NVIDIA GeForce GTX 650 Ti vs NVIDIA GeForce GTX 750 Comparison

Head-to-Head Benchmarks

The recorded data shows a split decision between these two NVIDIA cards. The NVIDIA GeForce GTX 750 takes the OpenCL test, while the NVIDIA GeForce GTX 650 Ti takes the Vulkan test. Each card wins one benchmark, and the margins are asymmetric.

In Geekbench OpenCL, the GTX 750 scores 9315 against the GTX 650 Ti's 7877. That is a 15.4% advantage for the GTX 750, a decisive margin in a compute-oriented workload. The GTX 650 Ti trails by a wide gap here, and the delta is the largest in either direction across both tests.

In Geekbench Vulkan, the GTX 650 Ti scores 8229 against the GTX 750's 8078. The margin is just 1.9%, a much closer result. The GTX 650 Ti takes the win, but the advantage is narrow enough that run-to-run variance could matter. Benchmark results indicate the GTX 650 Ti holds a slight edge in this API, while the GTX 750 dominates the OpenCL test.

The average benchmark score tells a mixed story. The GTX 650 Ti averages 8053 across its two recorded tests, while the GTX 750 averages 7222 across three tests. However, the GTX 750's average includes a Geekbench Metal score of 4274, which is substantially lower than either card's OpenCL or Vulkan scores. That Metal result drags down the GTX 750's average. When comparing only the overlapping tests, the GTX 750 is clearly ahead in OpenCL and marginally behind in Vulkan. The GTX 650 Ti's percentile ranking among all GPUs is 42, while the GTX 750 sits at 40. Those are close percentile positions, consistent with the head-to-head split.

For context against nearby rivals, the GTX 650 Ti's nearest competitor is the NVIDIA GeForce GTX 650 Ti Boost, which scores 8067, a delta of -0.2% against the GTX 650 Ti. The GTX 880M scores 8040, a 0.2% delta. The GTX 750's nearest rival is the AMD Radeon Vega 8 Mobile at 7203, a 0.3% delta. These tight clusters reinforce that both cards sit in a similar performance band, with the GTX 750's OpenCL strength being its standout feature.

Architecture Differences

The two cards come from different NVIDIA generations and architectures. The GTX 650 Ti uses the GK106S chip based on Kepler architecture, part of the GeForce 600 generation. The GTX 750 uses the GM107 chip based on Maxwell architecture, part of the GeForce 700 generation. Both are built on a 28 nm process at TSMC, but the chip designs diverge significantly.

The GTX 650 Ti packs 2,540 million transistors on a 221 mm² die, giving a transistor density of 11.5M per mm². The GTX 750 packs 1,870 million transistors on a 148 mm² die, giving a higher density of 12.6M per mm². The GTX 750 is a smaller, denser chip, reflecting the architectural efficiency gains of Maxwell over Kepler.

Compute resources differ notably. The GTX 650 Ti has 768 shading units, 64 texture mapping units, and 16 ROPs. The GTX 750 has 512 shading units, 32 TMUs, and 16 ROPs. The GTX 650 Ti has more shading units and double the TMUs, yet the GTX 750 still wins the OpenCL benchmark. This indicates Maxwell's architecture extracts more performance per shader and per texture unit than Kepler.

Clock behavior also differs. The GTX 650 Ti has no recorded base or boost clock in the database, only a memory clock of 1350 MHz, or 5.4 Gbps effective. The GTX 750 has a base clock of 1020 MHz and a boost clock of 1085 MHz, with a memory clock of 1253 MHz, or 5 Gbps effective. The GTX 750's explicit boost clock is an architectural feature that allows it to sustain higher frequencies under load.

Memory configurations are similar in capacity and bus width. Both cards have 1024 MB of GDDR5 on a 128-bit bus. The GTX 650 Ti has higher bandwidth at 86.40 GB/s, while the GTX 750 has 80.19 GB/s. The GTX 650 Ti's wider bandwidth comes from its higher memory clock. Pixel rates favor the GTX 750 at 17.36 GPixel/s versus 14.85 GPixel/s for the GTX 650 Ti, despite the latter's higher bandwidth. Texture rates favor the GTX 650 Ti at 59.39 GTexel/s versus 34.72 GTexel/s, reflecting its double TMU count. FP32 performance favors the GTX 650 Ti at 1,425.4 GFLOPS versus 1,111.0 GFLOPS for the GTX 750.

