NVIDIA GeForce GTX 1070 Ti vs NVIDIA Tesla K20Xm Comparison

FAQ

Q: How does the NVIDIA GeForce GTX 1070 Ti compare to the NVIDIA Tesla K20Xm in average benchmark score?

A: The GTX 1070 Ti has an average benchmark score of 14277, while the Tesla K20Xm has an average benchmark score of 12625. This places the GTX 1070 Ti in the 56th percentile of all GPUs, compared to the 52nd percentile for the Tesla K20Xm.

Q: Which card wins in the Geekbench Metal test?

A: The GTX 1070 Ti scores 22180 in Geekbench Metal, while the Tesla K20Xm scores 8035. The GTX 1070 Ti wins this test with a delta of 176%.

Q: What is the difference in Geekbench OpenCL scores?

A: The GTX 1070 Ti scores 49219 in Geekbench OpenCL, while the Tesla K20Xm scores 17215. The GTX 1070 Ti wins by 185.9%.

Q: What are the closest rivals to the GTX 1070 Ti according to the database?

A: The nearest rivals are the Intel Iris Xe MAX Graphics (average score 14315, delta -0.3%), AMD Radeon Vega 11 (14352, -0.5%), NVIDIA GeForce GTX TITAN (14373, -0.7%), and AMD Radeon RX Vega 11 (14385, -0.7%). All of these are slightly below the GTX 1070 Ti in average score.

Q: What are the closest rivals to the Tesla K20Xm?

A: The nearest rivals are the AMD Radeon RX 7600M XT (average score 12710, delta -0.7%), NVIDIA GeForce GTX 670 (12773, -1.2%), NVIDIA GeForce GTX 590 (12830, -1.6%), and AMD Radeon Pro 455 (12831, -1.6%). These rivals are all slightly above the Tesla K20Xm in average score.

Q: How many benchmark wins does each card have in head-to-head testing?

A: The GTX 1070 Ti wins 2 head-to-head benchmarks, while the Tesla K20Xm wins 0. The GTX 1070 Ti wins both available tests decisively.

Architecture Differences

The NVIDIA GeForce GTX 1070 Ti and the NVIDIA Tesla K20Xm represent two fundamentally different architectural eras. The GTX 1070 Ti uses the GP104 chip based on the Pascal architecture, fabricated on a 16 nm process at TSMC. In contrast, the Tesla K20Xm uses the GK110 chip based on the Kepler architecture, fabricated on a 28 nm process at TSMC. This process difference is significant: the 16 nm node allows the GTX 1070 Ti to pack 7,200 million transistors into a 314 mm² die, achieving a transistor density of 22.9M per mm². The Tesla K20Xm, with 7,080 million transistors on a 561 mm² die, has a transistor density of only 12.6M per mm².

The GTX 1070 Ti is part of the GeForce 10-series, while the Tesla K20Xm belongs to the Tesla Kepler (Kxx) generation. The former is a consumer gaming card, while the latter is a compute-oriented accelerator. This is reflected in their display outputs: the GTX 1070 Ti offers 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.4a outputs, while the Tesla K20Xm has no display outputs at all, indicating its server or workstation focus.

The GTX 1070 Ti features 2432 shading units, 152 TMUs, and 64 ROPs. The Tesla K20Xm has more shading units at 2688, but fewer TMUs at 224 and fewer ROPs at 48. The GTX 1070 Ti supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The Tesla K20Xm supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. Notably, the GTX 1070 Ti has a higher API feature level for DirectX 12.

The GTX 1070 Ti was released on 2017-11-01, while the Tesla K20Xm was released on 2012-11-11, a difference of five years. Both are now end-of-life products. The GTX 1070 Ti's predecessor is the GeForce 900 series and its successor is the GeForce 20 series. The Tesla K20Xm's predecessor is Tesla Fermi and its successor is Tesla Maxwell.

