NVIDIA GeForce 840M vs NVIDIA Quadro P400 Comparison

The NVIDIA GeForce 840M and the NVIDIA Quadro P400 occupy the same low-power corner of the GPU landscape, but they arrive there from very different directions. The 840M is a mobile Maxwell part from the GeForce 800M generation, an integrated-positioning chip with no power connector and a 33 W TDP. The Quadro P400 is a Pascal-based professional card from the Quadro Px000 line, a single-slot desktop board drawing 30 W. Both are end-of-life, both carry 2 GB of memory, and both sit in the lower third of the database's percentile rankings: the 840M at the 31st percentile, the P400 at the 27th. Yet the recorded benchmark data splits their head-to-head evenly, one win each, which makes this pairing more interesting than the raw specs suggest.

Head-to-Head Benchmarks

The database contains two direct comparisons between these GPUs, and they point in opposite directions.

In Geekbench OpenCL, the GeForce 840M scores 5764 against the Quadro P400's 4249, a 35.7 percent advantage for the older Maxwell mobile chip. That is not a marginal gap. For general-purpose compute workloads routed through OpenCL, the 840M's higher shading unit count and higher theoretical FP32 throughput (863.2 GFLOPS versus 641.0 GFLOPS) translate into a decisive win, and the benchmark result confirms the arithmetic rather than contradicting it.

In Geekbench Vulkan, the picture flips. The Quadro P400 scores 5119 against the 840M's 4880, a 4.7 percent win for the Pascal card. It is a narrow margin, nothing like the OpenCL blowout, but it is consistent with the P400's more modern API feature level and its GDDR5 memory subsystem. Vulkan's lower-level rendering model tends to reward memory bandwidth and driver-side efficiency, and the P400's 32.06 GB/s of bandwidth is exactly double the 840M's 16.02 GB/s.

The resulting tally is one win each. Averaging across both tests, the 840M posts an average benchmark score of 5322 against the P400's 4684, which is why the 840M ranks four percentile points higher against the full database. The context around each card is telling, though. The 840M's nearest rivals cluster almost impossibly tightly: the GeForce 930A sits 0.1 percent behind it, the GTX 980M-rated average sits 0.3 percent behind, the Radeon R7 M445 is 0.7 percent ahead, and the GeForce 940M is 0.7 percent behind. The P400's neighborhood is similarly compressed, with the Radeon RX 9060 XT 16 GB and R5 M320 both 0.6 percent behind and the R8 M445DX 0.9 percent ahead. Both GPUs sit in dense scoring bands where small differences barely separate neighboring entries.

Architecture Differences

The architectural gulf between these two is wide despite their similar performance envelopes.

The GeForce 840M is built on the GM108S chip, Maxwell architecture, on TSMC's 28 nm process. It packs 1,020 million transistors into a 77 mm² die, a density of 13.2 million transistors per square millimeter. The Quadro P400 uses the GP107 chip, Pascal architecture, manufactured by Samsung on a 14 nm node. Its die is larger at 132 mm² but holds 3,300 million transistors, nearly 25.0 million per square millimeter. The process node jump from 28 nm to 14 nm roughly doubles transistor density while enabling a much larger transistor budget overall.

The resource split is lopsided in curious ways. The 840M has more shading units, 384 against 256, which explains its higher FP32 throughput. Both chips have 16 TMUs. The P400 doubles the 840M's ROP count, 16 against 8, which shows up directly in pixel rate: 20.03 GPixel/s for the P400 versus 8.992 GPixel/s for the 840M, more than a two-to-one advantage for the Quadro. Texture rate also favors the P400, 20.03 GTexel/s against 17.98 GTexel/s, a reflection of its higher boost clock of 1252 MHz versus 1124 MHz.

Memory is where the platforms diverge most sharply. Both use a 64-bit bus and 2 GB of capacity, but the 840M relies on DDR3 running at 1001 MHz (2 Gbps effective) for 16.02 GB/s of bandwidth, while the P400 uses GDDR5 at 1002 MHz (4 Gbps effective) for 32.06 GB/s. Same bus width, double the bandwidth. The P400 also carries an FP16 capability at 10.02 GFLOPS (1:64), a capability the 840M's recorded data does not list.

Feature support differs at the API level. The P400 supports DirectX 12 at feature level 12_1, while the 840M tops out at 12 (11_0). Both support OpenGL 4.6 and Vulkan 1.4. The P400 offers three mini-DisplayPort 1.4a outputs and a full PCIe 3.0 x16 interface; the 840M connects over PCIe 3.0 x8 and has portable-device-dependent display output, as befits a mobile part. The P400 is a discrete single-slot card with physical dimensions recorded at 150 mm long and 69 mm high, and a suggested PSU of 200 W. Neither card requires a power connector.

Where Each One Wins

The data splits cleanly along workload lines.

