NVIDIA GeForce GTX 560 vs NVIDIA GeForce GTX 870M Comparison

The NVIDIA GeForce GTX 870M and NVIDIA GeForce GTX 560 represent two distinct eras of mobile and desktop GPU design, separated by three years of architectural evolution. The GTX 870M, a Kepler-based mobile part from the GeForce 800M generation, and the GTX 560, a Fermi 2.0 desktop card from the GeForce 500 series, deliver nearly identical average benchmark scores, with the former edging ahead. The data reveals a fascinating generational shift: the newer mobile chip achieves comparable compute results while consuming significantly less power and fitting into a slimmer MXM module, whereas the older desktop card relies on a larger die and higher thermal envelope to stay competitive.

Where Each One Wins

The benchmark data establishes a clear, albeit narrow, victory for the GTX 870M in the only head-to-head compute test available. In Geekbench OpenCL, the GTX 870M scores 12630, which is 39.4% ahead of the GTX 560’s 9058. This single dominant win indicates that the Kepler architecture’s compute throughput is substantially superior to Fermi 2.0 in raw GPU-accelerated workloads, making the 870M the clear choice for OpenCL-based applications like video encoding, physics simulation, and general-purpose GPU computing.

However, the GTX 560 is not entirely without merits. While it loses the compute benchmark, its design as a desktop card with a PCIe 2.0 x16 interface and dual-slot cooler suggests it was built for sustained, high-performance gaming in a stationary chassis. The 560’s 256-bit memory bus provides 128.0 GB/s of bandwidth, which is higher than the 870M’s 120.0 GB/s, potentially giving it an edge in memory-bandwidth-sensitive scenarios like high-resolution texture streaming, though no direct benchmark confirms this. The GTX 560 also benefits from a higher TDP of 150 W, allowing it to maintain boost clocks more aggressively under load, assuming adequate cooling, whereas the 870M’s 100 W limit is a constraint of its mobile form factor.

The percentile rankings reinforce this split. The GTX 870M sits at the 48th percentile of all GPUs, while the GTX 560 ranks at the 45th. This three-percentile gap is modest, but it places the 870M in a slightly stronger overall performance tier. The average benchmark score of 9959 for the 870M versus 9058 for the 560 confirms that the newer mobile part is, on aggregate, the more capable compute device. For users prioritizing portability or OpenCL performance, the 870M wins outright; for those needing a legacy desktop card with a wider memory bus and higher power ceiling, the 560 remains a functional, if dated, option.

Architecture Differences

The architectural gap between these two GPUs is vast, reflecting a generational leap in NVIDIA’s design philosophy. The GTX 870M is built on the Kepler architecture using the GK104 chip, fabricated on TSMC’s 28 nm process. This chip packs 3,540 million transistors into a 294 mm² die, achieving a transistor density of 12.0M per mm². In contrast, the GTX 560 uses the Fermi 2.0 architecture with the GF114 chip, produced on TSMC’s 40 nm node. The older chip contains only 1,950 million transistors spread across a larger 332 mm² die, resulting in a much lower density of 5.9M per mm². This density difference is the core of the performance-per-watt divide: the 28 nm process allows Kepler to pack nearly twice the transistors into a smaller area while consuming 33% less power (100 W vs 150 W).

The compute resources tell a similar story of generational scaling. The GTX 870M houses 1344 shading units, 112 texture mapping units (TMUs), and 24 ROPs, whereas the GTX 560 offers 336 shading units, 56 TMUs, and 32 ROPs. The 870M has four times the shading units and double the TMUs, but fewer ROPs, which explains its massive lead in texture rate (108.3 GTexel/s vs 45.36 GTexel/s) and pixel rate (27.08 GPixel/s vs 11.34 GPixel/s). The FP32 compute output is equally lopsided: the 870M delivers 2.599 TFLOPS, while the 560 manages just 1,088.6 GFLOPS. This is a direct consequence of Kepler’s more efficient shader design and higher clock speeds, with the 870M boosting to 967 MHz versus the 560’s fixed, unspecified base clock.

Memory configurations also diverge sharply. The 870M uses 3 GB of GDDR5 on a 192-bit bus, while the 560 uses 1 GB of GDDR5 on a 256-bit bus. Despite the narrower bus, the 870M’s higher memory clock (1250 MHz, 5 Gbps effective) yields 120.0 GB/s bandwidth, close to the 560’s 128.0 GB/s from a slower 1000 MHz (4 Gbps effective) clock. The 870M’s larger capacity is a clear win for modern workloads that demand more VRAM, though the 560’s wider bus offers marginally higher raw bandwidth. Finally, the 870M supports Vulkan 1.2.175, while the 560 has no Vulkan support listed, and it uses a portable MXM-B (3.0) interface versus the 560’s PCIe 2.0 x16 slot, reflecting their mobile versus desktop designations.

