NVIDIA GeForce GTX 650 Ti Boost vs NVIDIA GeForce GTX 880M Comparison

The NVIDIA GeForce GTX 650 Ti Boost and the NVIDIA GeForce GTX 880M are both Kepler-based parts, yet they target completely different market segments: a desktop card versus a mobile MXM module. Their average benchmark scores are nearly identical, but the underlying performance profiles diverge sharply depending on the workload. The data shows two GPUs with a 0.3% difference in average score, but with drastically different strengths in specific API tests.

Head-to-Head Benchmarks

The head-to-head comparison reveals a clear split based on the benchmark used. In Geekbench Metal, the NVIDIA GeForce GTX 880M delivers a score of 10458, which is a massive 53.5% higher than the GTX 650 Ti Boost’s 4858. This is a dominant win for the mobile part. The margin is substantial, indicating the 880M’s architecture is significantly better optimized for Metal compute tasks. In contrast, the Geekbench OpenCL test flips the result completely. The GTX 650 Ti Boost scores 9302, while the GTX 880M manages only 5622. This results in a 65.5% advantage for the desktop card. The delta between the two tests is stark: a 53.5% deficit in one, and a 65.5% lead in the other.

Looking at the overall average benchmark score, the two cards are essentially tied. The GTX 650 Ti Boost has an average score of 8067, while the GTX 880M sits at 8040. The nearest rival data confirms this proximity, with the GTX 650 Ti Boost being 0.3% ahead of the GTX 880M. In the GTX 880M’s rival list, the GTX 650 Ti Boost is shown as 0.3% ahead (a deltaPct of -0.3 from the 880M’s perspective). This means that in a generalized compute benchmark, you would not notice a difference between the two. The wins are entirely workload-dependent, with the 880M owning Metal and the 650 Ti Boost owning OpenCL. The Geekbench Vulkan score for the GTX 650 Ti Boost is 10041, which is not part of the head-to-head data but is its highest individual benchmark result, suggesting strong Vulkan performance relative to its own OpenCL and Metal scores.

Architecture Differences

Both GPUs are built on the same fundamental architecture, Kepler, and are fabricated by TSMC on a 28 nm process node. However, the silicon underneath is different. The GTX 650 Ti Boost uses the GK106 chip, while the GTX 880M uses the larger GK104. This difference in physical die size is significant: the GK104 measures 294 mm² and contains 3,540 million transistors, while the GK106 is smaller at 221 mm² with 2,540 million transistors. The transistor density is similar, with 12.0M per mm² for the 880M versus 11.5M per mm² for the 650 Ti Boost, but the raw count is 39% higher on the mobile chip.

The core configuration mirrors this difference. The GTX 880M has 1536 shading units, 128 TMUs, and 32 ROPs. The GTX 650 Ti Boost is equipped with 768 shading units, 64 TMUs, and 24 ROPs. This means the 880M has exactly double the shading units and TMUs, and 33% more ROPs. Consequently, the theoretical peak rates are much higher on the mobile part: the pixel rate is 31.78 GPixel/s versus 16.51 GPixel/s, and the texture rate is 127.1 GTexel/s versus 66.05 GTexel/s. FP32 compute is also nearly double, at 3.050 TFLOPS for the 880M against 1.585 TFLOPS for the 650 Ti Boost.

Clock speeds tell a different story. The desktop card runs at a base clock of 980 MHz with a boost clock of 1032 MHz, while the mobile card is slightly slower at 954 MHz base and 993 MHz boost. Memory configurations also differ. The GTX 650 Ti Boost uses 2 GB of GDDR5 on a 192-bit bus, with a memory clock of 1502 MHz (6 Gbps effective) yielding 144.2 GB/s of bandwidth. The GTX 880M offers 8 GB of GDDR5 on a wider 256-bit bus, but at a lower clock of 1250 MHz (5 Gbps effective), resulting in 160.0 GB/s. Despite the lower clock, the wider bus gives the 880M an 11% bandwidth advantage. Power characteristics are inverted from what you might expect: the desktop GTX 650 Ti Boost has a TDP of 134 W, while the mobile GTX 880M is rated at 122 W. The desktop card requires a 6-pin power connector and a 300 W suggested PSU, while the 880M uses an MXM-B (3.0) interface with no separate power connectors, as it is designed for portable devices.

FAQ

Q: Which GPU is faster in Metal-based workloads?

A: The NVIDIA GeForce GTX 880M is decisively faster in Geekbench Metal, scoring 10458 compared to the GTX 650 Ti Boost’s 4858. This represents a 53.5% lead for the 880M.

Q: Which GPU is faster in OpenCL-based workloads?

A: The NVIDIA GeForce GTX 650 Ti Boost wins the Geekbench OpenCL test with a score of 9302, while the GTX 880M scores 5622. The desktop card holds a 65.5% advantage.

Q: Are these two cards similar in overall performance?

