GPU Comparison

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA P106-100 records a higher average benchmark score of 23249, compared to the NVIDIA GeForce GTX 780's 19164. This places the P106-100 in the 68th percentile of all GPUs, while the GTX 780 sits in the 64th percentile.

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

A: The P106-100 scores 35951 in Geekbench OpenCL, which is 57.2% higher than the GTX 780's 22863. The P106-100 wins this test decisively.

Q: What is the difference in memory capacity and bus width?

A: The P106-100 has 6 GB of GDDR5 memory on a 192-bit bus, while the GTX 780 has 3 GB of GDDR5 memory on a 384-bit bus. Despite the narrower bus, the P106-100's memory runs at a higher effective speed of 8 Gbps versus 6 Gbps.

Q: Which card has a higher transistor count and larger die size?

A: The GTX 780 uses the GK110 chip with 7,080 million transistors on a 561 mm² die. The P106-100 uses the GP106 chip with 4,400 million transistors on a 200 mm² die. The GTX 780 is physically larger in both respects.

Q: What are the power requirements for each card?

A: The P106-100 has a TDP of 120 W and requires a single 6-pin power connector with a suggested PSU of 300 W. The GTX 780 has a TDP of 250 W, needs one 6-pin and one 8-pin connector, and suggests a 600 W PSU.

Q: Which card supports more modern APIs?

A: The P106-100 supports DirectX 12 (12_1) and Vulkan 1.4, while the GTX 780 supports DirectX 12 (11_0) and Vulkan 1.2.175. Both cards support OpenGL 4.6.

Architecture Differences

The two NVIDIA cards come from different architectural generations and target different use cases. The P106-100 is built on the Pascal architecture using the GP106 chip, manufactured on a 16 nm process at TSMC. The GTX 780 uses the older Kepler architecture with the GK110 chip, also from TSMC but on a 28 nm process. This process difference is substantial: the P106-100 packs 4,400 million transistors into a 200 mm² die, achieving a transistor density of 22.0 million per mm². The GTX 780 crams 7,080 million transistors into a much larger 561 mm² die, resulting in a lower density of 12.6 million per mm².

The GTX 780 has more raw processing resources: 2304 shading units and 192 texture mapping units, versus 1280 shading units and 80 TMUs on the P106-100. Both cards have 48 ROPs. However, the P106-100 compensates with much higher clock speeds. Its base clock is 1506 MHz and boost clock is 1709 MHz, while the GTX 780 runs at 863 MHz base and 902 MHz boost. This clock advantage helps the P106-100 deliver a slightly higher FP32 throughput of 4.375 TFLOPS versus 4.156 TFLOPS for the GTX 780.

Memory configurations also differ significantly. The P106-100 has 6 GB of GDDR5 on a 192-bit bus with memory clocked at 2002 MHz (8 Gbps effective), yielding 192.2 GB/s of bandwidth. The GTX 780 has 3 GB of GDDR5 on a wider 384-bit bus with memory at 1502 MHz (6 Gbps effective), delivering a higher 288.4 GB/s bandwidth. The GTX 780's wider bus gives it a bandwidth advantage despite the slower memory clock.

The P106-100 is a mining-oriented card with no display outputs, while the GTX 780 offers 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2. The P106-100 also uses a PCIe 1.0 x16 interface, whereas the GTX 780 uses PCIe 3.0 x16. Power characteristics differ sharply: the P106-100 draws 120 W with a 300 W suggested PSU, while the GTX 780 draws 250 W and suggests a 600 W PSU. The GTX 780 also has a launch MSRP of 649 USD.

The Verdict

The recorded data points to the NVIDIA P106-100 as the stronger performer in the benchmark suite. It wins both head-to-head tests, with a 57.2% margin in Geekbench OpenCL and a 34.2% margin in Geekbench Vulkan. Its average benchmark score of 23249 is roughly 21% higher than the GTX 780's 19164, and its 68th percentile ranking sits four points above the GTX 780's 64th.

For users who prioritize compute workloads such as OpenCL and Vulkan tasks, the P106-100 is the clear choice. It delivers higher raw throughput, more memory capacity, and lower power consumption. The lack of display outputs, however, means it cannot serve as a primary graphics card for a desktop workstation. The GTX 780, despite lower scores, offers full display connectivity and remains a functional option for legacy systems that need output capabilities.

For anyone building a system around a card that can render to a screen, the GTX 780 is the only option of the two. For compute-focused applications where display output is handled by another GPU or is unnecessary, the P106-100's benchmark results make it the superior pick. The data does not support choosing the GTX 780 for raw performance; its only advantages lie in memory bandwidth, display outputs, and PCIe interface generation.

Specification Differences

The two cards differ across nearly every major specification category. The P106-100 uses the GP106 chip on a 16 nm Pascal architecture, while the GTX 780 uses the GK110 chip on a 28 nm Kepler architecture. Transistor counts are 4,400 million versus 7,080 million, and die sizes are 200 mm² versus 561 mm², giving transistor densities of 22.0M per mm² and 12.6M per mm² respectively.

