NVIDIA GeForce GTX 460 v2 vs NVIDIA GeForce GTX 650 Ti Comparison

The NVIDIA GeForce GTX 460 v2 and the NVIDIA GeForce GTX 650 Ti represent two distinct approaches to the mid-range graphics card of their respective generations. The GTX 460 v2, built on Fermi 2.0, is the older, larger, and more power-hungry card, while the GTX 650 Ti, built on Kepler, is the newer, smaller, and more efficient option. Benchmark data from the Geekbench OpenCL test shows the GTX 460 v2 holds a clear performance lead, scoring 8743 against the GTX 650 Ti's 7877. This 11% delta places the older card in a higher percentile ranking (44th vs. 42nd), although the GTX 650 Ti counters with modern API support and a significantly more compact physical footprint.

FAQ

Q: Which GPU is faster in the Geekbench OpenCL benchmark?

A: The NVIDIA GeForce GTX 460 v2 is faster, scoring 8743 compared to the GTX 650 Ti's 7877. This represents an 11% advantage for the GTX 460 v2 in the head-to-head comparison.

Q: How does the GTX 650 Ti's average benchmark score compare to its nearest rivals?

A: The GTX 650 Ti has an average score of 8053. It sits within a tight pack, being 0.2% ahead of the GeForce GTX 880M and Quadro P5000, and 0.2% behind the GTX 650 Ti Boost. It is 0.3% behind the GRID K2.

Q: What are the key architectural differences between the two cards?

A: The GTX 460 v2 uses the GF114 chip based on the Fermi 2.0 architecture, manufactured on a 40 nm process at TSMC. The GTX 650 Ti uses the GK106S chip based on the Kepler architecture, manufactured on a 28 nm process, also at TSMC.

Q: Which card has a higher transistor density, and why does it matter?

A: The GTX 650 Ti has a much higher transistor density at 11.5M / mm², compared to the GTX 460 v2's 5.9M / mm². This is due to the newer 28 nm process node, allowing the 650 Ti to pack 2,540 million transistors into a smaller 221 mm² die, whereas the 460 v2 has 1,950 million transistors on a larger 332 mm² die.

Q: What is the difference in power requirements between the two cards?

A: The GTX 460 v2 has a TDP of 160 W and requires a 450 W power supply with two 6-pin power connectors. The GTX 650 Ti is more efficient, with a 110 W TDP, a 300 W power supply recommendation, and a single 6-pin connector.

Q: Does the GTX 650 Ti support any APIs that the GTX 460 v2 does not?

A: Yes, the GTX 650 Ti supports Vulkan 1.2.175, while the GTX 460 v2 has no listed Vulkan support. Both cards support DirectX 12 (11_0) and OpenGL 4.6.

Where Each One Wins

The benchmark data clearly favors the NVIDIA GeForce GTX 460 v2 in raw compute performance. Its Geekbench OpenCL score of 8743 is decisively higher than the GTX 650 Ti's 7877. This 11% lead suggests that in applications leveraging OpenCL for general-purpose computing, rendering, or physics simulation, the older Fermi card is the stronger performer. Its higher memory bandwidth of 96.19 GB/s, despite a slower clock, and larger 192-bit bus, also gives it an edge in memory-intensive workloads compared to the 650 Ti's 86.40 GB/s over a 128-bit bus.

The NVIDIA GeForce GTX 650 Ti, however, wins on efficiency and physical design. With a TDP of just 110 W, it consumes significantly less power than the 460 v2's 160 W. This translates to a single-slot design and a shorter 145 mm (5.7 inches) length, making it far easier to install in small form factor cases or for users with a 300 W power supply. Its modern Kepler architecture also provides access to Vulkan 1.2.175, a feature completely absent from the GTX 460 v2, which may be a deciding factor for users running specific Linux drivers or newer emulators.

Architecture Differences

The two cards are separated by a full architectural generation. The GTX 460 v2 is built on the Fermi 2.0 architecture with the GF114 chip, while the GTX 650 Ti uses the Kepler architecture with the GK106S chip. This generational shift is visible in almost every design decision.

