NVIDIA GeForce GTX 1660 Ti vs NVIDIA Quadro K620 Comparison

FAQ

Q: What are the key differences in architecture between the two GPUs?

A: The GeForce GTX 1660 Ti is built on the Turing architecture with a 12 nm process node from TSMC, while the Quadro K620 uses the older Maxwell architecture on a larger 28 nm node from TSMC with a smaller die size of 148 mm².

Q: Which GPU has more memory bandwidth and capacity?

A: The GTX 1660 Ti has a significant advantage with 6 GB of GDDR6 memory on a 192-bit interface, delivering 288.0 GB/s of bandwidth. The Quadro K620 has 2 GB of DDR3 memory on a 128-bit bus, offering much lower bandwidth at 28.80 GB/s.

Q: What is the performance difference in compute workloads?

A: The GeForce GTX 1660 Ti, scoring 662.4% ahead of the Quadro K620 in the Geekbench OpenCL test, wins by a massive margin in raw compute, with a score of 51029 versus 6693 for the Quadro.

Q: Are there differences in power consumption and form factor?

A: The Quadro K620 has a lower TDP of 45 W and is a single-slot card, requiring no extra power connectors and a 200 W PSU. The GTX 1660 Ti has a higher TDP of 120 W, dual-slot width, 1x 8-pin power connector, and is a longer card at 160mm.

Q: Which GPU has a higher average benchmark score and percentile ranking?

A: The GTX 1660 Ti has a higher average benchmark score of 7723, placing it in the 41st percentile of all GPUs. The Quadro K620 has a lower average score of 6282, sitting in the 36th percentile of the database.

Q: What is the release timeline and MSRP difference between the two GPUs?

A: The GTX 1660 Ti was released in 2019 as an end-of-life product with a launch MSRP of 279 USD. The Quadro K620 was released in 2014, is also end-of-life, and has no MSRP listed in the database.

Architecture Differences

The two NVIDIA GPUs are worlds apart architecturally. The GeForce GTX 1660 Ti uses the TU116 chip based on the Turing architecture, fabricated on a 12 nm process node at TSMC. It packs 6,600 million transistors across a 284 mm² die, giving it a transistor density of 23.2 million per square millimeter and a total FP32 throughput of 5.5 TFLOPS. Its memory subsystem includes 6 GB of GDDR6 on a 192-bit bus with a bandwidth of 288.0 GB/s. The Quadro K620, in contrast, uses the older GM107 chip with the Maxwell architecture on a 28 nm TSMC process with 1,870 million transistors on a smaller 148 mm² die size of 12.6 million per mm². The GTX 1660 Ti also has more shading units, with 1536 shading units, 96 texture mapping units, and 48 ROPs, compared to the Quadro's 384 shading units, 24 TMUs and 16 ROPs. The differences in the feature set reflect their different target markets: the GTX 1660 Ti supports DirectX 12 (12_1), OpenGL 4.6 and Vulkan 12_1 APIs, and DDR3 memory on a 128-bit bus, resulting in a far lower bandwidth of 28.80 GB/s.

The GTX 1660 Ti also has a faster memory clock of 1500 MHz (12 Gbps effective) versus the Quadro's 900 MHz. The GTX 1660 Ti has a higher base and boost clock, 1500 MHz and 1770 MHz, respectively, against the Quadro's 1058 MHz base and 1124 MHz boost. The GTX 1660 Ti also has a different set of display outputs, including 1x DVI, 1x HDMI 2.0, and 1x DisplayPort 1.4a, while the Quadro has 1x DVI and 1x DisplayPort 1.2. The GTX 1660 Ti card is physically larger, with a length of 229 mm, a height of 111 mm, and a width of 35 mm, compared to the Quadro's 160 mm length and 69 mm height. The bus interface of the GTX 1660 Ti is PCIe 3.0 x16, while the Quadro K620, released in 2014, uses PCIe 2.0 x16 and has a predecessor of Quadro Fermi, vs the GTX 1660 Ti's predecessor of GeForce 10.

The GTX 1660 Ti has a successor of GeForce 20, and the Quadro K620 has a successor of Quadro Maxwell and a predecessor of Quadro Fermi. The GTX 1660 Ti has a die size of 284 mm², while the Quadro has a die size of 148 mm².

Where Each One Wins

The data clearly shows that the GeForce GTX 1660 Ti is the dominant performer in compute-heavy tasks. In the Geekbench OpenCL benchmark, a test that measures raw compute performance across a wide range of workloads, the GTX 1660 Ti scores 51029, which is 662.4% higher than the Quadro K620's 6693.

In terms of gaming and DirectX performance, the GTX 1660 Ti also has a large win in the Passmark DirectX 11 benchmark, with a score of 103, and Passmark DirectX 9, with a score of 190. Its Passmark G3D score is 12871, indicating a much stronger overall graphics and gaming performance.

