NVIDIA GeForce GTX 950 vs NVIDIA GeForce GTX TITAN Comparison

FAQ

Q: Which GPU has the higher average benchmark score in the database?

A: The NVIDIA GeForce GTX TITAN records an average benchmark score of 14373, while the NVIDIA GeForce GTX 950 scores 13189. This places the GTX TITAN at the 56th percentile among all GPUs, compared to the GTX 950’s 53rd percentile.

Q: How do the two cards compare in the Geekbench Metal test?

A: The GTX TITAN wins the Metal test with 8218 points versus 7172 for the GTX 950, a lead of 14.6%. This is one of the two benchmark wins for the GTX TITAN in the head-to-head comparison.

Q: Which GPU performs better in the Vulkan benchmark?

A: The GTX 950 takes the Vulkan test decisively, scoring 16734 against the GTX TITAN’s 10027. The delta is 40.1% in favor of the GTX 950, marking its only win in the three-test comparison.

Q: What is the difference in FP32 floating-point performance?

A: The GTX TITAN delivers 4.709 TFLOPS of FP32 compute, while the GTX 950 reaches 1.825 TFLOPS. The TITAN’s shading unit count of 2688 versus 768 for the GTX 950 explains much of this gap.

Q: Do both cards use the same manufacturing process?

A: Yes, both are built on a 28 nm node at TSMC. However, the GTX TITAN packs 7,080 million transistors on a 561 mm² die, while the GTX 950 uses 2,940 million transistors on a 228 mm² die. The transistor densities are close, at 12.6M / mm² for the TITAN and 12.9M / mm² for the GTX 950.

Q: Which card has the higher memory bandwidth?

A: The GTX TITAN has a 384-bit memory bus with 288.4 GB/s of bandwidth, while the GTX 950 uses a 128-bit bus with 105.8 GB/s. The TITAN also has 6 GB of GDDR5 memory versus 2 GB on the GTX 950.

The Verdict

The data points to a split decision based on workload. The NVIDIA GeForce GTX TITAN wins two of the three recorded head-to-head benchmarks, with a commanding 58.8% lead in OpenCL and a 14.6% lead in Metal. Its average score of 14373 sits at the 56th percentile, and its nearest rivals are all within 0.4% in either direction, indicating a tightly clustered performance tier. The GTX 950, meanwhile, posts an average of 13189 at the 53rd percentile, with rivals ranging from 0.8% above to 1.2% below.

For users prioritizing compute-heavy OpenCL workloads or general GPU compute through Metal, the GTX TITAN is the clear choice from the recorded data. Its 2688 shading units, 224 texture mapping units, and 384-bit memory bus provide the raw throughput that drives its higher scores. For Vulkan-based applications, however, the GTX 950 is the stronger option, outperforming the TITAN by over 40%. The GTX 950 also carries a significantly lower power draw, with a 90 W TDP versus 250 W for the TITAN, and requires a 250 W suggested PSU instead of 600 W.

The GTX TITAN is an end-of-life product from the GeForce 700 generation, released in February 2013, while the GTX 950 belongs to the GeForce 900 series, launched in August 2015. The TITAN’s launch MSRP was 999 USD, and the GTX 950’s was 159 USD. Neither card is in current production. The verdict: choose the GTX TITAN for OpenCL and Metal compute tasks, and choose the GTX 950 for Vulkan workloads or lower system power requirements.

Head-to-Head Benchmarks

The three recorded benchmarks show a clear split in strengths. In Geekbench Metal, the GTX TITAN scores 8218 against 7172 for the GTX 950, a 14.6% advantage. This is consistent with the TITAN’s larger compute resources, including 2688 shading units and a 49.06 GPixel/s pixel rate versus 38.02 GPixel/s for the GTX 950.

In Geekbench OpenCL, the margin widens dramatically. The GTX TITAN posts 24873 points, while the GTX 950 manages 15662. That is a 58.8% difference, the largest gap in any test. The TITAN’s texture rate of 196.2 GTexel/s versus 57.02 GTexel/s for the GTX 950, along with its 4.709 TFLOPS FP32 throughput versus 1.825 TFLOPS, explains this overwhelming result. The memory subsystem also favors the TITAN, with 288.4 GB/s of bandwidth compared to 105.8 GB/s.

The Vulkan test reverses the outcome. The GTX 950 scores 16734, while the GTX TITAN records 10027. This is a 40.1% win for the GTX 950, and it is the only test where the newer card prevails. The GTX 950’s architecture, Maxwell 2.0 with DirectX 12 (12_1) support and Vulkan 1.4 API compatibility, appears better suited to this workload than the TITAN’s Kepler design, which supports DirectX 12 (11_0) and Vulkan 1.2.175.

The aggregate result is 2 wins for the GTX TITAN and 1 win for the GTX 950. However, the magnitude of the OpenCL victory (58.8%) is substantially larger than the Vulkan defeat (40.1%), which is why the TITAN’s average benchmark score of 14373 sits above the GTX 950’s 13189 despite losing the Vulkan test.

