NVIDIA GeForce GTX 560 vs NVIDIA GeForce GTX 950A Comparison

FAQ

Q: How do the two GPUs compare in overall benchmark performance?

A: The NVIDIA GeForce GTX 950A scores 10,273 in Geekbench OpenCL, while the NVIDIA GeForce GTX 560 scores 9,058. That makes the GTX 950A 13.4% faster in the only head-to-head benchmark recorded in the database.

Q: What is the transistor density of each chip?

A: The GTX 950A uses the GM107 chip with 1,870 million transistors on a 148 mm² die, yielding 12.6M transistors per mm². The GTX 560 uses the GF114 chip with 1,950 million transistors on a 332 mm² die, yielding 5.9M transistors per mm².

Q: Which GPU has higher memory bandwidth?

A: The GTX 560 has substantially higher memory bandwidth at 128.0 GB/s, thanks to a 256-bit bus with GDDR5 memory. The GTX 950A manages only 32.03 GB/s over a 128-bit bus with DDR3 memory.

Q: What are the power requirements for each card?

A: The GTX 950A draws 75 W and requires no power connectors. The GTX 560 draws 150 W and requires two 6-pin power connectors plus a 450 W suggested PSU.

Q: What is the release timeline for these GPUs?

A: The GTX 560 was released on May 16, 2011, and belongs to the GeForce 500 generation. The GTX 950A came later, on March 12, 2015, and belongs to the GeForce 900A generation. Both are end-of-life products.

Q: How do their percentile rankings compare across all GPUs?

A: The GTX 950A sits at the 48th percentile of all GPUs in the database, while the GTX 560 sits at the 45th percentile. This places both in the lower-middle range of recorded GPUs.

Architecture Differences

The two GPUs represent distinct architectural eras from NVIDIA. The GTX 950A is built on the Maxwell architecture using the GM107 chip, fabricated on a 28 nm process at TSMC. The GTX 560 uses the older Fermi 2.0 architecture with the GF114 chip, fabricated on a 40 nm process, also at TSMC.

The manufacturing process difference is significant. The 28 nm node allows the GTX 950A to pack 1,870 million transistors into just 148 mm², producing a transistor density of 12.6M per mm². The GTX 560's 40 nm node requires a much larger 332 mm² die to house 1,950 million transistors, resulting in a lower density of 5.9M per mm². Despite having fewer total transistors, the GTX 950A achieves higher density and better efficiency per area.

The shading unit counts differ notably. The GTX 950A has 640 shading units, while the GTX 560 has 336 shading units. However, the GTX 560 compensates with more texture mapping units (56 versus 40) and double the render output units (32 versus 16). These counts reflect different design priorities: Maxwell favors compute throughput per shader, while Fermi 2.0 allocated more resources to texture and pixel processing.

Memory architecture differs fundamentally. The GTX 950A uses 2 GB of DDR3 on a 128-bit bus, delivering 32.03 GB/s of bandwidth. The GTX 560 uses 1 GB of GDDR5 on a 256-bit bus, delivering 128.0 GB/s. This fourfold bandwidth advantage is the GTX 560's key architectural strength.

Clock behavior also differs. The GTX 950A has a base clock of 993 MHz and a boost clock of 1124 MHz. The GTX 560 has no base or boost clock recorded in the database. Memory clocks are similar in base frequency (1001 MHz versus 1000 MHz), but the effective data rates differ: 2 Gbps for the GTX 950A versus 4 Gbps for the GTX 560, reflecting the different memory types.

Both GPUs support DirectX 12 (11_0) and OpenGL 4.6. The GTX 950A supports Vulkan 1.4, while the GTX 560 has no recorded Vulkan support. This reflects the newer API stack available to Maxwell parts.

Form factor and interface differ substantially. The GTX 950A is an MXM module with an MXM-B (3.0) interface, designed for portable devices with dependent display outputs. The GTX 560 is a dual-slot PCIe 2.0 x16 card with 2x DVI and 1x mini-HDMI 1.3a outputs, measuring 210 mm (8.3 inches) in length.

The Verdict

The benchmark data points clearly to the GTX 950A as the stronger performer. Its Geekbench OpenCL score of 10,273 beats the GTX 560's 9,058 by 13.4%. The GTX 950A also ranks higher in the global percentile distribution, at 48 versus 45.

For users prioritizing raw compute performance in OpenCL workloads, the GTX 950A is the better choice. It delivers higher throughput, has double the memory capacity at 2 GB versus 1 GB, and requires half the power at 75 W versus 150 W. It also avoids the need for external power connectors, making it simpler to integrate.

However, the GTX 560 holds advantages in specific areas. Its memory bandwidth of 128.0 GB/s is four times that of the GTX 950A, which matters for bandwidth-sensitive workloads. It also has more texture mapping units and render output units, which can benefit certain rendering tasks. The GTX 560 launched at 199 USD as its launch MSRP, though this should not be interpreted as a current price indicator.

The GTX 950A is the better pick for most compute-oriented scenarios given its higher benchmark score, better efficiency, and newer architecture. The GTX 560 is the pick only for scenarios where memory bandwidth and texture throughput are the limiting factors, and where the older Fermi architecture's characteristics are acceptable.