Clock speeds favor the P106-100: base clock of 1506 MHz versus 863 MHz, boost clock of 1709 MHz versus 902 MHz, and memory clock of 2002 MHz (8 Gbps effective) versus 1502 MHz (6 Gbps effective). The P106-100 has 1280 shading units, 80 TMUs, and 48 ROPs, while the GTX 780 has 2304 shading units, 192 TMUs, and 48 ROPs. Pixel rates are 82.03 GPixel/s for the P106-100 and 43.30 GPixel/s for the GTX 780, while texture rates are 136.7 GTexel/s and 173.2 GTexel/s respectively. FP32 throughput is 4.375 TFLOPS versus 4.156 TFLOPS, and the P106-100 also lists FP16 at 68.36 GFLOPS (1:64) while the GTX 780 has no FP16 entry.

Memory specs differ: 6 GB versus 3 GB, both GDDR5, but bus widths are 192-bit versus 384-bit, and bandwidths are 192.2 GB/s versus 288.4 GB/s. Power profiles diverge with TDPs of 120 W and 250 W, power connectors of 1x 6-pin versus 1x 6-pin + 1x 8-pin, and suggested PSUs of 300 W and 600 W. The P106-100 uses PCIe 1.0 x16 while the GTX 780 uses PCIe 3.0 x16. Display outputs are absent on the P106-100 but present on the GTX 780. API support shows DirectX 12 (12_1) versus DirectX 12 (11_0), OpenGL 4.6 on both, and Vulkan 1.4 versus 1.2.175. The GTX 780 also has specified dimensions of 267 mm length, 111 mm height, and 38 mm width, while the P106-100 lists only a 250 mm length.

Head-to-Head Benchmarks

The head-to-head benchmark data contains two recorded tests, and the P106-100 wins both. In Geekbench OpenCL, the P106-100 scores 35951 against the GTX 780's 22863, a delta of 57.2%. This is the larger margin of the two tests and shows a significant advantage in general-purpose compute workloads. In Geekbench Vulkan, the P106-100 scores 32897 versus 24514, a 34.2% advantage. Both wins are substantial, though the OpenCL gap is roughly 23 percentage points wider than the Vulkan gap.

The average benchmark scores reinforce this pattern. The P106-100 averages 23249 across its benchmark suite, while the GTX 780 averages 19164. This represents an approximate 21% overall advantage for the P106-100. The P106-100's nearest rivals by average score are the AMD Radeon Pro Vega 16 at 23250 (0% delta), the AMD Radeon RX 6600M at 23273 (-0.1%), the AMD Radeon R9 M290X at 23276 (-0.1%), and the AMD Radeon AI PRO R9700 at 23315 (-0.3%). The GTX 780's nearest rivals include the NVIDIA TITAN Xp at 19177 (-0.1%), the NVIDIA Tesla K20m at 19089 (0.4%), the NVIDIA GeForce RTX 4050 Mobile at 19049 (0.6%), and the AMD Radeon RX 6600 at 19036 (0.7%). These proximity data points show that each card sits in a distinct performance tier, with the P106-100 roughly one tier above the GTX 780.

The GTX 780 does not win any recorded benchmark against the P106-100. Its higher memory bandwidth of 288.4 GB/s does not translate into a benchmark victory in the available data. The P106-100's clock speed advantage appears to outweigh the GTX 780's wider memory bus and higher shading unit count in these specific workloads.

Where Each One Wins

The P106-100 wins in compute-oriented workloads as measured by Geekbench OpenCL and Vulkan. Its 57.2% OpenCL lead and 34.2% Vulkan lead make it the stronger choice for applications that rely on these APIs, such as GPU-accelerated computation, rendering tasks, and Vulkan-based games. The card's 6 GB memory capacity is double that of the GTX 780, which helps in workloads that need more video memory. Its lower TDP of 120 W and single 6-pin connector also make it easier to integrate into systems with modest power delivery, requiring only a 300 W PSU.

The GTX 780 wins in areas not covered by the benchmark suite but present in the specification data. It offers display outputs, including 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2, making it the only card of the two that can output video to a monitor. Its memory bandwidth of 288.4 GB/s is roughly 50% higher than the P106-100's 192.2 GB/s, which could benefit memory-bandwidth-sensitive workloads even though the recorded benchmarks do not show it winning. The GTX 780 also has more shading units (2304 versus 1280) and TMUs (192 versus 80), which may help in scenarios where those resources scale well, though the benchmark data does not confirm this. Its PCIe 3.0 x16 interface is newer than the P106-100's PCIe 1.0 x16, which matters for compatibility with modern motherboards.

In practical terms, the P106-100 is the pick for compute-focused systems where display output is not needed. The GTX 780 is the pick for desktop systems that require a working video output and can tolerate higher power consumption. The recorded data gives the P106-100 two benchmark wins and a higher average score, but the GTX 780 retains the connectivity and interface advantages that make it the more flexible general-purpose card.