AMD Radeon R5 M320 vs NVIDIA GeForce 840M Comparison

Head-to-Head Benchmarks

The recorded data from the database shows a clear, consistent margin for the NVIDIA GeForce 840M across both benchmark tests. In the Geekbench OpenCL test, the 840M scores 5764, while the AMD Radeon R5 M320 trails at 5051. That is a 14.1% advantage for the NVIDIA part. The Vulkan test tells a similar story: the 840M posts 4880 against the R5 M320's 4262, a 14.5% lead. In both instances, the NVIDIA GPU wins outright, and the database records 2 wins for the 840M and 0 for the AMD part.

What stands out is the consistency of the gap. A 14% delta in compute-oriented OpenCL and a 14.5% delta in Vulkan suggest the 840M's advantage is not limited to one API or workload type. It is a broad, across-the-board edge. The average benchmark score reinforces this: the 840M sits at 5322, which is 14.3% above the R5 M320's 4657. The percentile ranking also separates them, with the 840M in the 31st percentile of all GPUs versus the R5 M320 in the 27th. That is a modest but real difference in overall standing.

Looking at the nearest rivals, the 840M's average score of 5322 places it just above the GeForce 930A (5317, 0.1% ahead) and the GTX 980M (5308, 0.3% ahead), while trailing the Radeon R7 M445 (5358) by 0.7%. The R5 M320's 4657 average matches the Radeon RX 9060 XT 16 GB exactly (0% delta), and it sits between the Quadro P400 (4684, 0.6% behind) and the GeForce GTX 970M (4628, 0.6% ahead). So while neither card is a top-tier performer, the 840M lands in a slightly stronger neighborhood, and its head-to-head record against the R5 M320 is unambiguous.

The data does not show any benchmark where the R5 M320 takes the lead. Its best recorded score, 5051 in OpenCL, is still 713 points behind the 840M's 5764. The Vulkan result, 4262, is 618 points short. This is not a close contest in raw performance terms.

Architecture Differences

The two GPUs come from different architectural families, and the database highlights several fundamental differences. The NVIDIA GeForce 840M is built on the Maxwell architecture, specifically the GM108S chip, while the AMD Radeon R5 M320 uses the GCN 1.0 architecture with the Jet chip. Both are manufactured on a 28 nm process at TSMC, so the process node is identical, but the chip designs diverge sharply.

Transistor counts reflect this. The 840M packs 1,020 million transistors on a 77 mm² die, giving a transistor density of 13.2 million per mm². The R5 M320 uses 690 million transistors on a 56 mm² die, with a density of 12.3 million per mm². The NVIDIA chip is physically larger and denser, which partially explains its performance edge.

The compute resources also differ. The 840M has 384 shading units, 16 texture mapping units, and 8 ROPs. The R5 M320 has 320 shading units but 20 TMUs and 8 ROPs. So NVIDIA fields more shaders, while AMD configures more texture units. The pixel rate and texture rate reflect this: the 840M delivers 8.992 GPixel/s and 17.98 GTexel/s, while the R5 M320 manages 6.840 GPixel/s and 17.10 GTexel/s. The 840M leads in both, though the texture gap is smaller (17.98 vs 17.10) than the pixel gap (8.992 vs 6.840).

Clock speeds are another differentiator. The 840M runs at a base clock of 1029 MHz with a boost up to 1124 MHz. The R5 M320 operates at 780 MHz base and 855 MHz boost. That is a substantial clock advantage for the NVIDIA part, and it shows up in the FP32 compute figures: the 840M delivers 863.2 GFLOPS, while the R5 M320 manages only 547.2 GFLOPS. That is a 57.8% compute advantage for the 840M, a larger gap than the benchmark scores suggest, likely because real-world workloads do not scale perfectly with raw GFLOPS.

Memory configurations are similar in bandwidth but different in capacity. Both use DDR3 with a 64-bit bus. The 840M has 2 GB and a bandwidth of 16.02 GB/s, while the R5 M320 has 4 GB and 16.00 GB/s. The effective memory clock is 2 Gbps on both. The R5 M320's extra capacity (4 GB vs 2 GB) does not translate into a bandwidth advantage, but it could matter for workloads that need more frame buffer space.