Specification Differences

| Specification | GTX 950A | GTX 560 |

|---|---|---|

| Architecture | Maxwell | Fermi 2.0 |

| Process Node | 28 nm | 40 nm |

| Transistors | 1,870 million | 1,950 million |

| Die Size | 148 mm² | 332 mm² |

| Transistor Density | 12.6M / mm² | 5.9M / mm² |

| Base Clock | 993 MHz | None recorded |

| Boost Clock | 1124 MHz | None recorded |

| Memory Clock | 1001 MHz, 2 Gbps effective | 1000 MHz, 4 Gbps effective |

| Memory Size | 2 GB | 1024 MB |

| Memory Type | DDR3 | GDDR5 |

| Memory Bus Width | 128 bit | 256 bit |

| Memory Bandwidth | 32.03 GB/s | 128.0 GB/s |

| Shading Units | 640 | 336 |

| TMUs | 40 | 56 |

| ROPs | 16 | 32 |

| Pixel Rate | 17.98 GPixel/s | 11.34 GPixel/s |

| Texture Rate | 44.96 GTexel/s | 45.36 GTexel/s |

| FP32 | 1,438.7 GFLOPS | 1,088.6 GFLOPS |

| TDP | 75 W | 150 W |

| Slot Width | MXM Module | Dual-slot |

| Power Connectors | None | 2x 6-pin |

| Suggested PSU | None recorded | 450 W |

| Bus Interface | MXM-B (3.0) | PCIe 2.0 x16 |

| Display Outputs | Portable Device Dependent | 2x DVI, 1x mini-HDMI 1.3a |

| Vulkan Support | 1.4 | None recorded |

| Release Date | 2015-03-12 | 2011-05-16 |

| Predecessor | GeForce 800A | GeForce 400 |

| Successor | None recorded | GeForce 600 |

Head-to-Head Benchmarks

The database records a single head-to-head benchmark between these two GPUs: Geekbench OpenCL. The GTX 950A scores 10,273 against the GTX 560's 9,058, a 13.4% advantage for the Maxwell part.

This gap is consistent with the FP32 compute figures. The GTX 950A delivers 1,438.7 GFLOPS of FP32 throughput, which is 32.2% higher than the GTX 560's 1,088.6 GFLOPS. The benchmark delta of 13.4% is smaller than the raw FP32 delta, which suggests that other factors such as memory bandwidth constrain the GTX 950A's real-world OpenCL performance.

The nearest rivals in the database provide additional context. The GTX 950A sits within 2% of several AMD parts: it is 0.9% behind the Radeon RX 6500M (10,362), 1.2% behind the Tesla C2075 (10,400), 2% ahead of the Radeon R9 M375 (10,070), and 2% behind the Radeon RX 550X (10,481). These close margins place the GTX 950A in a tight cluster of mid-range GPUs.

The GTX 560's nearest rivals tell a similar story at a lower performance tier. It is 0.2% ahead of the NVIDIA TITAN V CEO Edition (9,037), 0.4% ahead of the GeForce GTX 660 (9,022), 1.6% ahead of the AMD Radeon 550X (8,918), and 1.7% behind the AMD Radeon 890M (9,210). The GTX 560's score of 9,058 places it in a cluster with these parts, all within roughly 2% of each other.

The pixel rate comparison favors the GTX 950A at 17.98 GPixel/s versus 11.34 GPixel/s for the GTX 560, a 58.6% advantage. The texture rate is nearly identical: 44.96 GTexel/s for the GTX 950A versus 45.36 GTexel/s for the GTX 560, a 0.9% edge for the older card.

Where Each One Wins

The GTX 950A wins on compute throughput. Its FP32 output of 1,438.7 GFLOPS exceeds the GTX 560's 1,088.6 GFLOPS by a wide margin. This translates to a 13.4% lead in the Geekbench OpenCL benchmark, the only head-to-head test recorded. The GTX 950A also achieves a higher pixel rate at 17.98 GPixel/s versus 11.34 GPixel/s, indicating stronger fill-rate performance for pixel-heavy workloads.

The GTX 950A wins on efficiency and integration. Its 75 W TDP is exactly half of the GTX 560's 150 W. It requires no power connectors, while the GTX 560 needs two 6-pin connectors and a 450 W suggested PSU. The MXM form factor makes it suitable for portable devices, though this also means its display outputs are device-dependent.

The GTX 950A wins on memory capacity and API support. It has 2 GB of memory versus 1 GB, doubling the available framebuffer. It supports Vulkan 1.4, which the GTX 560 lacks entirely. It also has a higher transistor density (12.6M versus 5.9M per mm²) and a smaller die (148 mm² versus 332 mm²), indicating better manufacturing efficiency.

The GTX 560 wins on memory bandwidth. Its 128.0 GB/s bandwidth is exactly four times the GTX 950A's 32.03 GB/s. This is the single largest architectural advantage for the older card. For workloads that stream large amounts of data, such as high-resolution textures or certain compute kernels, this bandwidth advantage could offset some of the GTX 950A's compute lead.

The GTX 560 also wins on texture throughput, though narrowly. Its texture rate of 45.36 GTexel/s edges out the GTX 950A's 44.96 GTexel/s by 0.9%. It also has more TMUs (56 versus 40) and ROPs (32 versus 16), which can help in specific rendering scenarios.

The GTX 560 has a longer physical footprint at 210 mm (8.3 inches) as a dual-slot card, versus the MXM module form factor of the GTX 950A. It also has a documented predecessor and successor (GeForce 400 and GeForce 600, respectively), while the GTX 950A's lineage shows a predecessor (GeForce 800A) but no successor. The GTX 560's launch MSRP was 199 USD, a figure the database records but which should not be compared to any current pricing.