NVIDIA GeForce GTX TITAN vs NVIDIA P106-090 Comparison

Launched in February 2013, the NVIDIA GeForce GTX TITAN is a fully featured Kepler flagship with 6 GB of memory, while the NVIDIA P106-090 is a Pascal-based mining card from 2017 with 3 GB and no display outputs. Benchmark data shows these two cards are close in average score—the GTX TITAN averages 14,373 versus 13,470 for the P106-090—but they achieve parity through completely different architectural strengths. The GTX TITAN wins the OpenCL test with a 16.8% margin, while the P106-090 dominates the Vulkan test by 46.1%, making the choice between them entirely dependent on which API and workload matters more.

FAQ

Q: Which card has the higher average benchmark score?

A: The NVIDIA GeForce GTX TITAN averages 14,373 across all tested workloads, placing it in the 56th percentile of all GPUs. The NVIDIA P106-090 averages 13,470, which lands in the 54th percentile, a difference of roughly 903 points.

Q: How do the two cards compare in OpenCL performance?

A: The GTX TITAN scores 24,873 in Geekbench OpenCL, while the P106-090 scores 21,304. This gives the GTX TITAN a 16.8% lead in that test, representing one of its two head-to-head benchmark wins.

Q: Which card wins in Vulkan, and by how much?

A: The P106-090 wins the Geekbench Vulkan test decisively, scoring 18,596 against the GTX TITAN's 10,027. The delta is 46.1% in favor of the Pascal mining card, meaning it nearly doubles the TITAN's Vulkan output.

Q: What is the memory configuration difference?

A: The GTX TITAN comes with 6 GB of GDDR5 on a 384-bit bus, delivering 288.4 GB/s of bandwidth. The P106-090 has 3 GB of GDDR5 on a 192-bit bus, providing 192.2 GB/s, which is about two-thirds the bandwidth of the TITAN.

Q: Are these cards still in production?

A: Both are End-of-life. The GTX TITAN was released on February 18, 2013, and the P106-090 followed on July 30, 2017. Neither has a successor listed in the data.

Q: Does the P106-090 support display outputs?

A: No. The P106-090 has no display outputs at all, making it unsuitable for standard desktop use. The GTX TITAN, by contrast, offers 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2.

Architecture Differences

The GTX TITAN is built on the GK110 chip using the Kepler architecture, manufactured on a 28 nm process at TSMC. It packs 7,080 million transistors onto a 561 mm² die, yielding a transistor density of 12.6M per mm². In contrast, the P106-090 uses the GP106 chip in the Pascal architecture, produced on a more advanced 16 nm process. Its die is only 200 mm² yet holds 4,400 million transistors, giving it a much higher density of 22.0M per mm². This means the P106-090 achieves nearly twice the transistor density of the older Kepler design.

The compute resources differ sharply. The GTX TITAN has 2,688 shading units, 224 texture mapping units, and 48 ROPs. The P106-090 has just 768 shading units and 48 TMUs, but also 48 ROPs. Despite having fewer than a third of the shading units, the P106-090 compensates with much higher clocks: its base is 1,354 MHz and boost reaches 1,531 MHz, compared to the GTX TITAN's 836 MHz base and 876 MHz boost. The Pascal card's boost clock is 75% higher than the TITAN's.

Memory architecture is another major split. The GTX TITAN runs 6 GB of GDDR5 at 1,502 MHz (6 Gbps effective) over a 384-bit bus, producing 288.4 GB/s. The P106-090 halves the bus to 192-bit but uses faster 2,002 MHz GDDR5 (8 Gbps effective), still resulting in lower total bandwidth at 192.2 GB/s. The TITAN also has a much larger frame buffer, which matters for high-resolution textures and multi-monitor setups.

The P106-090's API support is newer: it lists DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The GTX TITAN supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.2.175. Both cards lack ray tracing and tensor cores. The P106-090 also has a 1:64 FP16 ratio, delivering 36.74 GFLOPS, while the TITAN does not list FP16 performance. Finally, the P106-090 is a mining-specific product with no video outputs, whereas the TITAN has a full complement of legacy display connectors.

Head-to-Head Benchmarks

Only two shared benchmarks exist in the data, and they split the victory. In Geekbench OpenCL, the GTX TITAN scores 24,873 against the P106-090's 21,304. That is a 16.8% margin, driven by the TITAN's massive 4.709 TFLOPS FP32 throughput versus 2.352 TFLOPS on the P106-090. The TITAN also has vastly more texture fill rate—196.2 GTexel/s compared to 73.49 GTexel/s—and more than double the shading units.

The Vulkan test flips the result dramatically. The P106-090 achieves 18,596, while the GTX TITAN manages only 10,027. This 46.1% delta is the largest performance gap in either direction. The Pascal architecture's superior Vulkan driver implementation and newer API feature level (Vulkan 1.4 versus 1.2.175) likely explain this, though the raw numbers show the P106-090 simply executes Vulkan workloads far more efficiently. Notably, the P106-090's pixel rate is 73.49 GPixel/s, which is higher than the TITAN's 49.06 GPixel/s despite the TITAN's overall compute advantage.

