GPU Comparison

Intel HD Graphics 530 and NVIDIA GeForce GT 640 are both end-of-life graphics solutions from different eras, targeting different segments of the market. The Intel part is a 2015 integrated graphics processor built on a 14 nm+ process, while the NVIDIA GeForce GT 640 is a 2012 discrete graphics card built on a 28 nm process. Benchmark results from Geekbench show a split decision: the Intel HD Graphics 530 wins one test decisively, while the NVIDIA GeForce GT 640 takes the other two, with the overall average benchmark score slightly favoring the Intel part at 3332 versus 3210 for the NVIDIA.

Head-to-Head Benchmarks

The most striking result in the head-to-head comparison is in the Geekbench Metal test, where the Intel HD Graphics 530 achieves a score of 5025 against the NVIDIA GeForce GT 640's 2086. This represents a 140.9% advantage for the Intel part, a massive margin that suggests the Intel integrated GPU is significantly more efficient in this specific API workload. The NVIDIA GeForce GT 640, despite being a discrete card with dedicated memory, trails far behind in this test, which may reflect architectural differences in how each handles Metal compute tasks or driver optimization levels.

In contrast, the Geekbench OpenCL test shows a much closer contest. The NVIDIA GeForce GT 640 scores 3736, while the Intel HD Graphics 530 scores 3550, giving the NVIDIA card a 5% lead. This is a modest but real advantage, suggesting that in OpenCL workloads, the discrete card's dedicated memory bandwidth and higher shading unit count provide a measurable benefit, even though the Intel part is not far behind. The 5% delta is within the range of run-to-run variance seen in some benchmarks, but it does consistently favor the NVIDIA part.

The largest defeat for the Intel HD Graphics 530 comes in the Geekbench Vulkan test, where it scores 1422 versus the NVIDIA GeForce GT 640's 3809. This is a 62.7% deficit for the Intel part, meaning the NVIDIA card is roughly 2.7 times faster in this workload. The Vulkan API is designed to reduce driver overhead and provide more direct hardware control, and the results indicate that the NVIDIA Kepler architecture handles this API far more effectively than Intel's Generation 9.0 architecture. The NVIDIA card's Vulkan support is also more mature, with version 1.2.175 compared to the Intel part's 1.3, though the Intel part nominally supports a higher version number.

Looking at the broader benchmark context, the Intel HD Graphics 530 has an average benchmark score of 3332, placing it in the 20th percentile of all GPUs. Its nearest rivals include the NVIDIA GeForce GT 730M with an average score of 3316 (a 0.5% delta), the NVIDIA GeForce 920M at 3287 (1.4% delta), and the Intel HD Graphics P4600 at 3389 (-1.7% delta). The NVIDIA GeForce GT 640, meanwhile, has an average benchmark score of 3210, also in the 20th percentile, with rivals including the Intel Arc Pro B60 at 3182 (0.9% delta), the NVIDIA Quadro P1000 at 3163 (1.5% delta), and the NVIDIA GeForce 920M at 3287 (-2.3% delta). Both parts sit in the same performance tier overall, though their individual benchmark profiles are quite different.

The Verdict

The data shows a clear split based on API and workload type. The Intel HD Graphics 530 is the better choice for Metal-based applications, where its 5025 score dwarfs the NVIDIA GeForce GT 640's 2086. If a user's primary workload is Metal compute or graphics, the Intel part is overwhelmingly superior, with a 140.9% performance advantage that is difficult to overstate. This makes it the preferred option for macOS environments or any application that heavily relies on Metal.

The NVIDIA GeForce GT 640 is the better choice for Vulkan workloads, where its 3809 score versus the Intel part's 1422 represents a 62.7% advantage. This is a substantial margin that makes the NVIDIA card clearly superior for Vulkan-based games or compute applications. The NVIDIA card also edges out the Intel part in OpenCL, though the 5% delta is smaller and less decisive than the other two benchmarks.

For users who need a balanced solution across multiple APIs, the choice depends on which API dominates their usage. The Intel HD Graphics 530 has a higher average benchmark score overall (3332 versus 3210), but this average is heavily influenced by its exceptional Metal result. The NVIDIA GeForce GT 640 is more consistent across OpenCL and Vulkan, while the Intel part is a one-trick pony with a massive Metal advantage and significant deficits elsewhere. Given that the NVIDIA card wins two out of three head-to-head tests, it may be the more versatile option for mixed workloads, despite the Intel part's higher average score.

Architecture Differences

The two graphics solutions differ fundamentally in their architecture and design philosophy. The Intel HD Graphics 530 is an integrated graphics processor (IGP) based on the Skylake GT2 chip, using Intel's Generation 9.0 architecture on a 14 nm+ process node fabricated by Intel. It has a die size of 123 mm² and uses a Ring Bus interface. The NVIDIA GeForce GT 640, in contrast, is a discrete card based on the GK107 chip, using the Kepler architecture on a 28 nm process node fabricated by TSMC. It has a die size of 118 mm² and uses a PCIe 3.0 x16 bus interface.

