Intel UHD Graphics P750 vs NVIDIA GeForce GTX 650 Ti Boost Comparison

Head-to-Head Benchmarks

The database contains a single direct comparison between these two graphics solutions: the Geekbench OpenCL test. In this measurement, the NVIDIA GeForce GTX 650 Ti Boost scores 9,302 points, while the Intel UHD Graphics P750 scores 6,554 points. The delta is a substantial 41.9 percent in favor of the NVIDIA card. This is not a marginal win; it is a clear, decisive margin that separates the two parts by a wide gulf in raw computational throughput.

To put the NVIDIA score into context, it sits in the 42nd percentile of all GPUs in the database. The Intel UHD Graphics P750 lands at the 38th percentile. While both are mid-range performers in the grand scheme of all recorded GPUs, the NVIDIA part is meaningfully closer to the median of the entire database. The average benchmark score for the NVIDIA card is 8,067, while the Intel part averages 6,554. This means the NVIDIA card carries an average score that is over 23 percent higher than the Intel part when all recorded tests (not just the head-to-head) are considered.

The nearest rivals in the database further clarify the NVIDIA’s position. The NVIDIA GRID K2 averages 8,080 points, a delta of -0.2 percent relative to the GTX 650 Ti Boost, which places the two essentially at parity. The GeForce GTX 650 Ti (non-Boost) sits at 8,053, a delta of 0.2 percent, so the Boost variant’s extra performance is minimal when averaged over all tests. The GeForce GTX 880M averages 8,040 (delta 0.3 percent) and the Quadro P5000 averages 8,039 (delta 0.3 percent). This cluster of scores around the 8,000 mark indicates that the GTX 650 Ti Boost is firmly in a performance tier where many competing NVIDIA parts land within a fraction of a percent of each other.

For the Intel side, the nearest rivals reveal a different pattern. The AMD Radeon HD 7730M averages 6,581, which is 0.4 percent above the Intel part. The NVIDIA GeForce GTX 670M averages 6,513, a delta of 0.6 percent below the Intel. The AMD Radeon R7 M460 averages 6,612, putting it 0.9 percent above the Intel. The NVIDIA GeForce GT 555M averages 6,493, which is 0.9 percent below. So the Intel UHD Graphics P750 lands in a tight cluster of mobile and low-power GPUs, all within about one percent of each other. This suggests the Intel part is competitive within its own segment but that segment is a full step below the discrete desktop GPU tier where the GTX 650 Ti Boost resides.

The only shared test, OpenCL, is the one where the NVIDIA demonstrates a 41.9 percent advantage. That is a substantial gap for any benchmark, and it indicates that in compute workloads, the NVIDIA card has a definitive edge. The Intel part, however, has no Vulkan or Metal benchmark results recorded; only OpenCL is available, so the database cannot show how these two would compare in those other APIs.

FAQ

Q: Which GPU is faster in the OpenCL benchmark?

A: The NVIDIA GeForce GTX 650 Ti Boost scores 9,302 points in OpenCL, while the Intel UHD Graphics P750 scores 6,554 points. That is a 41.9 percent difference in favor of the NVIDIA card.

Q: How do these two GPUs rank against all other GPUs in the database?

A: The NVIDIA GeForce GTX 650 Ti Boost sits at the 42nd percentile, meaning about 42 percent of all GPUs score lower. The Intel UHD Graphics P750 sits at the 58th percentile, meaning it outperforms roughly 58 percent of all GPUs. Despite the higher percentile for Intel, the actual average score is lower because the percentile distribution is based on the entire population of GPUs, which includes many low-power integrated and entry-level parts.

Q: What is the average benchmark score for each GPU across all tests?

A: The NVIDIA GeForce GTX 650 Ti Boost has an average benchmark score of 8,067 points. The Intel UHD Graphics P750 has an average of 6,554 points. The NVIDIA’s average is about 23 percent higher.

Q: Are these GPUs close to any specific rival in performance?

A: Yes. For the NVIDIA, the closest rival is the NVIDIA GRID K2, which averages 8,080 points, a delta of only 0.2 percent from the GTX 650 Ti Boost. For the Intel, the closest rival is the AMD Radeon HD 7730M, averaging 6,581 points, a delta of 0.4 percent from the UHD P750.