The manufacturing process is a primary differentiator. The GTX 460 v2 uses a 40 nm TSMC node, resulting in a large 332 mm² die with 1,950 million transistors. The GTX 650 Ti uses a 28 nm TSMC node, which is a significant shrink. This allows the 650 Ti to fit 2,540 million transistors onto a much smaller 221 mm² die, achieving a transistor density of 11.5M / mm² compared to the 460 v2's 5.9M / mm².

The compute core layouts are also fundamentally different. The GTX 460 v2 has 336 shading units, 56 TMUs, and 24 ROPs. In contrast, the GTX 650 Ti has a much higher count of 768 shading units and 64 TMUs, but a lower count of 16 ROPs. This configuration suggests the Kepler card is designed for higher arithmetic throughput, while the Fermi card holds an advantage in pixel processing and memory operations due to its larger ROP count and wider memory bus. These architectural choices explain why the 650 Ti has a higher theoretical FP32 rate of 1,425.4 GFLOPS and texture rate of 59.39 GTexel/s, yet the 460 v2 scores higher in the OpenCL benchmark.

Specification Differences

The specification sheets reveal a direct contrast in nearly every measurable category. The most significant difference is in the process node and die size: the GTX 460 v2 is a 40 nm chip with a 332 mm² die, while the GTX 650 Ti is a 28 nm chip with a 221 mm² die.

Memory subsystems differ in bandwidth and bus width. The GTX 460 v2 offers a 192-bit bus with 96.19 GB/s bandwidth, while the GTX 650 Ti has a narrower 128-bit bus with 86.40 GB/s bandwidth. Although the 650 Ti has a higher effective memory clock of 5.4 Gbps versus 4 Gbps, the wider bus on the 460 v2 yields higher total bandwidth.

Power and physical requirements are starkly different. The GTX 460 v2 is a dual-slot card, 210 mm (8.3 inches) long, with a 160 W TDP, requiring a 450 W PSU and two 6-pin connectors. The GTX 650 Ti is a single-slot card, 145 mm (5.7 inches) long, with a 110 W TDP, requiring only a 300 W PSU and a single 6-pin connector.

The bus interface also differs, with the GTX 460 v2 using PCIe 2.0 x16 and the GTX 650 Ti using the newer PCIe 3.0 x16. Display outputs are similar, but the 650 Ti supports HDMI 1.4a while the 460 v2 supports HDMI 1.3a. Both have 1024 MB of GDDR5 memory, but the launch MSRP differed, with the GTX 460 v2 at 199 USD and the GTX 650 Ti at 149 USD.

Head-to-Head Benchmarks

The only direct benchmark comparison available is the Geekbench OpenCL test, and it provides a clear verdict. The NVIDIA GeForce GTX 460 v2 wins this test with a score of 8743, against the GTX 650 Ti's 7877. The deltaPct is 11%, indicating that the GTX 460 v2 is 11% faster in this specific OpenCL workload.

This result is surprising given the architectural differences. The GTX 650 Ti has more than double the shading units (768 vs. 336) and a higher FP32 compute rate (1,425.4 GFLOPS vs. 1,046.3 GFLOPS). However, the GTX 460 v2's superior memory bandwidth (96.19 GB/s vs. 86.40 GB/s) and higher ROP count (24 vs. 16) likely contribute to its performance advantage in this test. The 460 v2's pixel rate is 10.91 GPixel/s, which is lower than the 650 Ti's 14.85 GPixel/s, but its texture rate of 43.62 GTexel/s is also lower than the 650 Ti's 59.39 GTexel/s. Despite these theoretical disadvantages, the real-world OpenCL score favors the older card.

The GTX 650 Ti's average benchmark score of 8053, which includes a Geekbench Vulkan score of 8229, places it in the 42nd percentile. The GTX 460 v2, with only the OpenCL score of 8743, sits in the 44th percentile. While the 650 Ti has a second benchmark result, its average is dragged down by the lower OpenCL score. The data indicates that for users prioritizing OpenCL compute performance, the GTX 460 v2 is the clear choice, while the GTX 650 Ti offers the benefit of Vulkan support and a more efficient design.