NVIDIA GeForce GTX 750 vs NVIDIA Quadro K620 Comparison

Head-to-Head Benchmarks

The database records two compute-focused benchmark comparisons between the NVIDIA GeForce GTX 750 and the NVIDIA Quadro K620. In both, the GTX 750 comes out decisively ahead, but the margins reveal a clear pattern about where each card is optimized.

In Geekbench OpenCL, the GTX 750 scores 9,315 against the Quadro K620’s 6,693. That is a 39.2% advantage, a substantial gap that places the consumer card in a different performance class for general compute workloads. The OpenCL test exercises raw throughput across the GPU, and the GTX 750’s higher shading-unit count and faster memory subsystem show up directly in this result.

The Vulkan test tells a similar story. The GTX 750 records 8,078 points, while the Quadro K620 manages 5,870. The 37.6% delta is slightly smaller than the OpenCL margin but still firmly in favor of the GeForce card. Vulkan is a lower-level API that benefits from efficient driver overhead and raw shader throughput, both of which favor the GTX 750’s architecture.

What stands out in the head-to-head data is not just that the GTX 750 wins both tests, but that the wins are consistent and large. A 39.2% lead in one test and a 37.6% lead in another suggests the performance gap is structural, not workload-specific. The Quadro K620 does not come close to matching the GTX 750 in either benchmark, despite sharing the same GM107 chip and Maxwell architecture.

Looking at the broader database context, the GTX 750’s average benchmark score is 7,222, placing it in the 40th percentile of all GPUs. Its nearest rivals include the AMD Radeon Vega 8 Mobile at 7,203, the NVIDIA GeForce GTX 560 SE at 7,171, and the Intel Iris Pro Graphics P580 at 7,170, all within 0.7% of the GTX 750’s score. Interestingly, the NVIDIA GeForce GTX 970 appears as a rival with an average score of 7,157, only 0.9% behind the GTX 750. This is a reminder that average scores can compress differences that are much larger in specific benchmark tests.

The Quadro K620’s average benchmark score is 6,282, placing it in the 36th percentile. Its nearest rivals include the NVIDIA GeForce RTX 5070 Ti SUPER and the RTX 4070 Ti SUPER AD102, both at 6,270 (0.2% behind), as well as the AMD Radeon R7 M350 at 6,327 (0.7% ahead) and the AMD Radeon Pro WX 4100 at 6,330 (0.8% ahead). These rivalries show the Quadro K620 sitting in a tightly packed cluster of low-to-mid-range GPUs, whereas the GTX 750 sits in a slightly higher cluster.

Where Each One Wins

The benchmark data points to a clear split: the GTX 750 wins in raw compute performance, while the Quadro K620 has no benchmark victories in this comparison. That does not mean the Quadro K620 is without merit, but its strengths lie outside the recorded tests.

For OpenCL workloads, the GTX 750’s 39.2% lead over the Quadro K620 is the single largest margin in this comparison. OpenCL is often used for general-purpose GPU computing, including scientific simulations, image processing, and some rendering tasks. The GTX 750’s 512 shading units, compared to the Quadro K620’s 384, explain much of this advantage. More shading units mean more parallel ALUs available for compute kernels, and the GTX 750’s faster memory (80.19 GB/s vs 28.80 GB/s) reduces bottlenecks when feeding those ALUs.

Vulkan performance also favors the GTX 750 by 37.6%. Vulkan’s explicit API design rewards GPUs with strong raw compute and minimal driver overhead. The GTX 750’s higher clock speeds (1,020 MHz base, 1,085 MHz boost) versus the Quadro K620’s (1,058 MHz base, 1,124 MHz boost) are interesting: the Quadro actually has higher clocks, but the GTX 750 still wins because it has 33% more shading units and 33% more texture units. Clock speed alone does not compensate for a shader-count deficit of that magnitude.

Where might the Quadro K620 hold an edge? The data shows it has 2 GB of memory versus the GTX 750’s 1 GB, and it uses DDR3 instead of GDDR5. While DDR3 is slower (28.80 GB/s vs 80.19 GB/s), the larger capacity could matter for certain workloads that need more memory footprint rather than more bandwidth. A 2 GB frame buffer can hold larger textures or datasets than a 1 GB one, even if feeding that memory is slower. The Quadro K620 also lists a DisplayPort 1.2 output, which may be preferable for professional display setups compared to the GTX 750’s mini-HDMI 1.4a.

