading units, nearly double the GTX 1060's 1,280. However, the GTX 1060 still achieves higher FP32 throughput at 4.375 TFLOPS versus 3.524 TFLOPS.

Q: What is the memory configuration difference?

A: The GTX 1060 has 6 GB of GDDR5 on a 192-bit bus with 192.2 GB/s bandwidth. The Tesla K20m has 5 GB of GDDR5 on a 320-bit bus with 208.0 GB/s bandwidth.

Architecture Differences

The two GPUs come from entirely different architectural generations and design philosophies. The GTX 1060 uses the GP106 chip built on TSMC's 16 nm process, part of the Pascal architecture from the GeForce 10-series. It packs 4,400 million transistors into a 200 mm² die, yielding a transistor density of 22.0M per mm². Its base clock is 1506 MHz with a boost of 1709 MHz, and memory runs at 2002 MHz (8 Gbps effective).

The Tesla K20m, by contrast, is built on the GK110 chip using TSMC's older 28 nm process, part of the Kepler architecture from the Tesla Kxx generation. It contains 7,080 million transistors on a much larger 561 mm² die, but the density is lower at 12.6M per mm². The K20m has no listed base or boost clocks, only a memory clock of 1300 MHz (5.2 Gbps effective). This older node and larger die explain its higher 225 W TDP versus 120 W for the GTX 1060.

The shading unit count heavily favors the Tesla: 2,496 versus 1,280. The Tesla also has more texture mapping units (208 vs 80), but fewer ROPs (40 vs 48). Despite fewer cores, the GTX 1060 achieves higher pixel rate (82.03 GPixel/s vs 36.71 GPixel/s) and nearly comparable texture rate (136.7 GTexel/s vs 146.8 GTexel/s). The FP32 output tells the story: the GTX 1060 delivers 4.375 TFLOPS, while the Tesla K20m delivers 3.524 TFLOPS. The GTX 1060 also supports FP16 at 68.36 GFLOPS (1:64), a feature absent from the Tesla.

The Tesla K20m is a compute-oriented card with no display outputs, while the GTX 1060 offers 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.4a. The Tesla also uses PCIe 2.0 x16, whereas the GTX 1060 uses PCIe 3.0 x16. DirectX support differs: the GTX 1060 reaches 12 (12_1), while the Tesla only reaches 12 (11_0). Vulkan support is newer on the GTX 1060 (1.4) versus 1.2.175 on the Tesla.

The Verdict

The data points to a clear split in intended use. For general-purpose compute, particularly OpenCL workloads, the GTX 1060 is the stronger choice by a wide margin. Its 99% lead in Geekbench OpenCL and higher FP32 throughput make it the more capable compute card in raw numbers, despite having fewer shading units. The GTX 1060 also offers better energy efficiency, with a 120 W TDP versus 225 W, and a smaller physical footprint at 250 mm length versus 267 mm.

However, the Tesla K20m wins in Vulkan workloads, posting a 14.9% higher score. For applications that leverage Vulkan's lower-level API, the Tesla's larger core count and wider memory bus appear to provide an edge. The Tesla also has slightly higher memory bandwidth (208.0 GB/s vs 192.2 GB/s) and a wider 320-bit bus, which could benefit memory-intensive tasks.

The GTX 1060 is the more versatile card overall, thanks to its display outputs and modern feature set. The Tesla K20m is a legacy compute accelerator with no display outputs, suited for server or workstation environments where rendering to screen is unnecessary. For a desktop user, the GTX 1060 is the obvious pick. For a compute node relying on Vulkan, the Tesla K20m has a specific advantage worth considering.

