NVIDIA GeForce GTX 1650 SUPER vs NVIDIA Tesla K20Xm Comparison

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA Tesla K20Xm records an average benchmark score of 12625, while the NVIDIA GeForce GTX 1650 SUPER averages 11047. The Tesla K20Xm sits at the 52nd percentile among all GPUs, and the GTX 1650 SUPER sits at the 50th percentile.

Q: How does the Tesla K20Xm compare to its nearest rivals?

A: The Tesla K20Xm is within 1.6% of all its nearest rivals. It is 0.7% behind the AMD Radeon RX 7600M XT (12710), 1.2% behind the NVIDIA GeForce GTX 670 (12773), 1.6% behind the NVIDIA GeForce GTX 590 (12830), and 1.6% behind the AMD Radeon Pro 455 (12831).

Q: What is the closest rival to the GTX 1650 SUPER?

A: The closest rival is the AMD Radeon RX 550 with an average score of 11075, just 0.2% ahead. The NVIDIA RTX PRO 6000D Blackwell Max-Q and NVIDIA RTX PRO 6000 Blackwell Max-Q both score 11088, 0.4% ahead, while the NVIDIA GeForce MX350 scores 10883, 1.5% behind.

Q: Which GPU has more shading units?

A: The Tesla K20Xm has 2688 shading units, while the GTX 1650 SUPER has 1280. The Tesla K20Xm also has 224 texture mapping units versus 80, and 48 ROPs versus 32.

Q: What is the difference in memory configuration?

A: The Tesla K20Xm has 6 GB of GDDR5 on a 384-bit bus with 249.6 GB/s bandwidth. The GTX 1650 SUPER has 4 GB of GDDR6 on a 128-bit bus with 192.0 GB/s bandwidth.

Q: Which GPU has higher FP32 performance?

A: The GTX 1650 SUPER delivers 4.416 TFLOPS FP32, compared to 3.935 TFLOPS for the Tesla K20Xm. The GTX 1650 SUPER also offers 8.832 TFLOPS FP16 (2:1), while the Tesla K20Xm has no recorded FP16 figure.

Architecture Differences

The Tesla K20Xm is built on the GK110 chip using the Kepler architecture, fabricated on a 28 nm process at TSMC. It integrates 7,080 million transistors on a 561 mm² die, resulting in a transistor density of 12.6M per mm². The GTX 1650 SUPER uses the TU116 chip with the Turing architecture, fabricated on a 12 nm process, also at TSMC. It contains 6,600 million transistors on a 284 mm² die, giving a higher transistor density of 23.2M per mm².

The two GPUs belong to entirely different product generations. The Tesla K20Xm is from the Tesla Kepler generation (Kxx), released on 2012-11-11, with the Tesla Fermi as its predecessor and Tesla Maxwell as its successor. The GTX 1650 SUPER belongs to the GeForce 16 series, released on 2019-11-21, with the GeForce 10 series as its predecessor and GeForce 20 as its successor. The Tesla K20Xm is a compute-oriented accelerator with no display outputs, while the GTX 1650 SUPER is a consumer graphics card with 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a outputs.

Memory architecture differs substantially. The Tesla K20Xm uses 6 GB of GDDR5 on a 384-bit interface, running at 1300 MHz with 5.2 Gbps effective speed. The GTX 1650 SUPER uses 4 GB of GDDR6 on a 128-bit interface, running at 1500 MHz with 12 Gbps effective speed. The wider bus of the Tesla K20Xm yields higher bandwidth at 249.6 GB/s versus 192.0 GB/s, despite the newer memory type in the GTX 1650 SUPER.

Shading resources differ greatly. The Tesla K20Xm has 2688 shading units, 224 TMUs, and 48 ROPs. The GTX 1650 SUPER has 1280 shading units, 80 TMUs, and 32 ROPs. Neither GPU has ray tracing cores or tensor cores. The Tesla K20Xm has a higher texture rate at 164.0 GTexel/s versus 138.0 GTexel/s, but the GTX 1650 SUPER has a higher pixel rate at 55.20 GPixel/s versus 40.99 GPixel/s.

API support differs in DirectX versions. The Tesla K20Xm supports DirectX 12 (11_0), while the GTX 1650 SUPER supports DirectX 12 (12_1). Both support OpenGL 4.6. The Vulkan versions differ: the Tesla K20Xm supports Vulkan 1.2.175, while the GTX 1650 SUPER supports Vulkan 1.4.

Power requirements are significantly lower for the GTX 1650 SUPER. The Tesla K20Xm has a 235 W TDP with a suggested PSU of 550 W. The GTX 1650 SUPER has a 100 W TDP with a suggested PSU of 300 W and uses a single 6-pin power connector. The Tesla K20Xm lists no power connector details.

Physical dimensions differ as well. The Tesla K20Xm is 267 mm (10.5 inches) long. The GTX 1650 SUPER is 229 mm (9 inches) long, 111 mm (4.4 inches) high, and 35 mm (1.4 inches) wide. Both are dual-slot cards.

Head-to-Head Benchmarks

The recorded database contains one direct head-to-head benchmark between these two GPUs: the Geekbench OpenCL test. In this test, the GTX 1650 SUPER scores 43875, while the Tesla K20Xm scores 17215. The GTX 1650 SUPER wins by a margin of 60.8%. This is a substantial difference, putting the newer consumer card far ahead of the older compute accelerator in this particular compute workload.

