NVIDIA GeForce GTX 650 Ti vs NVIDIA GeForce GTX 850M Comparison

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX 850M records an average benchmark score of 9302, while the NVIDIA GeForce GTX 650 Ti scores 8053. The 850M sits at the 46th percentile among all GPUs, and the 650 Ti sits at the 42nd percentile.

Q: How much faster is the GTX 850M in OpenCL workloads?

A: The GTX 850M achieves a Geekbench OpenCL score of 9821 compared to 7877 for the 650 Ti, a 24.7% advantage. In the database, the 850M is 0.1% ahead of the NVIDIA GeForce GTX 465, while the 650 Ti is 0.2% ahead of the GTX 650 Ti Boost.

Q: Does the GTX 650 Ti offer Vulkan support?

A: Yes, both GPUs support Vulkan. The GTX 850M scores 8782 in Geekbench Vulkan, and the 650 Ti scores 8229. The 850M wins by 6.7% in this test.

Q: Which card has more shading units?

A: The GTX 650 Ti has 768 shading units, while the GTX 850M has 640. However, the 850M's memory runs at 1001 MHz (2 Gbps effective) compared to the 650 Ti's 1350 MHz (5.4 Gbps effective).

Q: What memory configuration does each GPU use?

A: The GTX 850M uses 2 GB of DDR3 memory on a 128-bit bus with 32.03 GB/s bandwidth. The GTX 650 Ti uses 1024 MB of GDDR5 memory on a 128-bit bus with 86.40 GB/s bandwidth.

Q: Which GPU is newer?

A: The GTX 850M was released on 2014-03-11, while the GTX 650 Ti was released on 2012-10-08. Both are end-of-life products.

Where Each One Wins

The benchmark data splits cleanly between these two GPUs. The GTX 850M wins both recorded tests, but the margin tells a different story depending on the workload type.

In Geekbench OpenCL, the 850M holds a decisive 24.7% lead over the 650 Ti. This is a compute-oriented workload, and the Maxwell architecture's efficiency shows clearly. The 850M's score of 9821 places it near the GTX 465 (9294, within 0.1%) and slightly above the AMD Radeon R7 M380 (9313, 0.1% behind) and GTX 960 (9273, 0.3% ahead). The 650 Ti's 7877 score sits among the GTX 880M (8040, 0.2% ahead) and GTX 650 Ti Boost (8067, 0.2% behind), a noticeably lower performance tier.

In Geekbench Vulkan, the gap narrows considerably. The 850M scores 8782 versus 8229 for the 650 Ti, a 6.7% win. Vulkan exercises driver overhead and command processing more than raw compute throughput, so the 650 Ti closes some ground here. Still, the 850M remains ahead in both API paths.

The 650 Ti has no benchmark wins in this comparison. Its strength lies in memory bandwidth (86.40 GB/s versus 32.03 GB/s) and raw fill rates, but the recorded Geekbench results do not favor it. The 850M wins both head-to-head tests, making it the default choice for compute-heavy tasks.

Architecture Differences

The GTX 850M uses the GM107 chip built on NVIDIA's Maxwell architecture, while the GTX 650 Ti uses the GK106S chip on the older Kepler architecture. Both are fabricated on TSMC's 28 nm process, but the designs diverge significantly.

The Maxwell chip packs 1,870 million transistors into a 148 mm² die, yielding a transistor density of 12.6 million per mm². The Kepler chip is physically larger at 221 mm² with 2,540 million transistors, but its density is lower at 11.5 million per mm². Maxwell achieves higher density on the same process node, a sign of more efficient layout and design.

The 850M belongs to the GeForce 800M generation, following the GeForce 700M series and preceding the GeForce 900M. The 650 Ti belongs to the GeForce 600 generation, following GeForce 500 and preceding GeForce 700. This generational gap explains the architectural shift: Maxwell was designed to improve performance per watt, which matters for a mobile part like the 850M.

The 850M is an integrated-style GPU with an IGP slot width and no power connectors, drawing 45 W. The 650 Ti is a discrete single-slot card requiring a 1x 6-pin power connector with a 110 W TDP and a suggested 300 W power supply. The physical form factors reflect their intended platforms: the 850M for laptops, the 650 Ti for desktop towers.

API support differs slightly. Both support DirectX 12 (11_0) and OpenGL 4.6. The 850M supports Vulkan 1.4, while the 650 Ti supports Vulkan 1.2.175. The mobile part carries a newer Vulkan revision, consistent with its later release date.

Specification Differences

The GTX 850M has 640 shading units, 40 texture mapping units, and 16 ROPs. The GTX 650 Ti has 768 shading units, 64 TMUs, and 16 ROPs. The 650 Ti leads in raw shader count and texture hardware, but the 850M compensates elsewhere.