Power consumption is a major split. The GTX 650 Ti has a TDP of 110 W and requires a 1x 6-pin power connector. The GTX 750 has a TDP of 55 W and requires no power connector at all. The suggested PSU drops from 300 W on the GTX 650 Ti to 250 W on the GTX 750. This is a halving of power draw while maintaining competitive benchmark scores.

API support is nearly identical. Both cards support DirectX 12 (11_0) and OpenGL 4.6. The Vulkan version differs: the GTX 650 Ti supports Vulkan 1.2.175, while the GTX 750 supports Vulkan 1.4. The GTX 750's newer Vulkan support is a feature-level advantage.

Where Each One Wins

The GTX 750 wins in compute workloads that favor OpenCL. Its Geekbench OpenCL score of 9315 is 15.4% higher than the GTX 650 Ti's 7877. This makes the GTX 750 the better choice for general-purpose compute tasks that use OpenCL, such as certain rendering, simulation, or data-processing applications. The GTX 750 also wins on power efficiency. Its 55 W TDP is half the GTX 650 Ti's 110 W, and it needs no auxiliary power connector. For builds with limited power delivery or smaller power supplies, the GTX 750 is the safer pick. Its smaller die at 148 mm² and lower transistor count of 1,870 million also indicate a more efficient design.

The GTX 650 Ti wins in Vulkan-based workloads. Its Geekbench Vulkan score of 8229 edges out the GTX 750's 8078 by 1.9%. For applications that use Vulkan, the GTX 650 Ti has a slight performance edge. The GTX 650 Ti also has higher raw compute throughput on paper: 1,425.4 GFLOPS FP32 versus 1,111.0 GFLOPS, and a higher texture rate of 59.39 GTexel/s versus 34.72 GTexel/s. Memory bandwidth also favors the GTX 650 Ti at 86.40 GB/s versus 80.19 GB/s. For texture-heavy or bandwidth-sensitive workloads that are not OpenCL-bound, the GTX 650 Ti's specifications suggest an advantage.

The GTX 750's higher pixel rate of 17.36 GPixel/s versus 14.85 GPixel/s suggests it handles pixel-fill operations more efficiently, which could benefit certain rasterization-heavy scenarios. The GTX 650 Ti's higher shading unit count of 768 versus 512 gives it more parallel execution units, but the GTX 750's architecture compensates in practice.

FAQ

Q: Which card has the higher average benchmark score?

A: The GTX 650 Ti has an average benchmark score of 8053 across two tests, while the GTX 750 has 7222 across three tests. However, the GTX 750's average includes a Metal score of 4274, which is much lower than its OpenCL and Vulkan scores.

Q: How do the two cards compare in OpenCL performance?

A: The GTX 750 scores 9315 in Geekbench OpenCL, which is 15.4% ahead of the GTX 650 Ti's 7877. This is the largest performance gap between the two cards in any recorded test.

Q: Does the GTX 750 require a power connector?

A: No. The GTX 750 has no power connectors and a 55 W TDP. The GTX 650 Ti, by contrast, has a 110 W TDP and requires a single 6-pin power connector.

Q: Which card has better Vulkan support?

A: The GTX 750 supports Vulkan 1.4, while the GTX 650 Ti supports Vulkan 1.2.175. Despite this newer API version, the GTX 650 Ti scores higher in the Vulkan benchmark at 8229 versus 8078.

Q: What are the memory specifications of each card?

A: Both cards have 1024 MB of GDDR5 memory on a 128-bit bus. The GTX 650 Ti has a memory clock of 1350 MHz with 5.4 Gbps effective speed and 86.40 GB/s bandwidth. The GTX 750 has a memory clock of 1253 MHz with 5 Gbps effective speed and 80.19 GB/s bandwidth.

Q: How does the GTX 650 Ti compare to its nearest rival, the GTX 650 Ti Boost?

A: The GTX 650 Ti Boost scores 8067 on average, a delta of -0.2% against the GTX 650 Ti's 8053. The two cards are nearly identical in average performance.

