NVIDIA GeForce 945M vs NVIDIA GeForce GTX 650 Ti Boost Comparison

The NVIDIA GeForce 945M and the NVIDIA GeForce GTX 650 Ti Boost are two end-of-life graphics solutions from different eras of NVIDIA’s lineup, with the former targeting mobile platforms and the latter serving as a desktop part. While both share the same 28 nm process node from TSMC, their architectural foundations and memory subsystems create distinct performance profiles. The benchmark data shows a clear, albeit narrow, advantage for the desktop-oriented GTX 650 Ti Boost in raw compute workloads, but the 945M’s design efficiency and modern feature set make it a compelling option for portable systems.

Head-to-Head Benchmarks

The only direct comparison available in the data is the Geekbench OpenCL test, which measures general-purpose compute performance across a range of workloads. In this test, the NVIDIA GeForce GTX 650 Ti Boost scores 9,302 points, while the NVIDIA GeForce 945M scores 8,099 points. This represents a delta of -12.9% from the 945M’s perspective, meaning the GTX 650 Ti Boost outperforms the mobile chip by approximately 12.9%. This is a substantial margin in a synthetic benchmark, indicating that the desktop card’s higher shading unit count—768 versus 640—and significantly higher memory bandwidth are decisive factors in compute-heavy tasks.

However, the overall average benchmark scores tell a more nuanced story. The 945M’s average benchmark score is 8,099, while the GTX 650 Ti Boost’s average is 8,067. This puts the two cards remarkably close, with the 945M holding a slight 0.4% advantage over the GTX 650 Ti Boost in the aggregated data. This discrepancy between the single OpenCL test and the overall average suggests that the GTX 650 Ti Boost’s advantage may be specific to OpenCL workloads, while other benchmarks—possibly those that favor the 945M’s Maxwell architecture’s efficiency—could narrow or reverse the gap. The nearest rivals data reinforces this: the GTX 650 Ti Boost’s closest competitor is the NVIDIA GRID K2, with a delta of -0.2%, while the 945M’s closest rival is the AMD Radeon R9 M360, which scores 0.4% higher. Both cards sit at the 42nd percentile of all GPUs, indicating that they are statistically equivalent in overall performance distribution.

The GTX 650 Ti Boost also has additional benchmark results beyond OpenCL, including a Geekbench Metal score of 4,858 and a Geekbench Vulkan score of 10,041. These results highlight its versatility across different API environments, particularly in Vulkan, where it achieves a significantly higher score than its OpenCL result. The 945M, by contrast, only has a single OpenCL score in the data, leaving its performance in Metal and Vulkan undefined. This lack of data makes it difficult to assess the 945M’s modern API performance, though its Vulkan 1.4 support (versus the GTX 650 Ti Boost’s Vulkan 1.2.175) suggests architectural readiness for newer workloads.

FAQ

Q: Which GPU has a higher peak single-precision floating-point performance?

A: The NVIDIA GeForce GTX 650 Ti Boost has a higher FP32 throughput at 1.585 TFLOPS, compared to the NVIDIA GeForce 945M’s 1,305.6 GFLOPS. This is a direct consequence of the GTX 650 Ti Boost’s 768 shading units versus the 945M’s 640.

Q: How do their memory configurations affect bandwidth?

A: The GTX 650 Ti Boost uses a 192-bit GDDR5 interface with 2 GB of memory, delivering 144.2 GB/s of bandwidth. The 945M uses a 128-bit DDR3 interface with 2 GB of memory, providing only 28.80 GB/s. This five-fold difference in bandwidth is a major factor in the GTX 650 Ti Boost’s superior OpenCL score.

Q: Are these GPUs from the same architecture generation?

A: No. The 945M is based on the Maxwell architecture (chip GM107) from the GeForce 900M generation, while the GTX 650 Ti Boost is based on the Kepler architecture (chip GK106) from the GeForce 600 generation. Both are built on a 28 nm process at TSMC.

Q: What are the power consumption differences?

A: The 945M has a TDP of 75 W and requires no power connectors, reflecting its mobile MXM Module design. The GTX 650 Ti Boost has a TDP of 134 W, requires a single 6-pin power connector, and suggests a 300 W power supply. This makes the 945M significantly more power-efficient.

Q: Which card offers better overall performance according to average benchmark scores?

A: The 945M has a slightly higher average benchmark score of 8,099 versus the GTX 650 Ti Boost’s 8,067, a 0.4% difference. However, the GTX 650 Ti Boost wins the direct OpenCL comparison by 12.9%, indicating the overall average may be influenced by other tests.

Q: Do both cards support modern graphics APIs?

A: Both support DirectX 12 (11_0) and OpenGL 4.6. The 945M supports Vulkan 1.4, while the GTX 650 Ti Boost supports Vulkan 1.2.175. The 945M also has a higher API version, suggesting better long-term compatibility with newer Vulkan applications.

Architecture Differences

The architectural divide between these two GPUs is stark, stemming from their different generations and design goals. The 945M is built on NVIDIA’s Maxwell architecture, specifically the GM107 chip, which was engineered for efficiency in mobile and low-power environments. It packs 1,870 million transistors into a 148 mm² die, yielding a transistor density of 12.6 million per square millimeter. In contrast, the GTX 650 Ti Boost uses the older Kepler architecture with the GK106 chip, which houses 2,540 million transistors on a much larger 221 mm² die, resulting in a lower density of 11.5 million per square millimeter. This means the GTX 650 Ti Boost has more raw hardware—more transistors, more die area—but it is less efficiently packed.