Specification Differences

The two cards diverge on nearly every major specification. The GTX TITAN has 2688 shading units, 224 TMUs, and 48 ROPs, while the GTX 950 has 768 shading units, 48 TMUs, and 32 ROPs. Clock speeds also differ: the TITAN runs at a base of 836 MHz and boost of 876 MHz, whereas the GTX 950 runs at 1024 MHz base and 1188 MHz boost. Despite the higher clocks, the GTX 950’s smaller core count leaves it far behind in aggregate throughput.

Memory configurations are starkly different. The TITAN offers 6 GB of GDDR5 on a 384-bit bus with 288.4 GB/s bandwidth, while the GTX 950 has 2 GB on a 128-bit bus with 105.8 GB/s. Memory clock rates also differ, with the TITAN at 1502 MHz (6 Gbps effective) and the GTX 950 at 1653 MHz (6.6 Gbps effective). The TITAN’s wider bus compensates for its lower clock speed.

Power and physical dimensions favor the GTX 950. The TITAN has a 250 W TDP with a 600 W suggested PSU and requires both a 6-pin and an 8-pin power connector. The GTX 950 has a 90 W TDP, a 250 W suggested PSU, and only a single 6-pin connector. The TITAN measures 267 mm in length, 111 mm in height, and 38 mm in width; the GTX 950 is shorter at 202 mm, with no recorded height or width data. Both are dual-slot cards.

Display outputs differ as well. The TITAN provides 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2. The GTX 950 provides 1x DVI, 1x HDMI 2.0, and 3x DisplayPort 1.2. API support also differs, with the TITAN at DirectX 12 (11_0) and Vulkan 1.2.175, versus the GTX 950 at DirectX 12 (12_1) and Vulkan 1.4.

Architecture Differences

The GTX TITAN is built on the GK110 chip using the Kepler architecture, part of the GeForce 700 generation. The GTX 950 uses the GM206 chip with Maxwell 2.0, part of the GeForce 900 generation. Both are fabricated at TSMC on a 28 nm process, but the transistor counts diverge sharply: 7,080 million for the TITAN versus 2,940 million for the GTX 950. The die sizes are 561 mm² and 228 mm² respectively, yielding nearly identical transistor densities of 12.6M / mm² and 12.9M / mm².

The Kepler architecture in the TITAN was designed for high compute throughput, as evidenced by its 2688 shading units and 4.709 TFLOPS FP32 performance. Maxwell 2.0 in the GTX 950 emphasizes efficiency and modern API features, with 768 shading units and 1.825 TFLOPS. The GTX 950 supports DirectX 12 (12_1) and Vulkan 1.4, while the TITAN is limited to DirectX 12 (11_0) and Vulkan 1.2.175. This API gap is directly reflected in the Vulkan benchmark results, where the GTX 950 leads by 40.1%.

The TITAN’s memory architecture is older but wider: 384-bit bus versus 128-bit, 288.4 GB/s versus 105.8 GB/s. The GTX 950 compensates with a higher effective memory clock (6.6 Gbps versus 6 Gbps), but the bandwidth gap remains enormous. The TITAN’s pixel rate of 49.06 GPixel/s and texture rate of 196.2 GTexel/s dwarf the GTX 950’s 38.02 GPixel/s and 57.02 GTexel/s.

Where Each One Wins

The GTX TITAN wins in OpenCL and Metal workloads. In OpenCL, its 58.8% advantage over the GTX 950 is the largest margin in the entire comparison. The TITAN’s 4.709 TFLOPS FP32 throughput, 2688 shading units, and 288.4 GB/s memory bandwidth make it the stronger choice for compute-heavy tasks such as rendering, scientific simulations, or any workload that leverages OpenCL. In Metal, the 14.6% lead reinforces the TITAN’s compute advantage, though the margin is smaller.

The GTX 950 wins in Vulkan workloads. Its 40.1% lead in the Vulkan benchmark is substantial and likely stems from its more modern Maxwell 2.0 architecture with Vulkan 1.4 support, compared to the TITAN’s Vulkan 1.2.175. For gaming or applications that use Vulkan, the GTX 950 is the better performer according to the recorded data.

The GTX 950 also wins on power efficiency and system requirements. Its 90 W TDP versus 250 W, and its 250 W suggested PSU versus 600 W, make it far easier to integrate into a modest system. The single 6-pin power connector simplifies installation compared to the TITAN’s 6-pin plus 8-pin requirement.

For users who need maximum compute performance in OpenCL or Metal, the GTX TITAN is the clear winner despite its age and higher power draw. For users targeting Vulkan-based applications or building a lower-power system, the GTX 950 is the appropriate choice. The recorded data shows no scenario where both cards are equally suited; each has a distinct domain of superiority.