Head-to-Head Benchmarks

The database records two head-to-head benchmark comparisons between these cards, and the GTX 1070 Ti wins both by wide margins. The first test is Geekbench Metal, where the GTX 1070 Ti scores 22180 against the Tesla K20Xm's 8035. This represents a delta of 176%, meaning the GTX 1070 Ti is nearly three times faster in this particular workload. This is a substantial gap, and it reflects the architectural improvements in the Pascal generation over Kepler.

The second head-to-head test is Geekbench OpenCL, where the GTX 1070 Ti scores 49219 and the Tesla K20Xm scores 17215. The delta here is 185.9%, an even larger margin than in the Metal test. The GTX 1070 Ti's advantage in OpenCL is particularly pronounced, showing that the newer architecture handles general-purpose compute workloads with significantly greater efficiency.

The overall head-to-head record stands at 2 wins for the GTX 1070 Ti and 0 wins for the Tesla K20Xm. The GTX 1070 Ti also holds a higher average benchmark score of 14277 compared to the Tesla K20Xm's 12625. This difference of 1652 points translates to the GTX 1070 Ti sitting in the 56th percentile of all GPUs, while the Tesla K20Xm sits in the 52nd percentile.

The GTX 1070 Ti's nearest rivals, such as the Intel Iris Xe MAX Graphics and AMD Radeon Vega 11, are within 0.3% to 0.7% of its average score, indicating that the GTX 1070 Ti sits at the upper edge of its performance tier. The Tesla K20Xm's nearest rivals, including the AMD Radeon RX 7600M XT and NVIDIA GeForce GTX 670, are within 0.7% to 1.6% above its average score, placing the Tesla K20Xm slightly below its nearest competitors.

Specification Differences

The GTX 1070 Ti and Tesla K20Xm differ across nearly every specification category. The GTX 1070 Ti uses a 16 nm process, while the Tesla K20Xm uses 28 nm. The GTX 1070 Ti has a die size of 314 mm², while the Tesla K20Xm has a die size of 561 mm². Transistor counts are similar at 7,200 million and 7,080 million respectively, but transistor density differs sharply: 22.9M per mm² for the GTX 1070 Ti versus 12.6M per mm² for the Tesla K20Xm.

Clock speeds differ significantly. The GTX 1070 Ti has a base clock of 1607 MHz and a boost clock of 1683 MHz. The Tesla K20Xm has no recorded base or boost clock in the database. Memory clocks also differ: the GTX 1070 Ti runs at 2002 MHz with 8 Gbps effective, while the Tesla K20Xm runs at 1300 MHz with 5.2 Gbps effective.

Memory configurations are notably different. The GTX 1070 Ti has 8 GB of GDDR5 memory on a 256-bit bus, yielding 256.3 GB/s of bandwidth. The Tesla K20Xm has 6 GB of GDDR5 memory on a 384-bit bus, yielding 249.6 GB/s of bandwidth. Despite the wider bus on the Tesla K20Xm, the higher memory clock on the GTX 1070 Ti gives it slightly more bandwidth.

Compute rates show a clear advantage for the GTX 1070 Ti. It delivers 8.186 TFLOPS of FP32 performance, while the Tesla K20Xm delivers 3.935 TFLOPS. The GTX 1070 Ti also has FP16 capability at 127.9 GFLOPS (1:64), while the Tesla K20Xm has no recorded FP16 performance. Pixel rates are 107.7 GPixel/s for the GTX 1070 Ti versus 40.99 GPixel/s for the Tesla K20Xm. Texture rates are 255.8 GTexel/s versus 164.0 GTexel/s.

Power consumption differs substantially. The GTX 1070 Ti has a TDP of 180 W with a suggested PSU of 450 W. The Tesla K20Xm has a TDP of 235 W with a suggested PSU of 550 W. Both are dual-slot cards, and both are 267 mm (10.5 inches) in length. The GTX 1070 Ti has a height of 112 mm and width of 40 mm, while the Tesla K20Xm has no recorded height or width. The GTX 1070 Ti uses a 1x 8-pin power connector, while the Tesla K20Xm has no recorded power connector configuration.