The 840M wins compute. Its 35.7 percent Geekbench OpenCL lead is the single largest margin in this comparison, and it aligns with the hardware: 384 shading units and 863.2 GFLOPS of FP32 give it substantially more raw number-crunching throughput than the P400's 256 units and 641.0 GFLOPS. Any OpenCL-based workload in the database's records favors the 840M decisively.

The P400 wins modern graphics rendering paths. Its Vulkan victory of 4.7 percent is modest, but the underlying hardware points the same way more strongly than the score does: double the memory bandwidth, double the ROPs, more than double the pixel rate, and a newer DirectX feature level. For display-driving scenarios its three dedicated mini-DisplayPort 1.4a outputs matter too, since the 840M's output configuration depends entirely on the host device.

Power is nearly a wash. The 840M is rated at 33 W as an IGP-class mobile part; the P400 at 30 W as a single-slot desktop card. Neither needs auxiliary power.

FAQ

Q: Which GPU is faster overall in the recorded benchmarks?

A: They split the two head-to-head tests. The 840M wins Geekbench OpenCL by 35.7 percent (5764 vs 4249), while the P400 wins Geekbench Vulkan by 4.7 percent (5119 vs 4880). On average score the 840M leads, 5322 vs 4684.

Q: How do they rank against all GPUs in the database?

A: The 840M sits at the 31st percentile and the P400 at the 27th, so both occupy the lower third, with the 840M slightly ahead.

Q: Do they have the same amount of memory?

A: Yes, both have 2 GB on a 64-bit bus, but the types differ. The 840M uses DDR3 at 16.02 GB/s, while the P400 uses GDDR5 at 32.06 GB/s.

Q: Which has the newer architecture and manufacturing process?

A: The P400. It is Pascal on Samsung's 14 nm node with 3,300 million transistors, while the 840M is Maxwell on TSMC's 28 nm node with 1,020 million transistors.

Q: Do either of these cards need a power connector?

A: No. The 840M draws its 33 W as an IGP-class part, and the P400's 30 W fits within its PCIe slot, with a suggested PSU of 200 W.

Q: Which supports the newer DirectX feature level?

A: The P400, at DirectX 12 (12_1), compared with the 840M's DirectX 12 (11_0). Both support OpenGL 4.6 and Vulkan 1.4.

Specification Differences

• Chip: GM108S (840M) vs GP107 (P400)
• Architecture: Maxwell vs Pascal
• Generation: GeForce 800M vs Quadro Pascal (Px000)
• Process node: 28 nm TSMC vs 14 nm Samsung
• Transistors / die size: 1,020 million on 77 mm² vs 3,300 million on 132 mm²
• Transistor density: 13.2M/mm² vs 25.0M/mm²
• Clocks: 1029 MHz base / 1124 MHz boost vs 1228 MHz base / 1252 MHz boost
• Memory type / bandwidth: DDR3, 16.02 GB/s vs GDDR5, 32.06 GB/s
• Shading units: 384 vs 256
• ROPs: 8 vs 16
• Pixel rate: 8.992 GPixel/s vs 20.03 GPixel/s
• Texture rate: 17.98 GTexel/s vs 20.03 GTexel/s
• FP32: 863.2 GFLOPS vs 641.0 GFLOPS
• FP16: not listed vs 10.02 GFLOPS (1:64)
• TDP: 33 W vs 30 W
• Form factor: IGP vs single-slot card, 150 mm by 69 mm
• Bus: PCIe 3.0 x8 vs PCIe 3.0 x16
• Display outputs: portable device dependent vs 3x mini-DisplayPort 1.4a
• DirectX feature level: 12 (11_0) vs 12 (12_1)
• Release date: March 2014 vs February 2017
• Predecessor / successor: GeForce 700M / 900M vs Quadro Maxwell / Quadro Volta

The Verdict

The data frames this as a compute-versus-graphics split rather than a clean winner.

Pick the GeForce 840M if OpenCL throughput is the priority. Its 35.7 percent win in the OpenCL test is backed by more shading units and higher FP32 performance, and its higher average score and 31st percentile placement reflect that. It is the stronger compute part in this pairing, full stop.

Pick the Quadro P400 if the workload runs through modern graphics APIs or needs display connectivity. Its Vulkan win, double the memory bandwidth, double the ROPs, more than double the pixel rate, newer DirectX feature level, and three dedicated DisplayPort outputs make it the more complete rendering and display platform, particularly in a desktop where its single-slot, connector-free 30 W design drops in easily.

Notably, the P400's superior silicon does not convert into superior measured performance across the board. Three years newer, on a denser process, with far more transistors, it still loses the compute benchmark by a wide margin because the 840M simply has more shading resources. That is the central lesson of this comparison: architecture generation sets the feature set and efficiency, but the resource counts decide who wins a given benchmark. Both cards are end-of-life, both are modest performers in absolute percentile terms, and the right choice depends entirely on whether the workload leans on OpenCL or on Vulkan-class rendering.