The memory configurations are starkly different. The Intel HD Graphics 530 uses system shared memory, with its size, type, and bus width all listed as "System Shared" and bandwidth as "System Dependent." The NVIDIA GeForce GT 640 has 2 GB of dedicated DDR3 memory on a 128-bit bus, providing 28.51 GB/s of bandwidth. This dedicated memory gives the NVIDIA card a significant advantage in memory-intensive workloads, though the Intel part's shared memory can benefit from faster system RAM.

The compute resources also differ. The Intel HD Graphics 530 has 192 shading units, 24 texture mapping units (TMUs), and 3 raster output units (ROPs). The NVIDIA GeForce GT 640 has 384 shading units, 32 TMUs, and 16 ROPs. This gives the NVIDIA card double the shading units and significantly more ROPs, which explains its higher pixel rate of 7.216 GPixel/s versus the Intel part's 2.850 GPixel/s. The texture rate also favors the NVIDIA card at 28.86 GTexel/s versus 22.80 GTexel/s, and the FP32 performance is nearly double at 692.7 GFLOPS versus 364.8 GFLOPS.

Clock speeds tell a different story. The Intel HD Graphics 530 has a base clock of 350 MHz and a boost clock of 950 MHz, while the NVIDIA GeForce GT 640 has no listed base or boost clocks, only a memory clock of 891 MHz (1782 Mbps effective). The Intel part's boost clock is higher than the NVIDIA card's memory clock, but this comparison is not directly meaningful given the different architectures and roles.

Power consumption is another major difference. The Intel HD Graphics 530 has a TDP of 15 W and is an IGP, meaning it requires no additional power connectors and is integrated into the motherboard. The NVIDIA GeForce GT 640 has a TDP of 65 W, is a single-slot card with no power connectors, and has a suggested PSU of 250 W. The Intel part is far more power-efficient, which is expected for an integrated solution.

API support differs as well. The Intel HD Graphics 530 supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. The NVIDIA GeForce GT 640 supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The Intel part has a higher DirectX feature level and a newer Vulkan version, though the benchmark results show the NVIDIA card performs better in Vulkan despite the lower version number.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The Intel HD Graphics 530 has an average benchmark score of 3332, while the NVIDIA GeForce GT 640 has an average of 3210, giving the Intel part a 3.8% advantage overall.

Q: How much faster is the Intel HD Graphics 530 in Geekbench Metal?

A: The Intel HD Graphics 530 scores 5025 in Geekbench Metal, which is 140.9% higher than the NVIDIA GeForce GT 640's score of 2086.

Q: What is the memory bandwidth of the NVIDIA GeForce GT 640?

A: The NVIDIA GeForce GT 640 has 2 GB of DDR3 memory on a 128-bit bus, providing 28.51 GB/s of bandwidth.

Q: Which GPU has better Vulkan performance?

A: The NVIDIA GeForce GT 640 has significantly better Vulkan performance, scoring 3809 versus the Intel HD Graphics 530's 1422, a 62.7% advantage.

Q: What are the TDP ratings for each GPU?

A: The Intel HD Graphics 530 has a TDP of 15 W, while the NVIDIA GeForce GT 640 has a TDP of 65 W.

Q: How do the shading unit counts compare?

A: The NVIDIA GeForce GT 640 has 384 shading units, which is double the 192 shading units of the Intel HD Graphics 530.

Where Each One Wins

The Intel HD Graphics 530 wins decisively in Metal-based workloads, with a 140.9% performance advantage over the NVIDIA GeForce GT 640. This makes it the clear choice for applications that leverage Metal, whether for graphics rendering, compute tasks, or any other Metal-accelerated functionality. The Intel part's 5025 Metal score places it well above not only the NVIDIA card but also above its own nearest rivals, all of which score below 3400 in average benchmark terms. For users in ecosystems where Metal is the primary API, the Intel HD Graphics 530 is the superior option.

The NVIDIA GeForce GT 640 wins in Vulkan workloads by a wide margin, with a 62.7% advantage over the Intel part. This makes it the better choice for Vulkan-based games, compute applications, or any workflow that relies on this modern, low-overhead API. The NVIDIA card's 3809 Vulkan score is more than double the Intel part's 1422, indicating a fundamental architectural advantage in handling Vulkan's direct hardware access model. The NVIDIA card also holds a 5% edge in OpenCL, making it the better choice for OpenCL-based compute tasks, though this margin is smaller and less decisive.

The NVIDIA GeForce GT 640 also wins in scenarios that benefit from dedicated memory and higher raw throughput. Its 2 GB of DDR3 memory with 28.51 GB/s bandwidth, combined with 384 shading units and 16 ROPs, gives it a pixel rate of 7.216 GPixel/s and a texture rate of 28.86 GTexel/s. These figures are substantially higher than the Intel HD Graphics 530's 2.850 GPixel/s and 22.80 GTexel/s, suggesting the NVIDIA card is better suited for traditional rasterization workloads that depend on fill rate and memory bandwidth. The Intel part, with its 15 W TDP and system shared memory, is better suited for power-constrained environments or basic graphics tasks where its lower power draw and integrated nature are advantages.