Q: Which GPU has a higher pixel rate and texture rate?

A: The Intel UHD Graphics P750 has a higher pixel rate at 41.60 GPixel/s compared to the NVIDIA’s 16.51 GPixel/s. The Intel also has a higher texture rate of 83.20 GTexel/s versus the NVIDIA’s 66.05 GTexel/s. This suggests the Intel may excel in fill-rate-bound tasks despite losing in compute.

Q: Does the Intel GPU have any advantage in FP16 performance?

A: Yes, the Intel UHD Graphics P750 has a FP16 rating of 1,331.2 GFLOPS, while the NVIDIA GTX 650 Ti Boost has no listed FP16 rating. However, the NVIDIA’s FP32 rating is 1,585 TFLOPS, which is higher than the Intel’s FP32 of 665.6 GFLOPS.

Architecture Differences

The NVIDIA GeForce GTX 650 Ti Boost is built on the Kepler architecture, specifically using the GK106 chip. This is a discrete GPU manufactured by TSMC on a 28 nm process. It contains 2,540 million transistors on a die size of 221 mm², giving a transistor density of 11.5 million per mm². The Intel UHD Graphics P750 uses a completely different approach: it is an integrated graphics processor (IGP) within the Rocket Lake chip, using Intel’s 14 nm+++ process. The Intel chip has no separate transistor count or die size listed in the database, because it is part of a larger CPU package.

The NVIDIA GPU is a dedicated graphics card with its own VRAM. It has 2 GB of GDDR5 memory on a 192-bit bus, yielding a bandwidth of 144.2 GB/s. The Intel UHD Graphics P750 shares system memory, using a "System Shared" memory type and bus width. Its bandwidth is described as "System Dependent," meaning it varies based on the host system’s RAM configuration. This is a fundamental difference: one uses fixed, dedicated memory with a known bandwidth, and the other depends on the PC’s main memory.

The number of execution units also differs significantly. The NVIDIA has 768 shading units and 64 texture mapping units, with 24 render output units. The Intel has 256 shading units, also 64 texture mapping units, and 32 render output units. The NVIDIA has three times the shading units of the Intel, but the Intel has more render output units (32 versus 24) and the same number of texture mapping units.

Architecturally, the NVIDIA supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The Intel supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The Intel therefore supports a higher feature level of DirectX 12 (12_1 versus 11_0) and a newer Vulkan version (1.4 versus 1.2.175). This is notable despite the lower raw compute performance.

The NVIDIA is a dual-slot card requiring a 6-pin power connector and a 300 W power supply. The Intel is an IGP with no power connectors. The NVIDIA has a TDP of 134 W, while the Intel has a TDP of only 15 W, a massive difference in power consumption. The NVIDIA is also a discrete card measuring 241 mm in length, while the Intel has no physical dimensions because it soldered into the motherboard.

Specification Differences

The two components differ in nearly every specification field recorded in the database. The NVIDIA is a discrete GPU with a 28 nm process node, while the Intel uses a 14 nm+++ process node. The NVIDIA has a base clock of 980 MHz and a boost clock of 1,032 MHz, while the Intel has a base clock of 350 MHz and a boost clock of 1,300 MHz. So the Intel has a higher boost clock but a lower base clock.

Memory is a major differentiator: the NVIDIA has 2 GB of GDDR5 with a 192-bit bus and 144.2 GB/s bandwidth. The Intel uses system shared memory, with no fixed size, type, or bandwidth. The NVIDIA’s memory clock is 1,502 MHz with an effective rate of 6 Gbps, while the Intel’s memory clock is not listed.

The shading units count is 768 for NVIDIA versus 256 for Intel. Texture mapping units are equal at 64 for both, but render output units are 24 for NVIDIA and 32 for Intel. The pixel rate is 16.51 GP/s for NVIDIA versus 41.60 GP/s for Intel. The texture rate is 66.05 GT/s versus 83.20 GT/s. The FP32 performance is 1.585 TFLOPS for NVIDIA versus 665.6 GFLOPS for Intel. The Intel also has an FP16 rating of 1,331.2 GFLOPS, while the NVIDIA has none listed.