Memory configurations differ sharply. The 850M uses 2 GB of DDR3 at 1001 MHz (2 Gbps effective) with a 128-bit bus, producing 32.03 GB/s bandwidth. The 650 Ti uses 1024 MB of GDDR5 at 1350 MHz (5.4 Gbps effective) with the same 128-bit bus, producing 86.40 GB/s. The 650 Ti has nearly 2.7 times the memory bandwidth of the 850M.

Pixel rates are similar: 14.43 GPixel/s for the 850M versus 14.85 GPixel/s for the 650 Ti. Texture rates diverge more: 36.08 GTexel/s for the 850M versus 59.39 GTexel/s for the 650 Ti. FP32 compute favors the 650 Ti at 1,425.4 GFLOPS versus 1,154.6 GFLOPS for the 850M.

Power draw is the most striking difference. The 850M consumes 45 W with no external power connectors, while the 650 Ti draws 110 W with a 6-pin connector. The 650 Ti also has a suggested 300 W power supply requirement. The 850M's display outputs are listed as portable device dependent, while the 650 Ti offers 2x DVI and 1x mini-HDMI 1.4a.

The 650 Ti has a physical length of 145 mm (5.7 inches). The 850M has no listed dimensions. The 650 Ti has a launch MSRP of 149 USD. The 850M has no launch MSRP recorded.

Head-to-Head Benchmarks

The Geekbench OpenCL test produces the largest gap between these two GPUs. The GTX 850M scores 9821 against 7877 for the GTX 650 Ti, a 24.7% difference. This is the 850M's strongest result, positioning it alongside desktop GPUs like the GTX 465 (9294) and GTX 960 (9273) rather than mobile parts. The 650 Ti's 7877 score places it with the GTX 880M (8040), a high-end mobile GPU from the same era, and the GTX 650 Ti Boost (8067), its own slightly improved sibling.

The Geekbench Vulkan test narrows the gap. The 850M scores 8782, and the 650 Ti scores 8229, a 6.7% margin. Vulkan results often depend on driver maturity and API implementation; the 650 Ti's older Kepler architecture still holds up reasonably well here. The 850M's Vulkan 1.4 support versus the 650 Ti's 1.2.175 may contribute to the difference.

The head-to-head record shows 2 wins for the 850M and 0 for the 650 Ti. However, the 650 Ti's hardware specifications suggest it should compete better in bandwidth-sensitive tasks. Its 86.40 GB/s memory bandwidth is a major advantage over the 850M's 32.03 GB/s, but the Geekbench workloads do not appear to favor that strength. The 850M's Maxwell architecture, with its higher transistor density and newer design, delivers better results in both recorded tests.

The average benchmark scores reinforce the pattern: 9302 for the 850M versus 8053 for the 650 Ti. The 850M sits at the 46th percentile, while the 650 Ti sits at the 42nd. In the nearest rivals table, the 850M trades places with the GTX 465 (0.1% difference) and the AMD Radeon R7 M380 (0.1% behind), while the 650 Ti sits within 0.3% of the GTX 880M, GTX 650 Ti Boost, and Quadro P5000.

The Verdict

The data points clearly to the GTX 850M as the better performer in the recorded benchmarks. It wins both Geekbench tests, holds a higher average score (9302 versus 8053), and ranks higher at the 46th percentile versus the 42nd. For compute workloads as measured by OpenCL and Vulkan, the 850M is the superior GPU.

The 850M achieves this with dramatically lower power consumption: 45 W versus 110 W. It requires no power connectors and fits an IGP form factor, making it suitable for portable devices. The 650 Ti demands a 6-pin connector and a 300 W power supply, limiting it to desktop builds. For mobile users or anyone prioritizing efficiency, the 850M is the clear choice.

The 650 Ti does have advantages in raw specification: more shading units (768 versus 640), more TMUs (64 versus 40), higher texture rate (59.39 GTexel/s versus 36.08), higher FP32 throughput (1,425.4 GFLOPS versus 1,154.6), and nearly 2.7 times the memory bandwidth. These specs suggest the 650 Ti could outperform the 850M in bandwidth-heavy or texture-heavy scenarios, but no recorded benchmark confirms this. The Geekbench suite does not include such workloads.

Buyers who need a mobile GPU or a low-power solution should pick the GTX 850M. Its benchmark lead and efficiency make it the stronger product overall. Desktop users who already have a 6-pin power connector and a 300 W power supply might consider the GTX 650 Ti for its higher memory bandwidth and larger shader count, but the recorded data does not show it winning any test. The 850M is the verdict from the measurements.