Intel UHD Graphics P750 vs NVIDIA GeForce GTX 675M Comparison

The GeForce GTX 675M and Intel UHD Graphics P750 sit in nearly the same performance tier, separated by a single benchmark result that shows the NVIDIA part holding a 6% lead. Both are end-of-life products, but they come from opposite ends of the hardware spectrum: a 2012 discrete mobile GPU versus an integrated graphics processor built into a Rocket Lake desktop chip. The data reveals a tighter race than the architectural gulf suggests.

Head-to-Head Benchmarks

The only direct comparison available is the Geekbench OpenCL test, and it favors the NVIDIA GeForce GTX 675M. The GTX 675M scores 6,946, while the Intel UHD Graphics P750 scores 6,554. That works out to a 6% delta, with the NVIDIA card winning the only head-to-head matchup. In percentile terms, the GTX 675M sits at the 39th percentile of all GPUs, while the P750 is one notch lower at the 38th percentile — a negligible gap in the broader landscape.

Looking at nearest rivals, the GTX 675M is bracketed by a cluster of similarly performing parts. The AMD Radeon R5 M240 scores 6,975, which is 0.4% higher than the GTX 675M. The NVIDIA GeForce GTX 680M is 1.1% ahead at 7,023, and the NVIDIA T600 is 1.3% ahead at 7,035. On the other side, the AMD FirePro M5100 trails by 1.7% at 6,830. This puts the GTX 675M squarely in a dense pack where a few points either way determine rank.

The Intel UHD Graphics P750's nearest rivals tell a similar story. The AMD Radeon HD 7730M scores 6,581, which is 0.4% ahead. The AMD Radeon R7 M460 is 0.9% ahead at 6,612. Meanwhile, the NVIDIA GeForce GTX 670M trails by 0.6% at 6,513, and the NVIDIA GeForce GT 555M trails by 0.9% at 6,493. The P750's score of 6,554 places it in the middle of that group, with roughly 1% separating the closest competitors on either side.

The 6% delta between the two featured GPUs is meaningful but not transformative. In practical terms, a 6% performance advantage in OpenCL compute translates to a noticeable but not dramatic edge in workloads that scale with raw throughput. The GTX 675M's win is consistent with its higher FP32 rating, but the P750's higher pixel and texture rates suggest it is not uniformly inferior.

Architecture Differences

The two GPUs could hardly be more different in construction. The GTX 675M uses the GF114 chip built on NVIDIA's Fermi 2.0 architecture, manufactured on a 40 nm process at TSMC. It packs 1,950 million transistors into a 332 mm² die, yielding a transistor density of 5.9 million per square millimeter. The Intel UHD Graphics P750, by contrast, uses the Rocket Lake chip with Intel's Generation 12.1 architecture, built on a 14 nm+++ process at Intel. Its transistor count and die size are not listed, but the process node alone signals a much smaller physical footprint for the integrated solution.

The compute resources differ significantly. The GTX 675M has 384 shading units, 64 texture mapping units, and 32 raster operation units. The P750 has fewer shading units at 256, but matches the GTX 675M with 64 TMUs and 32 ROPs. This creates an interesting split. The GTX 675M's higher shading unit count gives it a theoretical FP32 throughput of 952.3 GFLOPS, while the P750's 256 units produce 665.6 GFLOPS. However, the P750 compensates with a much higher pixel rate of 41.60 GPixel/s versus 9.920 GPixel/s for the GTX 675M, and a texture rate of 83.20 GTexel/s versus 39.68 GTexel/s.

The clock behavior also diverges. The GTX 675M has no listed base or boost clock, only a memory clock of 750 MHz running at 3 Gbps effective. The P750 lists a base clock of 350 MHz and a boost clock of 1300 MHz. Those clocks, combined with the higher pixel and texture rates, explain why the P750 can compete despite having fewer shading units — it is running much faster per unit of work.

Memory is another fundamental divide. The GTX 675M uses 2 GB of GDDR5 on a 256-bit bus, delivering 96.00 GB/s of bandwidth. The P750 uses system shared memory, with bandwidth described as "System Dependent." This is a critical distinction for workloads that are memory-bound, as the GTX 675M's dedicated GDDR5 provides predictable bandwidth, while the P750's performance depends entirely on the host system's memory configuration.

The APIs tell a different story. The GTX 675M supports DirectX 12 (11_0) and OpenGL 4.6, but has no listed Vulkan support. The P750 supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The P750's higher DirectX feature level and Vulkan support give it a modern software advantage, even though it is the slower part in the one benchmark measured.

Power consumption is starkly different. The GTX 675M has a TDP of 100 W and uses an MXM module slot with no power connectors. The P750 has a TDP of just 15 W and is an integrated graphics processor on a Ring Bus interface. The P750 draws 85 W less, which is a massive difference in thermal and power budgets.

FAQ

Q: Which GPU wins the Geekbench OpenCL benchmark?

A: The NVIDIA GeForce GTX 675M wins with a score of 6,946 versus 6,554 for the Intel UHD Graphics P750, a 6% difference.

Q: How does the GTX 675M compare to its closest rival, the AMD Radeon R5 M240?

A: The Radeon R5 M240 scores 6,975, which is 0.4% higher than the GTX 675M's 6,946. The two are essentially tied.

Q: What is the most significant architectural difference between the two GPUs?