A: Yes, the average benchmark scores are nearly identical. The GTX 650 Ti Boost averages 8067, and the GTX 880M averages 8040, a difference of only 0.3%. The GTX 650 Ti Boost is ranked as 0.3% ahead of the GTX 880M in its nearest rivals list.

Q: Do these GPUs use the same chip?

A: No. The GTX 650 Ti Boost uses the GK106 chip, while the GTX 880M uses the GK104 chip. The GK104 is physically larger, with a die size of 294 mm² versus 221 mm², and contains 3,540 million transistors versus 2,540 million.

Q: Which card has more memory and bandwidth?

A: The GTX 880M has 8 GB of GDDR5 memory on a 256-bit bus, providing 160.0 GB/s of bandwidth. The GTX 650 Ti Boost has 2 GB on a 192-bit bus, providing 144.2 GB/s. The 880M has more capacity and higher bandwidth.

Q: Which GPU has higher theoretical compute rates?

A: The GTX 880M has higher theoretical rates across the board. Its FP32 performance is 3.050 TFLOPS, its pixel rate is 31.78 GPixel/s, and its texture rate is 127.1 GTexel/s. The GTX 650 Ti Boost is lower in all three: 1.585 TFLOPS, 16.51 GPixel/s, and 66.05 GTexel/s.

Specification Differences

The two cards differ in nearly every core specification beyond their shared Kepler architecture and 28 nm process node. The GTX 650 Ti Boost is built on the GK106 chip, whereas the GTX 880M uses the larger GK104. The GTX 880M has a larger die size (294 mm² versus 221 mm²) and more transistors (3,540 million versus 2,540 million). The shading units, TMUs, and ROPs are all higher on the 880M: 1536 versus 768 shading units, 128 versus 64 TMUs, and 32 versus 24 ROPs. Theoretical pixel rate (31.78 GPixel/s versus 16.51 GPixel/s), texture rate (127.1 GTexel/s versus 66.05 GTexel/s), and FP32 compute (3.050 TFLOPS versus 1.585 TFLOPS) are all roughly double on the 880M. Clock speeds are slightly higher on the 650 Ti Boost, with a base of 980 MHz versus 954 MHz and a boost of 1032 MHz versus 993 MHz. Memory capacity differs dramatically (2 GB versus 8 GB), as does bus width (192-bit versus 256-bit), memory clock (1502 MHz versus 1250 MHz), and effective speed (6 Gbps versus 5 Gbps). Bandwidth is higher on the 880M at 160.0 GB/s versus 144.2 GB/s. TDP is lower on the 880M (122 W versus 134 W). The form factor and interface are completely different: the 650 Ti Boost is a dual-slot PCIe 3.0 x16 card with a 6-pin connector and a 300 W suggested PSU, while the 880M is an MXM Module using an MXM-B (3.0) interface with no power connectors. Display outputs also differ: the desktop card has 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2, while the mobile card’s outputs are portable device dependent. The GTX 650 Ti Boost has a set length of 241 mm (9.5 inches), while the 880M has no listed dimensions.

Where Each One Wins

The GTX 650 Ti Boost wins in scenarios where OpenCL performance is critical. Its 65.5% lead in the Geekbench OpenCL test is the single largest performance gap between the two cards. This makes it the better choice for compute tasks that rely on OpenCL acceleration, such as certain physics simulations or heterogeneous computing workloads. It also runs at higher clock speeds (980 MHz base versus 954 MHz base), which may benefit lightly threaded or latency-sensitive tasks. As a desktop card with a PCIe 3.0 x16 interface and dedicated display outputs, it is suited for a traditional desktop workstation or gaming rig. Its lower memory capacity (2 GB) is a limitation, but its higher memory clock (1502 MHz versus 1250 MHz) suggests lower latency per access. The GTX 650 Ti Boost also has a higher average benchmark score (8067 versus 8040), though the margin is negligible.

The GTX 880M wins in Metal-based workloads, where its 53.5% advantage in the Geekbench Metal test is decisive. This makes it the superior option for applications that leverage Apple’s Metal API, which is common in macOS environments. Its massive 8 GB memory capacity is a clear advantage for large datasets or high-resolution textures that exceed the 2 GB limit of the desktop card. The 880M also has double the shading units, TMUs, and ROPs, along with higher pixel, texture, and FP32 rates. This gives it a raw theoretical throughput advantage that can be exploited in workloads that scale well with parallel execution units, even if the OpenCL results do not reflect it. The 880M’s higher memory bandwidth (160.0 GB/s versus 144.2 GB/s) and wider 256-bit bus further support memory-intensive operations. Its lower TDP (122 W versus 134 W) makes it a more power-efficient part in terms of raw wattage, though it is constrained to a mobile form factor. The 880M is also newer, with a release date of March 2014 versus March 2013 for the 650 Ti Boost, and it sits in a different generation (GeForce 800M versus GeForce 600). For users in a portable environment requiring Metal support and large memory capacity, the 880M is the clear winner. For desktop users prioritizing OpenCL compute and higher clock speeds, the 650 Ti Boost holds the edge.