NVIDIA GeForce GTX 960 vs NVIDIA GeForce GTX TITAN BLACK Comparison

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GTX TITAN BLACK leads with an average benchmark score of 9474, while the NVIDIA GeForce GTX 960 trails at 9273. That is a difference of roughly 2.2%.

Q: How do the two cards compare in Geekbench Vulkan performance?

A: The TITAN BLACK scores 29561 in Geekbench Vulkan, which is 220.2% higher than the GTX 960’s 9231. This is the largest single-test gap between the two cards.

Q: Which card has the higher memory bandwidth?

A: The TITAN BLACK offers 336.0 GB/s of memory bandwidth thanks to a 384-bit bus and 6 GB of GDDR5 memory. The GTX 960 provides 112.2 GB/s over a 128-bit bus with 2 GB of GDDR5.

Q: What is the TDP difference between these two GPUs?

A: The TITAN BLACK has a TDP of 250 W, while the GTX 960 draws only 120 W. Consequently, the suggested power supply for the TITAN BLACK is 600 W versus 300 W for the GTX 960.

Q: Do both cards support DirectX 12?

A: Yes, but at different feature levels. The TITAN BLACK supports DirectX 12 (11_1), while the GTX 960 supports DirectX 12 (12_1), meaning the GTX 960 has a higher feature level.

Q: Which card has better rasterization throughput?

A: The TITAN BLACK achieves a pixel rate of 58.80 GPixel/s and a texture rate of 235.2 GTexel/s. The GTX 960 delivers 37.70 GPixel/s and 75.39 GTexel/s, so the TITAN BLACK is substantially ahead in both metrics.

The Verdict

The data is unambiguous: the NVIDIA GeForce GTX TITAN BLACK wins all three head-to-head benchmark comparisons against the GTX 960, taking 3 wins to 0. In Geekbench Metal, the TITAN BLACK scores 24172 versus 8773, a 175.5% advantage. In OpenCL, it leads 27218 to 18925, a 43.8% margin. In Vulkan, the gap widens to 220.2% (29561 vs 9231).

The TITAN BLACK is the clear choice for users prioritizing raw compute and graphics throughput. Its 2880 shading units, 240 TMUs, and 48 ROPs dwarf the GTX 960’s 1024 shading units, 64 TMUs, and 32 ROPs. FP32 performance is 5.645 TFLOPS versus 2.413 TFLOPS. Memory capacity and bandwidth also favor the TITAN BLACK: 6 GB at 336.0 GB/s versus 2 GB at 112.2 GB/s.

However, the GTX 960 is not without merit. It delivers 43.8% of the TITAN BLACK’s FP32 throughput while consuming only 120 W versus 250 W. Its transistor density is actually slightly higher (12.9M / mm² vs 12.6M / mm²), reflecting the more efficient Maxwell 2.0 architecture. The GTX 960 also supports DirectX 12 (12_1) versus the TITAN BLACK’s DirectX 12 (11_1), and it offers newer display outputs including HDMI 2.0 and three DisplayPort 1.2 connectors.

In terms of percentile ranking, the TITAN BLACK sits at the 46th percentile of all GPUs, while the GTX 960 sits at the 45th. Their average scores (9474 vs 9273) place them in nearly the same performance tier, despite the TITAN BLACK’s massive advantage in specific compute workloads. The GTX 960’s nearest rivals include the GTX 465 (9294, -0.2%), GTX 850M (9302, -0.3%), and Radeon R7 M380 (9313, -0.4%), while the TITAN BLACK’s nearest rivals include the Quadro K5000 (9637, -1.7%) and GTX 960M (9645, -1.8%).

For users needing maximum compute performance, the TITAN BLACK is the obvious pick. For those prioritizing power efficiency and modern API support, the GTX 960 is the more sensible option. Neither card is in production; both are end-of-life products.

Head-to-Head Benchmarks

The TITAN BLACK dominates every shared benchmark. In Geekbench Metal, it scores 24172 against the GTX 960’s 8773, representing a 175.5% lead. This massive margin indicates that the TITAN BLACK’s Kepler architecture, with its 2880 shading units, is far better suited to Metal compute workloads than the GTX 960’s 1024 shading units.

Geekbench OpenCL shows a narrower but still decisive gap. The TITAN BLACK posts 27218, which is 43.8% ahead of the GTX 960’s 18925. This suggests that OpenCL performance scales better with the GTX 960’s Maxwell design, but the TITAN BLACK’s raw shading resources still carry it to victory.

Geekbench Vulkan is where the TITAN BLACK truly separates itself. Its score of 29561 is 220.2% higher than the GTX 960’s 9231. This is the largest delta in any test, indicating that Vulkan compute workloads heavily favor the TITAN BLACK’s larger GPU configuration.

The GTX 960 has no benchmark wins in the head-to-head comparisons. However, it does have one unique benchmark entry not shared with the TITAN BLACK: 3DMark Steel Nomad DX12, where it scores 162. The TITAN BLACK has additional PassMark scores (DirectX 9: 133, DirectX 10: 47, DirectX 11: 73, DirectX 12: 34, G2D: 595, G3D: 9184, GPU Compute: 3726) that are not present for the GTX 960.