Head-to-Head Benchmarks

The only direct benchmark comparison in the data is Geekbench OpenCL, and it paints a decisive picture. The GTX 870M scores 12630, while the GTX 560 scores 9058, yielding a 39.4% delta in favor of the 870M. This is not a marginal victory; it is a substantial performance gap that underscores Kepler’s compute superiority. For context, the 870M’s score places it within 0.5% of the AMD Radeon Pro 5300M and 0.8% of the NVIDIA Quadro K5100M, according to its nearest rivals list, while the 560 sits just 0.2% above the NVIDIA TITAN V CEO Edition and 0.4% above the GeForce GTX 660. These rival comparisons show that the 870M is competitive with professional mobile GPUs, while the 560 hangs near mid-range desktop parts from its own generation.

The 39.4% delta translates to the 870M completing OpenCL workloads nearly 1.4 times faster than the 560. This advantage is likely driven by the 870M’s 1344 shading units operating at a higher boost clock, delivering 2.599 TFLOPS of FP32 throughput versus the 560’s 1,088.6 GFLOPS. Even accounting for the 560’s slightly higher memory bandwidth, the compute bottleneck is overwhelmingly in the shader and texture units, where the 870M holds a 4x and 2x advantage, respectively. The pixel rate also favors the 870M by 2.4x, though the 560’s higher ROP count (32 vs 24) suggests it was optimized for fill-rate-bound scenarios at lower resolutions, a legacy design choice that cannot overcome Kepler’s raw compute lead.

There are no other head-to-head benchmarks in the data, meaning the Geekbench Metal score of 7288 for the 870M has no direct counterpart for the 560. This single-test dataset limits the scope of comparison, but the OpenCL result is sufficient to establish the 870M as the clear winner in compute-oriented tasks. The absence of a gaming benchmark, such as 3DMark or Unigine, means that real-world frame rates cannot be inferred from this data, but the architectural analysis strongly suggests the 870M would dominate in most modern titles due to its superior texture rate and FP32 output.

FAQ

Q: Which GPU has a higher average benchmark score?

A: The NVIDIA GeForce GTX 870M has an average benchmark score of 9959, while the NVIDIA GeForce GTX 560 scores 9058, making the 870M approximately 10% faster on average.

Q: How much faster is the GTX 870M in OpenCL?

A: In the Geekbench OpenCL test, the GTX 870M scores 12630 versus the GTX 560’s 9058, a delta of 39.4% in favor of the 870M.

Q: What are the memory capacity differences?

A: The GTX 870M comes with 3 GB of GDDR5 memory, while the GTX 560 has 1 GB of GDDR5. The 870M’s larger capacity is better suited for high-resolution textures and modern game assets.

Q: Do both GPUs support the same APIs?

A: No. The GTX 870M supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. The GTX 560 supports DirectX 12 (11_0) and OpenGL 4.6, but has no listed Vulkan support.

Q: Which GPU has a smaller manufacturing process?

A: The GTX 870M is built on a 28 nm TSMC process, whereas the GTX 560 uses a 40 nm TSMC process. The smaller node allows the 870M to pack 3,540 million transistors into a 294 mm² die, versus 1,950 million in 332 mm² for the 560.

Q: How do their power requirements compare?

A: The GTX 870M has a TDP of 100 W and uses no power connectors (MXM module), while the GTX 560 has a TDP of 150 W and requires 2x 6-pin power connectors with a suggested PSU of 450 W.

Specification Differences

The two GPUs differ across nearly every core specification, reflecting their distinct generations and form factors. The GTX 870M uses the GK104 chip on a 28 nm process with 3,540 million transistors, while the GTX 560 uses the GF114 chip on a 40 nm process with 1,950 million transistors. The die sizes are 294 mm² for the 870M and 332 mm² for the 560, giving the 870M a transistor density of 12.0M per mm² versus 5.9M per mm² for the 560.

In terms of compute resources, the 870M has 1344 shading units, 112 TMUs, and 24 ROPs, whereas the 560 has 336 shading units, 56 TMUs, and 32 ROPs. The 870M’s base clock is 941 MHz with a boost of 967 MHz, while the 560 has no listed base or boost clock. Memory clocks are 1250 MHz (5 Gbps effective) for the 870M and 1000 MHz (4 Gbps effective) for the 560. The 870M offers 3 GB of GDDR5 on a 192-bit bus with 120.0 GB/s bandwidth, while the 560 offers 1 GB of GDDR5 on a 256-bit bus with 128.0 GB/s bandwidth.

Performance rates diverge significantly: the 870M delivers 27.08 GPixel/s pixel rate, 108.3 GTexel/s texture rate, and 2.599 TFLOPS FP32, compared to the 560’s 11.34 GPixel/s, 45.36 GTexel/s, and 1,088.6 GFLOPS. The TDP is 100 W for the 870M and 150 W for the 560, with the former using an MXM-B (3.0) interface and the latter using PCIe 2.0 x16. The 870M has no power connectors and portable-device-dependent display outputs, while the 560 has 2x 6-pin connectors, a 450 W suggested PSU, and outputs of 2x DVI and 1x mini-HDMI 1.3a. The 560 also has a listed length of 210 mm (8.3 inches) and a launch MSRP of 199 USD, while the 870M has no MSRP or dimensions listed.