The compute resources differ accordingly. The GTX 650 Ti Boost has 768 shading units, 64 texture mapping units (TMUs), and 24 raster output units (ROPs). The 945M has 640 shading units, 40 TMUs, and 16 ROPs. These numbers directly explain the GTX 650 Ti Boost’s higher texture rate (66.05 GTexel/s versus 40.80 GTexel/s) and its FP32 throughput advantage. However, the pixel rates are nearly identical: 16.51 GPixel/s for the GTX 650 Ti Boost and 16.32 GPixel/s for the 945M. This suggests that while the GTX 650 Ti Boost excels in texture-heavy and compute-heavy workloads, its advantage in pixel fill rate is negligible.

Memory architecture is another major differentiator. The GTX 650 Ti Boost uses GDDR5 memory on a 192-bit bus, achieving 144.2 GB/s of bandwidth. The 945M uses DDR3 memory on a 128-bit bus, achieving only 28.80 GB/s. This is a massive gap that impacts any bandwidth-sensitive operation, from high-resolution textures to compute kernels that rely on memory access. The GTX 650 Ti Boost’s memory clock is also higher at 6 Gbps effective, versus the 945M’s 1800 Mbps effective. This disparity is a primary reason for the GTX 650 Ti Boost’s OpenCL victory, as memory-bound tasks will see significant gains from the wider, faster interface.

The physical and power profiles contrast sharply as well. The 945M is an MXM Module with no power connectors and a 75 W TDP, designed to slot into laptops. The GTX 650 Ti Boost is a dual-slot desktop card, 241 mm in length, requiring one 6-pin power connector and a 300 W power supply, with a TDP of 134 W. The GTX 650 Ti Boost also offers a fixed set of display outputs (2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2), while the 945M’s outputs are portable device dependent. This reflects their intended use cases: one is a modular mobile component, the other a fixed desktop peripheral.

The Verdict

The data presents a clear choice based on workload and form factor. The NVIDIA GeForce GTX 650 Ti Boost is the superior performer in compute-heavy applications, as evidenced by its 12.9% lead in the OpenCL benchmark and its higher FP32 throughput of 1.585 TFLOPS. Its GDDR5 memory with 144.2 GB/s bandwidth provides a massive advantage for any task that saturates memory bandwidth, making it the better option for users who prioritize raw performance in a desktop system with adequate power delivery and cooling. The GTX 650 Ti Boost also has additional benchmark data in Metal and Vulkan, showing scores of 4,858 and 10,041 respectively, which suggests it is well-tested across multiple API environments.

On the other hand, the NVIDIA GeForce 945M is the pragmatic choice for mobile platforms. Its 75 W TDP and MXM Module form factor make it suitable for laptops where power and thermal constraints are paramount. Despite its lower raw compute performance, it achieves nearly identical pixel rate (16.32 GPixel/s) to the GTX 650 Ti Boost, and its average benchmark score of 8,099 is actually 0.4% higher, indicating that its efficiency-focused Maxwell architecture can hold its own in diverse workloads. The 945M also supports a newer Vulkan version (1.4 versus 1.2.175), which may offer better compatibility with future applications. For users who need a capable GPU in a portable device, the 945M is the logical pick, while desktop users seeking maximum compute performance should favor the GTX 650 Ti Boost.

Specification Differences

The two GPUs differ across nearly every major specification category. The 945M uses the GM107 chip on a 28 nm process with 1,870 million transistors on a 148 mm² die, while the GTX 650 Ti Boost uses the GK106 chip with 2,540 million transistors on a 221 mm² die. The 945M operates at a base clock of 928 MHz with a boost of 1020 MHz, while the GTX 650 Ti Boost runs at 980 MHz base and 1032 MHz boost. Memory specifications diverge significantly: the 945M has 2 GB of DDR3 on a 128-bit bus with 28.80 GB/s bandwidth and 1800 Mbps effective memory clock, whereas the GTX 650 Ti Boost has 2 GB of GDDR5 on a 192-bit bus with 144.2 GB/s bandwidth and 6 Gbps effective memory clock.

Compute resources differ, with the 945M featuring 640 shading units, 40 TMUs, and 16 ROPs, while the GTX 650 Ti Boost has 768 shading units, 64 TMUs, and 24 ROPs. This results in different fill rates: the 945M has a texture rate of 40.80 GTexel/s and a pixel rate of 16.32 GPixel/s, while the GTX 650 Ti Boost has a texture rate of 66.05 GTexel/s and a pixel rate of 16.51 GPixel/s. FP32 performance is 1,305.6 GFLOPS for the 945M and 1.585 TFLOPS for the GTX 650 Ti Boost. Power requirements are starkly different, with the 945M at 75 W TDP and no power connectors, versus the GTX 650 Ti Boost at 134 W TDP with a 6-pin connector and 300 W suggested PSU. The bus interfaces are MXM-B (3.0) for the 945M and PCIe 3.0 x16 for the GTX 650 Ti Boost, with the latter being a dual-slot, 241 mm card. API support also differs in Vulkan version (1.4 for 945M, 1.2.175 for GTX 650 Ti Boost), though both share DirectX 12 (11_0) and OpenGL 4.6. The GTX 650 Ti Boost carries a launch MSRP of 169 USD.