The Verdict

The data clearly favors the NVIDIA GeForce GTX 1070 Ti across every measurable benchmark. In the two head-to-head tests recorded in the database, the GTX 1070 Ti wins both, with deltas of 176% and 185.9%. The average benchmark score also confirms this dominance: 14277 for the GTX 1070 Ti versus 12625 for the Tesla K20Xm. The GTX 1070 Ti sits in the 56th percentile of all GPUs, while the Tesla K20Xm sits in the 52nd percentile.

The GTX 1070 Ti is the newer product, released in 2017 versus 2012 for the Tesla K20Xm, and it benefits from a more advanced 16 nm manufacturing process. Its FP32 compute rate of 8.186 TFLOPS is more than double the Tesla K20Xm's 3.935 TFLOPS. The GTX 1070 Ti also has more memory at 8 GB versus 6 GB, and slightly higher memory bandwidth at 256.3 GB/s versus 249.6 GB/s.

For any user choosing between these two cards, the GTX 1070 Ti is the superior option. It offers better performance in every recorded benchmark, higher compute throughput, more memory, better API support, and lower power consumption at 180 W versus 235 W. The Tesla K20Xm's only potential advantage is its higher shading unit count at 2688 versus 2432, but this does not translate into better benchmark results.

The Tesla K20Xm was positioned as a compute accelerator with no display outputs, while the GTX 1070 Ti is a fully featured consumer graphics card with multiple display outputs. The GTX 1070 Ti supports DirectX 12 (12_1), while the Tesla K20Xm only supports DirectX 12 (11_0), which may matter for modern gaming titles. The GTX 1070 Ti also supports Vulkan 1.4 versus Vulkan 1.2.175 for the Tesla K20Xm.

Where Each One Wins

The NVIDIA GeForce GTX 1070 Ti wins in all recorded benchmark categories. In Geekbench Metal, it scores 22180 versus 8035 for the Tesla K20Xm, a 176% advantage. In Geekbench OpenCL, it scores 49219 versus 17215, a 185.9% advantage. These results indicate that the GTX 1070 Ti is the clear winner for compute workloads, whether measured through Metal or OpenCL APIs.

The GTX 1070 Ti also wins in terms of overall performance class. Its average benchmark score of 14277 places it just above rivals like the Intel Iris Xe MAX Graphics (14315, -0.3%) and AMD Radeon Vega 11 (14352, -0.5%). The Tesla K20Xm, with an average score of 12625, sits below rivals like the AMD Radeon RX 7600M XT (12710, -0.7%) and NVIDIA GeForce GTX 670 (12773, -1.2%). This means the GTX 1070 Ti is competitive within its performance tier, while the Tesla K20Xm is slightly behind its nearest peers.

For gaming and consumer use, the GTX 1070 Ti is the only viable choice, as the Tesla K20Xm has no display outputs. The GTX 1070 Ti offers 8 GB of GDDR5 memory, which is ample for modern games, and its higher pixel rate of 107.7 GPixel/s and texture rate of 255.8 GTexel/s suggest strong rasterization performance. The Tesla K20Xm's lower pixel rate of 40.99 GPixel/s and texture rate of 164.0 GTexel/s would be a bottleneck in graphics-heavy workloads.

For compute-oriented tasks, the GTX 1070 Ti still wins based on the recorded data. Its FP32 throughput of 8.186 TFLOPS is more than double the Tesla K20Xm's 3.935 TFLOPS. The GTX 1070 Ti also has FP16 capability at 127.9 GFLOPS, which the Tesla K20Xm lacks entirely. The Tesla K20Xm's only potential niche would be in legacy compute environments that specifically require Kepler architecture, but the benchmark data does not show any workload where it outperforms the GTX 1070 Ti.