The GTX 1650 SUPER also has additional benchmark entries that the Tesla K20Xm lacks. In Geekbench Vulkan, it scores 50519. In 3DMark Steel Nomad DX12, it scores 352. In Passmark tests, it records 10179 in G3D, 4477 in GPU compute, 749 in G2D, 148 in DirectX 9, 73 in DirectX 11, 50 in DirectX 10, and 45 in DirectX 12. The Tesla K20Xm has no recorded scores in these tests, limiting direct comparison to the one overlapping benchmark.

The Tesla K20Xm does have a Geekbench Metal score of 8035, which the GTX 1650 SUPER does not have recorded. However, the overall average scores tell a more nuanced story. The Tesla K20Xm averages 12625 across its two recorded benchmarks, while the GTX 1650 SUPER averages 11047 across ten recorded benchmarks. The discrepancy between the head-to-head result and the overall averages suggests the benchmark sets measure different aspects of performance, with the Tesla K20Xm's limited benchmark pool skewing its average upward.

When comparing percentile positions, the Tesla K20Xm sits at the 52nd percentile, above the GTX 1650 SUPER at the 50th percentile. This ranking is based on the full GPU database and places both cards near the middle of the distribution. The nearest rival data reinforces that these are mid-range performers in their respective eras.

The wins tally in the head-to-head data shows the GTX 1650 SUPER with 1 win and the Tesla K20Xm with 0 wins. This reflects the single overlapping benchmark result. For the GTX 1650 SUPER, the Geekbench OpenCL score of 43875 is roughly 2.5 times the Tesla K20Xm's 17215, a clear generational improvement in compute throughput despite the Tesla K20Xm's larger die and higher transistor count.

The Verdict

The data points to a clear split in use cases. For raw compute performance in the Geekbench OpenCL workload, the GTX 1650 SUPER is decisively ahead, delivering 60.8% higher performance than the Tesla K20Xm. This makes the GTX 1650 SUPER the better choice for general compute tasks that leverage OpenCL.

The Tesla K20Xm counters with a higher average benchmark score of 12625 versus 11047, and a slightly better percentile ranking at 52 versus 50. Its 6 GB memory capacity and 384-bit bus with 249.6 GB/s bandwidth give it an advantage in memory-heavy workloads, though the GTX 1650 SUPER's GDDR6 memory operates at a higher effective speed of 12 Gbps.

For consumers, the GTX 1650 SUPER is the practical choice. It has display outputs, a much lower 100 W TDP, a smaller physical footprint, and a suggested PSU of only 300 W. The Tesla K20Xm has no display outputs, a 235 W TDP, and a suggested PSU of 550 W, making it unsuitable for standard desktop use.

For compute-focused deployments, the Tesla K20Xm still shows value in its higher average score and proximity to rivals like the AMD Radeon RX 7600M XT (within 0.7%) and the NVIDIA GeForce GTX 670 (within 1.2%). However, the single head-to-head result indicates the GTX 1650 SUPER has superior OpenCL compute performance.

Users prioritizing memory bandwidth and capacity should consider the Tesla K20Xm. Users prioritizing power efficiency, modern API support including Vulkan 1.4 and DirectX 12 (12_1), and compute throughput should choose the GTX 1650 SUPER.

Specification Differences

| Specification | NVIDIA Tesla K20Xm | NVIDIA GeForce GTX 1650 SUPER |

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

| Chip | GK110 | TU116 |

| Architecture | Kepler | Turing |

| Generation | Tesla Kepler (Kxx) | GeForce 16 |

| Process Node | 28 nm | 12 nm |

| Transistors | 7,080 million | 6,600 million |

| Die Size | 561 mm² | 284 mm² |

| Transistor Density | 12.6M / mm² | 23.2M / mm² |

| Memory Size | 6 GB | 4 GB |

| Memory Type | GDDR5 | GDDR6 |

| Memory Bus Width | 384 bit | 128 bit |

| Memory Bandwidth | 249.6 GB/s | 192.0 GB/s |

| Memory Clock | 1300 MHz, 5.2 Gbps effective | 1500 MHz, 12 Gbps effective |

| Base Clock | Not specified | 1530 MHz |

| Boost Clock | Not specified | 1725 MHz |

| Shading Units | 2688 | 1280 |

| TMUs | 224 | 80 |

| ROPs | 48 | 32 |

| Pixel Rate | 40.99 GPixel/s | 55.20 GPixel/s |

| Texture Rate | 164.0 GTexel/s | 138.0 GTexel/s |

| FP32 | 3.935 TFLOPS | 4.416 TFLOPS |

| FP16 | Not specified | 8.832 TFLOPS (2:1) |

| TDP | 235 W | 100 W |

| Power Connectors | Not specified | 1x 6-pin |

| Suggested PSU | 550 W | 300 W |

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

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

| Vulkan | 1.2.175 | 1.4 |

| Length | 267 mm (10.5 inches) | 229 mm (9 inches) |

| Height | Not specified | 111 mm (4.4 inches) |

| Width | Not specified | 35 mm (1.4 inches) |

| Release Date | 2012-11-11 | 2019-11-21 |

| Predecessor | Tesla Fermi | GeForce 10 |

| Successor | Tesla Maxwell | GeForce 20 |

| Launch MSRP | 7,699 USD | 159 USD |