The average benchmark scores tell a slightly different story than the head-to-head results. The TITAN BLACK averages 9474, while the GTX 960 averages 9273. Both cards land within 2% of each other in aggregate, yet the TITAN BLACK’s wins in the shared tests are consistently large. This implies that the GTX 960 performs relatively better in other benchmark suites not covered by the head-to-head data.

For the GTX 960, its nearest rival is the GTX 465 at 9294 (-0.2%), followed by the GTX 850M at 9302 (-0.3%) and Radeon R7 M380 at 9313 (-0.4%). The TITAN BLACK’s nearest rival is the Quadro K5000 at 9637 (-1.7%), with the GTX 960M close behind at 9645 (-1.8%). These numbers show that both GPUs sit in a crowded mid-range performance band, despite their architectural differences.

Specification Differences

The two GPUs differ substantially in nearly every hardware specification. The TITAN BLACK uses the GK110B chip, while the GTX 960 uses GM206. Transistor counts diverge sharply: 7,080 million for the TITAN BLACK versus 2,940 million for the GTX 960. Die size also differs, at 561 mm² versus 228 mm². Interestingly, transistor density is nearly identical: 12.6M / mm² for the TITAN BLACK and 12.9M / mm² for the GTX 960.

Clock speeds favor the GTX 960. It runs at a base clock of 1127 MHz and a boost of 1178 MHz, while the TITAN BLACK operates at 889 MHz base and 980 MHz boost. Memory clocks are essentially the same: 1750 MHz (7 Gbps effective) for the TITAN BLACK and 1753 MHz (7 Gbps effective) for the GTX 960.

Memory configuration is a major differentiator. The TITAN BLACK has 6 GB of GDDR5 on a 384-bit bus, yielding 336.0 GB/s bandwidth. The GTX 960 has 2 GB on a 128-bit bus, yielding 112.2 GB/s. This means the TITAN BLACK offers triple the memory capacity and triple the bandwidth.

Pixel and texture throughput favor the TITAN BLACK overwhelmingly. It achieves 58.80 GPixel/s and 235.2 GTexel/s, versus 37.70 GPixel/s and 75.39 GTexel/s for the GTX 960. FP32 compute is 5.645 TFLOPS versus 2.413 TFLOPS.

Power draw differs significantly. The TITAN BLACK is rated at 250 W with a suggested PSU of 600 W, using one 6-pin and one 8-pin connector. The GTX 960 draws 120 W with a suggested PSU of 300 W, using a single 6-pin connector. Both are dual-slot cards.

Physical dimensions differ as well. The TITAN BLACK measures 267 mm (10.5 inches) in length, 111 mm (4.4 inches) in height, and 38 mm (1.5 inches) in width. The GTX 960 is shorter at 241 mm (9.5 inches), with height and width not specified.

Display outputs differ. The TITAN BLACK offers 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2. The GTX 960 offers 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2, providing more modern connectivity options.

Architecture Differences

The TITAN BLACK is built on Kepler architecture, while the GTX 960 uses Maxwell 2.0. Both are manufactured on TSMC’s 28 nm process node, but they represent different design philosophies. Kepler, as implemented in the GK110B chip, is a large-die design with massive shading resources: 2880 shading units, 240 TMUs, and 48 ROPs. Maxwell 2.0 in the GM206 chip is more compact, with 1024 shading units, 64 TMUs, and 32 ROPs.

The memory subsystem reflects these architectural differences. The TITAN BLACK’s 384-bit bus is typical of high-end Kepler parts, while the GTX 960’s 128-bit bus is characteristic of the mid-range Maxwell lineup. The TITAN BLACK’s 6 GB frame buffer is triple the GTX 960’s 2 GB, making it better suited for high-resolution textures and compute workloads.

API support reveals generational differences. The TITAN BLACK supports DirectX 12 (11_1), whereas the GTX 960 supports DirectX 12 (12_1). Both support OpenGL 4.6. Vulkan support differs: the TITAN BLACK supports Vulkan 1.2.175, while the GTX 960 supports Vulkan 1.4. The GTX 960’s higher DirectX feature level and newer Vulkan version reflect its Maxwell 2.0 architecture’s more modern feature set.

The GTX 960 is part of the GeForce 900 generation, succeeding the GeForce 700 series and preceding the GeForce 10 series. The TITAN BLACK is part of the GeForce 700 generation, succeeding the GeForce 600 and preceding the GeForce 900. This places the GTX 960 one generation ahead of the TITAN BLACK in NVIDIA’s product timeline.

Power efficiency is a key architectural differentiator. The GTX 960 delivers 2.413 TFLOPS at 120 W, while the TITAN BLACK delivers 5.645 TFLOPS at 250 W. Per watt, the GTX 960 is more efficient (20.1 GFLOPS/W vs 22.6 GFLOPS/W), though the TITAN BLACK still provides 2.3 times the absolute FP32 performance.

The TITAN BLACK’s transistor density of 12.6M / mm² versus the GTX 960’s 12.9M / mm² shows that Maxwell 2.0 packs slightly more transistors per square millimeter, despite the GTX 960 having far fewer total transistors. This efficiency gain is a hallmark of the Maxwell architecture’s design improvements over Kepler.

Both cards share a PCIe 3.0 x16 bus interface. Neither has dedicated ray tracing or tensor cores, as those features were introduced in later NVIDIA architectures.