The Quadro K620, being an older, lower-power card, has a much smaller footprint, with lower power consumption (45W). Its advantage comes in a niche area: the Geekbench Vulkan test, where it scores 5870 versus the GTX 1660 Ti's 5827, a difference of 0.7%. This is a narrow win, but it shows the Quadro can hold its own in Vulkan API workloads. The Quadro also has a lower power draw, making it suitable for environments where power and space are at a premium. The GTX 1660 Ti, with its higher TDP of 120 W, requires a 300 W PSU and a dual-slot cooler. The Quadro's single-slot design and 160 mm length make it more fitting for small form factor systems.

The GTX 1660 Ti also wins in terms of average benchmark scores, with an average score of 7723. The Quadro's average score is 6282.

In summary, the GTX 1660 Ti wins decisively in raw compute power and is positioned as a far more capable GPU for gaming, CAD, and general 3D work.

The Quadro K620 wins in Vulkan API performance and is a more power-efficient card in terms of TDP and slot width.

The GTX 1660 Ti wins in OpenCL compute, DirectX 11, DirectX 9, Passmark G3D, Passmark GPU Compute, and Geekbench OpenCL. The Quadro K620 wins in the Geekbench Vulkan test.

Specification Differences

The GTX 1660 Ti has 1536 shading units, 96 TMUs, and 48 ROPs, while the Quadro K620 has 384 shading units, 24 TMUs, and 16 ROPs. The GTX 1660 Ti has 6 GB of GDDR6 memory, a 192-bit bus, and 288.0 GB/s bandwidth, while the Quadro has 2 GB of DDR3, a 128-bit bus, and 28.80 GB/s bandwidth. The GTX 1660 Ti is built on the Turing architecture with a 12 nm process node from TSMC, while the Quadro uses Maxwell on a 28 nm node. The GTX 1660 Ti has a base clock of 1500 MHz and a boost clock of 1770 MHz, while the Quadro has a base clock of 1058 MHz and a boost clock of 1124 MHz. The GTX 1660 Ti has a TDP of 120 W and requires a 300 W PSU, while the Quadro has a TDP of 45 W and a 200 W PSU. The GTX 1660 Ti is dual-slot with a 1x 8-pin power connector, while the Quadro is single-slot with no power connector. The GTX 1660 Ti uses a PCIe 3.0 x16 interface, the Quadro uses PCIe 2.0 x16. The GTX 1660 Ti has a die size of 284 mm² and 6,600 million transistors, while the Quadro has a die size of 148 mm² and 1,870 million transistors. The GTX 1660 Ti has a transistor density of 23.2M/mm², the Quadro has 12.6M/mm².

Head-to-Head Benchmarks

The recorded data contains two shared benchmarks between these cards, and they tell a lopsided story. In Geekbench OpenCL, the GTX 1660 Ti achieved a score of 51029, while the Quadro K620 managed 6693. The delta is 662.4% in favor of the GTX 1660 Ti, a decisive win that shows the GTX 1660 Ti's compute advantage. The GTX 1660 Ti's FP32 throughput is 5.437 TFLOPS, while the Quadro's is only 863.2 GFLOPS, a massive difference that aligns with the OpenCL result. The GTX 1660 Ti's texture rate is 169.9 GTexel/s and its pixel rate is 84.96 GPixel/s, while the Quadro's texture rate is 26.98 GTexel/s and its pixel rate is 17.98 GPixel/s. The GTX 1660 Ti also has a much higher memory bandwidth of 288.0 GB/s versus the Quadro's 28.80 GB/s, which feeds its high fill rates.

The other shared test is Geekbench Vulkan. Here, the Quadro K620 narrowly wins, scoring 5870 against the GTX 1660 Ti's 5827, a delta of 0.7%. This is a unique outcome given the GTX 1660 Ti's overwhelming advantage in OpenCL. The Vulkan score may reflect driver optimizations or architectural differences in the Maxwell GPU, but the win is narrow and does not change the overall picture of the GTX 1660 Ti being the much faster card. The GTX 1660 Ti's average benchmark score is 7723, compared to the Quadro's 6282. The GTX 1660 Ti also scores 12871 in Passmark G3D, while the Quadro does not have a comparable G3D score. In Passmark DirectX 11, the GTX 1660 Ti scores 103, and the Quadro scores no DirectX 11 test. The GTX 1660 Ti's Passmark GPU Compute score is 5226. The data confirms that the GTX 1660 Ti wins the head-to-head in 1 out of 2 shared tests, with the Quadro winning the other. These are the only two benchmarks the database has for both products, so the overall picture is that the GTX 1660 Ti is the dominant performer.