Specification Differences

| Field | NVIDIA GeForce GTX 1060 6 GB | NVIDIA Tesla K20m |

|-------|------------------------------|-------------------|

| Process node | 16 nm | 28 nm |

| Transistors | 4,400 million | 7,080 million |

| Die size | 200 mm² | 561 mm² |

| Transistor density | 22.0M / mm² | 12.6M / mm² |

| Base clock | 1506 MHz | Not listed |

| Boost clock | 1709 MHz | Not listed |

| Memory clock | 2002 MHz (8 Gbps effective) | 1300 MHz (5.2 Gbps effective) |

| Memory size | 6 GB | 5 GB |

| Memory bus width | 192 bit | 320 bit |

| Memory bandwidth | 192.2 GB/s | 208.0 GB/s |

| Shading units | 1280 | 2496 |

| TMUs | 80 | 208 |

| ROPs | 48 | 40 |

| Pixel rate | 82.03 GPixel/s | 36.71 GPixel/s |

| Texture rate | 136.7 GTexel/s | 146.8 GTexel/s |

| FP32 | 4.375 TFLOPS | 3.524 TFLOPS |

| FP16 | 68.36 GFLOPS (1:64) | Not listed |

| TDP | 120 W | 225 W |

| Power connectors | 1x 6-pin | 1x 6-pin + 1x 8-pin |

| Suggested PSU | 300 W | 550 W |

| Bus interface | PCIe 3.0 x16 | PCIe 2.0 x16 |

| Display outputs | 1x DVI, 1x HDMI 2.0, 3x DisplayPort 1.4a | No outputs |

| DirectX version | 12 (12_1) | 12 (11_0) |

| Vulkan version | 1.4 | 1.2.175 |

| Length | 250 mm (9.8 inches) | 267 mm (10.5 inches) |

Head-to-Head Benchmarks

The recorded head-to-head data shows exactly one win for each card. In Geekbench OpenCL, the GTX 1060 delivers a score of 32,322 against 16,241 for the Tesla K20m. That is a 99% delta, meaning the GTX 1060 nearly doubles the Tesla's OpenCL performance. This is the largest single difference in any benchmark between the two.

In Geekbench Vulkan, the tables turn. The Tesla K20m scores 21,936, while the GTX 1060 scores 18,659. That gives the Tesla a 14.9% advantage. The Vulkan result is notable because it shows the Tesla's older architecture still has strengths in certain API environments. The delta is smaller than the OpenCL gap, but it is a clear and measurable win.

The average benchmark score for each card reflects these results. The GTX 1060 averages 21,856 across its three recorded tests (Geekbench Metal, OpenCL, and Vulkan), while the Tesla K20m averages 19,089 across its two tests (OpenCL and Vulkan). The GTX 1060's average is higher, but the Tesla's Vulkan performance keeps the overall picture competitive.

Looking at nearest rivals, the GTX 1060 sits between the NVIDIA Quadro RTX 5000 (21,629, 1% lower) and the AMD Radeon RX 6700S (22,154, 1.3% higher). The Tesla K20m sits just above the NVIDIA GeForce RTX 4050 Mobile (19,049, 0.2% lower) and the AMD Radeon RX 6600 (19,036, 0.3% lower). These proximity figures confirm that neither card is an outlier; both perform within a narrow band of their peers.

Where Each One Wins

The GTX 1060 wins decisively in OpenCL compute workloads, as shown by the 99% lead. This makes it the better option for general-purpose GPU computing, including tasks like physics simulation, image processing, or any OpenCL-based application. Its higher FP32 throughput (4.375 TFLOPS) and lower power draw (120 W) reinforce this advantage. The GTX 1060 also wins on efficiency: it achieves more performance per watt and per square millimeter of die area.

The Tesla K20m wins in Vulkan performance, with a 14.9% higher score. This suggests the Tesla's architecture, with its 2,496 shading units and 320-bit memory bus, handles Vulkan's explicit multi-threading and low-overhead model more effectively. For developers targeting Vulkan specifically, the Tesla K20m offers a measurable edge. Its higher memory bandwidth (208.0 GB/s) may also benefit workloads that saturate the memory subsystem.

For a desktop workstation with display needs, the GTX 1060 is the only viable option, as the Tesla K20m has no display outputs. For a headless compute server running Vulkan workloads, the Tesla K20m's advantage in that benchmark makes it worth considering. In all other measured areas, the GTX 1060 leads, making it the more balanced and practical choice for most users. The data shows a 1-1 split in head-to-head wins, but the magnitude of the GTX 1060's OpenCL victory (99%) outweighs the Tesla's Vulkan win (14.9%) when weighing overall utility.