A: The GTX 675M is a discrete 40 nm Fermi 2.0 part with 384 shading units and dedicated GDDR5 memory, while the P750 is an integrated 14 nm+++ Generation 12.1 part with 256 shading units and system shared memory.

Q: Does the Intel P750 have any advantages over the GTX 675M?

A: Yes. The P750 has a much higher pixel rate of 41.60 GPixel/s versus 9.920 GPixel/s, a texture rate of 83.20 GTexel/s versus 39.68 GTexel/s, and supports Vulkan 1.4, which the GTX 675M does not list.

Q: How do the TDPs compare?

A: The GTX 675M has a TDP of 100 W, while the P750 has a TDP of 15 W. The P750 draws 85 W less power.

Q: What is the percentile ranking for each GPU?

A: The GTX 675M is at the 39th percentile of all GPUs, and the P750 is at the 38th percentile.

The Verdict

The data points to a narrow win for the NVIDIA GeForce GTX 675M in raw compute, with a 6% lead in the only benchmark available. However, the Intel UHD Graphics P750 is not far behind, sitting just one percentile lower and matching the GTX 675M in TMU and ROP counts. The GTX 675M's advantage comes from its higher shading unit count and dedicated GDDR5 memory, which give it a 286.7 GFLOPS edge in FP32 throughput and a 96.00 GB/s bandwidth advantage over system shared memory.

The P750 counters with dramatically higher pixel and texture rates. Its 41.60 GPixel/s is more than four times the GTX 675M's 9.920 GPixel/s, and its 83.20 GTexel/s is more than double the GTX 675M's 39.68 GTexel/s. These figures suggest the P750 is better suited to fill-rate-bound workloads, even if its overall compute score is lower.

For a user choosing between these two, the decision hinges on context. The GTX 675M is a discrete MXM module that requires a compatible laptop or mobile workstation, draws 100 W, and uses GDDR5 memory. The P750 is an integrated GPU that comes with a Rocket Lake processor, draws 15 W, and uses system memory. The GTX 675M is the stronger compute part, but the P750 offers modern API support including Vulkan 1.4 and DirectX 12 (12_1), which the GTX 675M cannot match with its DirectX 12 (11_0) and no Vulkan.

Specification Differences

The two GPUs differ in nearly every measurable specification. The GTX 675M uses a GF114 chip on a 40 nm process at TSMC, with 1,950 million transistors on a 332 mm² die. The P750 uses a Rocket Lake chip on a 14 nm+++ process at Intel, with no listed transistor count or die size. The GTX 675M has 384 shading units, 64 TMUs, and 32 ROPs. The P750 has 256 shading units, 64 TMUs, and 32 ROPs.

Clock speeds differ sharply. The GTX 675M has no base or boost clock listed, only a memory clock of 750 MHz running at 3 Gbps effective. The P750 has a base clock of 350 MHz and a boost clock of 1300 MHz. Memory configuration is also divergent: the GTX 675M uses 2 GB of GDDR5 on a 256-bit bus with 96.00 GB/s bandwidth, while the P750 uses system shared memory with system dependent bandwidth.

Performance rates show the GTX 675M ahead in FP32 at 952.3 GFLOPS versus 665.6 GFLOPS for the P750. The P750 is ahead in pixel rate at 41.60 GPixel/s versus 9.920 GPixel/s, and in texture rate at 83.20 GTexel/s versus 39.68 GTexel/s. The P750 also supports FP16 at 1,331.2 GFLOPS (2:1), while the GTX 675M has no listed FP16 support. TDP is 100 W for the GTX 675M and 15 W for the P750. The GTX 675M uses an MXM-B (3.0) bus interface, while the P750 uses a Ring Bus. Display outputs are portable device dependent for the GTX 675M and motherboard dependent for the P750.

Where Each One Wins

The NVIDIA GeForce GTX 675M wins in compute-heavy workloads that benefit from higher shading unit counts and dedicated memory bandwidth. Its 384 shading units and 952.3 GFLOPS of FP32 throughput give it a clear edge in general-purpose compute tasks, as reflected in its 6% Geekbench OpenCL victory. The 2 GB of GDDR5 on a 256-bit bus provides 96.00 GB/s of bandwidth, which is a fixed resource that does not depend on the host system. The GTX 675M also has a higher transistor density at 5.9 million per square millimeter, though that is a manufacturing metric rather than a performance one.

The Intel UHD Graphics P750 wins in fill-rate-bound scenarios and efficiency-driven applications. Its 41.60 GPixel/s pixel rate is over four times higher than the GTX 675M's, and its 83.20 GTexel/s texture rate is more than double. These figures suggest the P750 handles tasks involving high-resolution textures and pixel output more effectively, despite its lower overall compute score. The P750 also supports Vulkan 1.4 and DirectX 12 (12_1), giving it a software compatibility edge that the GTX 675M cannot match. Its 15 W TDP makes it suitable for systems where power consumption is a primary constraint, whereas the GTX 675M's 100 W TDP demands a more robust cooling and power delivery solution.

The percentile data reinforces the split. Both GPUs sit within one percentile of each other, at 39 and 38 respectively, meaning neither holds a dominant position in the overall GPU landscape. The GTX 675M's nearest rivals are all within 1.7% of its score, and the P750's nearest rivals are within 0.9%. This is a tightly contested matchup where the winner depends entirely on the specific workload and system context.