Specification Differences

The database records the following differences between the two cards:

• Architecture: GTX 650 Ti uses Kepler, GTX 750 uses Maxwell
• Generation: GTX 650 Ti is GeForce 600, GTX 750 is GeForce 700
• Chip: GTX 650 Ti uses GK106S, GTX 750 uses GM107
• Transistors: GTX 650 Ti has 2,540 million, GTX 750 has 1,870 million
• Die size: GTX 650 Ti is 221 mm², GTX 750 is 148 mm²
• Transistor density: GTX 650 Ti is 11.5M per mm², GTX 750 is 12.6M per mm²
• Base clock: GTX 650 Ti has none recorded, GTX 750 has 1020 MHz
• Boost clock: GTX 650 Ti has none recorded, GTX 750 has 1085 MHz
• Memory clock: GTX 650 Ti is 1350 MHz (5.4 Gbps effective), GTX 750 is 1253 MHz (5 Gbps effective)
• Bandwidth: GTX 650 Ti is 86.40 GB/s, GTX 750 is 80.19 GB/s
• Shading units: GTX 650 Ti has 768, GTX 750 has 512
• TMUs: GTX 650 Ti has 64, GTX 750 has 32
• Pixel rate: GTX 650 Ti is 14.85 GPixel/s, GTX 750 is 17.36 GPixel/s
• Texture rate: GTX 650 Ti is 59.39 GTexel/s, GTX 750 is 34.72 GTexel/s
• FP32: GTX 650 Ti is 1,425.4 GFLOPS, GTX 750 is 1,111.0 GFLOPS
• TDP: GTX 650 Ti is 110 W, GTX 750 is 55 W
• Power connectors: GTX 650 Ti has 1x 6-pin, GTX 750 has none
• Suggested PSU: GTX 650 Ti is 300 W, GTX 750 is 250 W
• Vulkan version: GTX 650 Ti is 1.2.175, GTX 750 is 1.4
• Release date: GTX 650 Ti is 2012-10-08, GTX 750 is 2014-02-17
• Predecessor: GTX 650 Ti is GeForce 500, GTX 750 is GeForce 600
• Successor: GTX 650 Ti is GeForce 700, GTX 750 is GeForce 900
• Launch MSRP: GTX 650 Ti is 149 USD, GTX 750 is 119 USD
• OpenCL benchmark: GTX 650 Ti is 7877, GTX 750 is 9315
• Vulkan benchmark: GTX 650 Ti is 8229, GTX 750 is 8078
• Average benchmark score: GTX 650 Ti is 8053, GTX 750 is 7222
• Percentile: GTX 650 Ti is 42, GTX 750 is 40

Identical specifications include 1024 MB GDDR5 memory, 128-bit bus, 16 ROPs, PCIe 3.0 x16 interface, single-slot width, 145 mm length, 2x DVI and 1x mini-HDMI 1.4a outputs, DirectX 12 (11_0), OpenGL 4.6, 28 nm TSMC process, and end-of-life production status.

The Verdict

The data supports a clear split based on workload and system constraints. For compute-heavy tasks using OpenCL, the GTX 750 is the definitive pick. Its 9315 OpenCL score beats the GTX 650 Ti by 15.4%, and it does so while drawing 55 W versus 110 W, with no power connector required. The GTX 750's efficiency advantage is substantial, and its newer Vulkan 1.4 support adds a feature-level edge.

For Vulkan-based workloads, the GTX 650 Ti wins narrowly. Its 8229 Vulkan score is 1.9% ahead of the GTX 750's 8078. The GTX 650 Ti also offers higher raw specifications: more shading units, double the TMUs, higher FP32 throughput, higher texture rate, and higher memory bandwidth. If an application is sensitive to those factors rather than OpenCL, the GTX 650 Ti has the theoretical advantage.

The GTX 650 Ti's launch MSRP was 149 USD, while the GTX 750 launched at 119 USD. The GTX 750 is the newer card, released in 2014 versus 2012 for the GTX 650 Ti, and it belongs to a later generation with a different architecture. The GTX 750's Maxwell architecture delivers better OpenCL performance with half the power draw and a smaller die.

Users building or upgrading a system with a limited power supply, or those who prefer not to use a power connector, should choose the GTX 750. Users who prioritize Vulkan performance or who need the higher texture rate and shading unit count of the GTX 650 Ti should choose that card. The benchmark data shows one win each, so the decision comes down to which API and workload matters more, plus the power envelope. The GTX 750 is the more efficient and modern option, while the GTX 650 Ti retains a narrow Vulkan edge and superior raw compute specifications.