The NVIDIA requires a 134 W power draw, a dual-slot cooling solution, a 6-pin power connector, and a 300 W power supply. The Intel is an IGP with a 15 W TDP, no power connectors, and no recommended power supply. The NVIDIA uses a PCIe 3.0 x16 interface, while the Intel uses a Ring Bus. The NVIDIA has specific display outputs: 2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2, while the Intel’s outputs are motherboard-dependent.

The NVIDIA supports DirectX 12 (11_0), while the Intel supports DirectX 12 (12_1). Both support OpenGL 4.6, but the Vulkan versions differ: NVIDIA at 1.2.175 versus Intel at 1.4. The NVIDIA has a release date of March 25, 2013, while the Intel has no release date recorded. The NVIDIA has a launch MSRP of 169 USD, but the Intel has no MSRP because it is integrated into a larger processor package.

The Verdict

The database clearly shows that the NVIDIA GeForce GTX 650 Ti Boost is the superior performer in compute benchmarks. Its OpenCL score is 41.9 percent higher than the Intel UHD Graphics P750, and its average benchmark score across all tests is 8,067 versus 6,554. The NVIDIA also has a significantly higher FP32 performance (1.585 TFLOPS versus 665.6 GFLOPS) and more shading units (768 versus 256). For any task that relies on raw parallel computation, the NVIDIA wins.

However, the Intel UHD Graphics P750 is not without strengths. It has a higher pixel rate (41.60 GPixel/s versus 16.51), a higher texture rate (83.20 GTexel/s versus 66.05), and more render output units (32 versus 24). It also supports DirectX 12 (12_1) and Vulkan 1.4, both of which are newer than what the NVIDIA offers. This could be relevant for certain modern graphics workloads that leverage these newer APIs.

But the Intel is an integrated GPU with a TDP of only 15 W, while the NVIDIA is a discrete card with a 134 W TDP. The Intel shares system memory, while the NVIDIA uses dedicated GDDR5. The Intel is a part of a larger processor, so it does not require any additional hardware, while the NVIDIA requires a power connector and a 300 W power supply.

Thus, a user who needs maximum compute performance in an existing desktop system should choose the NVIDIA GeForce GTX 650 Ti Boost. A system builder who prioritizes the lowest power consumption and integration without a separate graphics card would choose the Intel UHD Graphics P750, but only if they accept the lower compute performance.

Where Each One Wins

NVIDIA GeForce GTX 650 Ti Boost Wins:

• In compute benchmarks: OpenCL score is 41.9 percent higher than the Intel.
• In average benchmark score: 8,067 versus 6,554.
• In FP32 throughput: 1.585 TFLOPS vs 665.6 GFLOPS.
• In shading units: 768 vs 256.
• In dedicated memory bandwidth: 144.2 GB/s vs system-dependent.
• In the only direct head-to-head benchmark, it wins all 1 recorded test.

Intel UHD Graphics P750 Wins:

• In pixel fill rate: 41.60 GPixel/s vs 16.51.
• In texture fill rate: 83.20 GTexel/s vs 66.05.
• In render output units: 32 vs 24.
• In DirectX feature level: 12_1 vs 11_0.
• In Vulkan version: 1.4 vs 1.2.175.
• In power efficiency: 15 W vs 134 W.
• In FP16 performance: 1,331.2 GFLOPS vs none listed.

The NVIDIA is the clear choice for compute-heavy tasks such as video encoding, rendering, or any workload that uses the OpenCL or Vulkan compute APIs. The Intel is better suited for display-only or fill-rate-dependent scenarios, but those scenarios are not covered by the recorded compute benchmarks. In the database’s single head-to-head test, the NVIDIA wins decisively. The Intel’s wins are in architectural features that may not translate to higher frame rates in typical gaming because the compute scores indicate a large deficiency. For a user who must choose based on the recorded data alone, the NVIDIA GeForce GTX 650 Ti Boost is the more powerful graphics solution, while the Intel UHD Graphics P750 offers a more modern feature set and radically lower power draw.