The Quadro K620’s generation label, "Quadro Kepler (Kx200)", is curious since it uses the Maxwell GM107 chip. The naming suggests a transitional period in NVIDIA’s product stack, but the architecture is the same Maxwell as the GTX 750.

FAQ

Q: Which card is faster in Geekbench OpenCL?

A: The NVIDIA GeForce GTX 750 scores 9,315, which is 39.2% higher than the Quadro K620’s 6,693.

Q: Does the Quadro K620 win any benchmark in this comparison?

A: No. The GTX 750 wins both recorded tests, Geekbench OpenCL and Geekbench Vulkan, with deltas of 39.2% and 37.6% respectively.

Q: How do the average benchmark scores compare?

A: The GTX 750 has an average score of 7,222, while the Quadro K620 averages 6,282. This places them in the 40th and 36th percentiles of all GPUs respectively.

Q: What memory configuration does each card use?

A: The GTX 750 has 1,024 MB of GDDR5 on a 128-bit bus with 80.19 GB/s bandwidth. The Quadro K620 has 2 GB of DDR3 on a 128-bit bus with 28.80 GB/s bandwidth.

Q: Are the two cards based on the same chip?

A: Yes, both use the GM107 chip built on TSMC’s 28 nm process, with 1,870 million transistors and a 148 mm² die size.

Q: Which card has more shading units?

A: The GTX 750 has 512 shading units, while the Quadro K620 has 384. The GTX 750 also has 32 texture units versus 24 on the Quadro K620.

Specification Differences

The two cards share a common foundation but diverge in several key specifications. Both use the GM107 chip, Maxwell architecture, TSMC 28 nm process, 1,870 million transistors, and a 148 mm² die size. From there, the differences become significant.

Clock speeds favor the Quadro K620 slightly. Its base clock is 1,058 MHz and boost clock is 1,124 MHz, compared to the GTX 750’s 1,020 MHz base and 1,085 MHz boost. The Quadro runs about 3.7% faster at base and 3.6% faster at boost. Memory clocks are a different story: the GTX 750 runs at 1,253 MHz with 5 Gbps effective speed, while the Quadro K620 runs at 900 MHz with 1,800 Mbps effective speed. This is a massive difference in memory clock rates.

The memory subsystem differs in both capacity and type. The GTX 750 has 1,024 MB of GDDR5, while the Quadro K620 has 2 GB of DDR3. Both use a 128-bit bus, but the bandwidth numbers diverge sharply: 80.19 GB/s for the GTX 750 versus 28.80 GB/s for the Quadro K620. The GTX 750 has nearly three times the memory bandwidth.

Compute resources also differ. The GTX 750 has 512 shading units, 32 texture units, and 16 ROPs. The Quadro K620 has 384 shading units, 24 texture units, and 16 ROPs. This translates to a pixel rate of 17.36 GPixel/s for the GTX 750 and 17.98 GPixel/s for the Quadro K620 (the Quadro is slightly ahead here due to higher clocks), but texture rate favors the GTX 750 at 34.72 GTexel/s versus 26.98 GTexel/s. FP32 performance is 1,111.0 GFLOPS for the GTX 750 versus 863.2 GFLOPS for the Quadro K620.

Power and physical specifications show the Quadro K620 as the more modest card. It has a 45 W TDP and a recommended 200 W PSU, while the GTX 750 has a 55 W TDP and a recommended 250 W PSU. Both are single-slot with no power connectors. The Quadro K620 is longer at 160 mm versus 145 mm for the GTX 750, and it has specified dimensions of 69 mm height. Bus interfaces differ: the GTX 750 uses PCIe 3.0 x16, while the Quadro K620 uses PCIe 2.0 x16.

Display outputs also differ. The GTX 750 offers 2x DVI and 1x mini-HDMI 1.4a, while the Quadro K620 offers 1x DVI and 1x DisplayPort 1.2. The Quadro’s DisplayPort is a notable feature for professional environments that often rely on DisplayPort connectivity.

Release dates differ by about five months, with the GTX 750 launching on February 17, 2014, and the Quadro K620 on July 21, 2014. Both are end-of-life products.