The average benchmark scores reflect this split. The GTX TITAN sits at 14,373, just 0.1% below the AMD Radeon RX Vega 11 (14,385) and 0.1% above the AMD Radeon Vega 11 (14,352). The P106-090's 13,470 places it 0.3% below the NVIDIA GeForce GTX 570 (13,515) and 0.5% above the AMD Radeon Pro 555 (13,407). In direct comparison, the TITAN leads by 6.7% in average score, but the P106-090's Vulkan advantage is far larger than that aggregate suggests.

Specification Differences

The two cards diverge on nearly every specification. The GTX TITAN uses the GK110 chip on 28 nm, while the P106-090 uses GP106 on 16 nm. Transistor count is 7,080 million versus 4,400 million, and die size is 561 mm² versus 200 mm². Transistor density is 12.6M/mm² on the TITAN and 22.0M/mm² on the P106-090.

Clock speeds are dramatically different: the TITAN runs at 836 MHz base and 876 MHz boost, while the P106-090 runs at 1,354 MHz base and 1,531 MHz boost. Memory clocks differ as well—1,502 MHz (6 Gbps) on the TITAN versus 2,002 MHz (8 Gbps) on the P106-090. Memory size is 6 GB versus 3 GB, bus width is 384-bit versus 192-bit, and bandwidth is 288.4 GB/s versus 192.2 GB/s.

Compute units show the TITAN's advantage: 2,688 shading units and 224 TMUs versus 768 shading units and 48 TMUs on the P106-090. Both have 48 ROPs. Pixel rate is 49.06 GPixel/s on the TITAN and 73.49 GPixel/s on the P106-090, while texture rate is 196.2 GTexel/s versus 73.49 GTexel/s. FP32 throughput is 4.709 TFLOPS versus 2.352 TFLOPS. The P106-090 lists FP16 at 36.74 GFLOPS (1:64), which the TITAN does not report.

Power and physical specs differ significantly. The TITAN has a 250 W TDP, requires a 600 W power supply, and needs 1x 6-pin plus 1x 8-pin connectors. The P106-090 has a 75 W TDP, a 250 W suggested PSU, and only needs 1x 6-pin. The TITAN is 267 mm long, 111 mm tall, and 38 mm wide; the P106-090 is 250 mm long with no listed height or width. Both are dual-slot cards. The TITAN uses PCIe 3.0 x16, while the P106-090 uses PCIe 1.0 x1. Display outputs are 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2 on the TITAN, versus no outputs on the P106-090. API support is also different, with the TITAN at DirectX 12 (11_0) and Vulkan 1.2.175, and the P106-090 at DirectX 12 (12_1) and Vulkan 1.4.

Where Each One Wins

The GTX TITAN wins in raw compute-bound workloads. Its OpenCL score of 24,873 is 16.8% higher, and its FP32 throughput of 4.709 TFLOPS is exactly double that of the P106-090. For tasks like scientific computing, rendering, or any OpenCL-accelerated application, the TITAN is the stronger card. Its 6 GB frame buffer and 288.4 GB/s bandwidth also make it suitable for large datasets or high-resolution textures. The TITAN's nearest rivals—the AMD Radeon RX Vega 11 at 14,385 and the NVIDIA GeForce GTX 965M at 14,404—are all within 0.2% of its score, indicating it sits at the top of its performance tier.

The P106-090 wins decisively in Vulkan. Its 18,596 score versus 10,027 represents a 46.1% lead, and its Vulkan 1.4 support is newer than the TITAN's 1.2.175. For modern games or applications that use Vulkan, the P106-090 is clearly faster despite having a third of the shading units. Its higher pixel rate (73.49 GPixel/s versus 49.06) suggests it also excels at fill-rate-heavy tasks. Additionally, its 75 W TDP and single 6-pin connector mean it can run in systems with much smaller power supplies—250 W versus 600 W—making it the better choice for low-power or mining-oriented rigs.

The P106-090's nearest rival list includes the NVIDIA GeForce GTX 570 at 13,515 and the AMD Radeon HD 7770M at 13,536, both within 0.5% of its average score. This puts it in a lower overall performance class, but its Vulkan dominance shows that API-specific performance can overturn aggregate rankings.

The Verdict

Pick the NVIDIA GeForce GTX TITAN if your workload is OpenCL-heavy or requires large memory capacity. It leads by 16.8% in OpenCL, offers 6 GB of memory versus 3 GB, and provides double the FP32 throughput. Its 250 W TDP and 600 W PSU requirement are the cost of that performance, and its 56th percentile ranking is slightly higher than the P106-090's 54th. The TITAN also has display outputs, making it usable as a standard desktop GPU, and its nearest rivals are all within 0.4% of its score, confirming it sits at the top of its peer group.

Pick the NVIDIA P106-090 if Vulkan performance or power efficiency is the priority. It wins the Vulkan test by 46.1%, a margin large enough to overcome its lower average score in that specific API. Its 75 W TDP is a third of the TITAN's, and it needs only a 250 W PSU and a single 6-pin connector. However, it has no display outputs, so it is strictly a compute or mining card. Its PCIe 1.0 x1 interface is also a severe bottleneck in most systems, but the data shows it still delivers competitive Vulkan numbers. For anyone running Vulkan-native applications, the P106-090 is the better choice; for everything else, the GTX TITAN's broader compute and memory advantage makes it the safer pick.