Architecture Differences

Despite sharing the GM107 chip and Maxwell architecture, the two cards implement that architecture with different priorities. The most obvious structural difference is the compute resource allocation: the GTX 750 has 512 shading units and 32 TMUs, while the Quadro K620 has 384 shading units and 24 TMUs. This means the GTX 750 has 33% more shading units and 33% more texture units, a direct scaling of the Maxwell SM (streaming multiprocessor) design.

The memory architecture is where the two diverge most philosophically. The GTX 750 uses GDDR5, which is a high-bandwidth memory type suited for frame-heavy workloads like gaming. The Quadro K620 uses DDR3, which is lower-bandwidth but allows for a larger 2 GB capacity at what would have been a lower cost. The 80.19 GB/s versus 28.80 GB/s bandwidth difference is a factor of 2.8, making the GTX 750 far better suited for bandwidth-intensive tasks. The Quadro K620’s larger capacity suggests NVIDIA prioritized memory footprint over speed for professional workloads that need to hold larger datasets.

Clock behavior also differs. The Quadro K620 runs higher base and boost clocks (1,058 MHz / 1,124 MHz) than the GTX 750 (1,020 MHz / 1,085 MHz), which partially compensates for its fewer shading units. However, the FP32 throughput numbers show the GTX 750 still leads at 1,111.0 GFLOPS versus 863.2 GFLOPS, a 28.7% advantage.

The PCIe interface differs: the GTX 750 uses PCIe 3.0 x16, while the Quadro K620 uses PCIe 2.0 x16. This affects host-GPU data transfer speeds, with PCIe 3.0 offering double the bandwidth of PCIe 2.0 per lane. For compute workloads that involve frequent data transfers between CPU and GPU, this could matter.

Power delivery is another differentiator. The GTX 750 has a 55 W TDP and recommends a 250 W PSU, while the Quadro K620 has a 45 W TDP and recommends a 200 W PSU. The Quadro is more power-efficient in terms of raw TDP, but the GTX 750 delivers far more performance per watt when considering the benchmark scores. Neither card requires external power connectors, making both suitable for simple system installations.

The display output configuration reflects their intended markets. The GTX 750 offers dual DVI and mini-HDMI, typical for consumer monitors and TVs. The Quadro K620 offers DVI and DisplayPort 1.2, which is standard in professional workstations and enterprise display setups. The Quadro’s generation label, "Quadro Kepler (Kx200)", is something of a misnomer given the Maxwell architecture, but it indicates NVIDIA’s naming scheme at the time.

The Verdict

The data is unambiguous: the NVIDIA GeForce GTX 750 outperforms the NVIDIA Quadro K620 in every recorded benchmark. In OpenCL, it leads by 39.2%, and in Vulkan, by 37.6%. Its average benchmark score of 7,222 sits in the 40th percentile of all GPUs, while the Quadro K620’s 6,282 sits in the 36th. For any workload measured by these tests, the GTX 750 is the clear choice.

The GTX 750 wins because of its compute architecture. With 512 shading units, 32 texture units, and 1,111.0 GFLOPS of FP32 performance, it has a structural advantage over the Quadro K620’s 384 shading units, 24 texture units, and 863.2 GFLOPS. The GDDR5 memory at 80.19 GB/s provides bandwidth that the Quadro K620’s DDR3 at 28.80 GB/s cannot match. These are not marginal differences; they are fundamental architectural choices that favor the consumer card.

Where the Quadro K620 makes its case is in specific professional use cases. Its 2 GB memory capacity doubles the GTX 750’s 1 GB, which could be decisive for workloads that require larger frame buffers or datasets. The DisplayPort 1.2 output is preferable for professional display ecosystems. Its lower 45 W TDP and 200 W recommended PSU make it easier to integrate into power-constrained systems. But none of these advantages show up in the benchmark scores.

The verdict is straightforward: for raw compute performance, the GTX 750 is the superior card. For professional environments where memory capacity, display connectivity, and lower power draw matter more than raw throughput, the Quadro K620 has a place. The benchmark data does not support choosing the Quadro K620 for performance, but the specification differences suggest it was designed for a different set of priorities. Buyers who need the GTX 750’s performance should choose it without hesitation. Buyers who need the Quadro K620’s specific professional features should understand they are trading away 37% or more of compute performance in exchange.