API support shows a small difference. The 840M supports DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4. The R5 M320 supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The NVIDIA part has a newer Vulkan version, while AMD's DirectX 12 feature level is one notch higher. Both are end-of-life products, with the 840M released in March 2014 and the R5 M320 in May 2015.

The power profile is notable: the 840M has a TDP of 33 W, while the R5 M320 has no listed TDP in the database. Both are IGP (integrated graphics processor) form factors with no power connectors, and both use a PCIe 3.0 x8 bus interface. Display outputs are portable-device dependent for both.

The Verdict

The benchmark data points to a single conclusion: the NVIDIA GeForce 840M is the stronger GPU in every measured category. It wins both head-to-head tests, has a higher average score, and sits in a higher percentile. For anyone choosing between these two, the 840M is the clear performance pick.

The 14.1% OpenCL lead and 14.5% Vulkan lead are substantial, and the 840M's FP32 compute (863.2 GFLOPS vs 547.2 GFLOPS) shows a deeper architectural advantage that could matter for compute-heavy tasks. The 840M also clocks higher (1029 MHz base vs 780 MHz base) and has more shading units (384 vs 320), which explains its consistent edge.

The R5 M320 does have one thing going for it: double the memory capacity (4 GB vs 2 GB). In scenarios where frame buffer size matters more than raw throughput, the AMD part could hold an advantage. But the bandwidth is virtually identical (16.00 GB/s vs 16.02 GB/s), so the extra capacity does not come with extra speed.

The data does not support choosing the R5 M320 for performance. Its 27th percentile ranking versus the 840M's 31st, its lower average score, and its zero wins in head-to-head tests all point the same way. The 840M is the superior part, and the numbers are not close enough to argue otherwise.

Specification Differences

The database records several specification differences between the two GPUs. These are the fields where the two parts diverge:

• Chip: NVIDIA GM108S vs AMD Jet
• Architecture: NVIDIA Maxwell vs AMD GCN 1.0
• Transistors: 1,020 million vs 690 million
• Die size: 77 mm² vs 56 mm²
• Transistor density: 13.2M / mm² vs 12.3M / mm²
• Base clock: 1029 MHz vs 780 MHz
• Boost clock: 1124 MHz vs 855 MHz
• Memory size: 2 GB vs 4 GB
• Memory bandwidth: 16.02 GB/s vs 16.00 GB/s
• Shading units: 384 vs 320
• TMUs: 16 vs 20
• Pixel rate: 8.992 GPixel/s vs 6.840 GPixel/s
• Texture rate: 17.98 GTexel/s vs 17.10 GTexel/s
• FP32: 863.2 GFLOPS vs 547.2 GFLOPS
• TDP: 33 W vs not listed
• Vulkan version: 1.4 vs 1.2.170
• DirectX version: 12 (11_0) vs 12 (11_1)
• Release date: March 2014 vs May 2015
• Predecessor: GeForce 700M vs Solar System
• Successor: GeForce 900M vs Polaris Mobile

The ROP count is identical at 8, and both use DDR3 memory with a 64-bit bus and 2 Gbps effective clock. Power connectors are absent on both, and both are IGP form factors with PCIe 3.0 x8.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce 840M has an average score of 5322, while the AMD Radeon R5 M320 averages 4657. That is a 14.3% difference in favor of the NVIDIA part.

Q: How much faster is the 840M in the OpenCL test?

A: The 840M scores 5764 in Geekbench OpenCL, while the R5 M320 scores 5051. The 840M leads by 14.1%.

Q: Does the R5 M320 win any benchmark test?

A: No. The database records 2 wins for the 840M and 0 for the R5 M320 across both the OpenCL and Vulkan tests.

Q: What is the memory capacity difference?

A: The R5 M320 has 4 GB of DDR3 memory, while the 840M has 2 GB. However, the bandwidth is nearly identical: 16.00 GB/s for the AMD part versus 16.02 GB/s for the NVIDIA part.

Q: Which GPU has more shading units?

A: The NVIDIA GeForce 840M has 384 shading units, compared to 320 for the AMD Radeon R5 M320.

Q: What are the clock speed differences?

A: The 840M has a base clock of 1029 MHz and a boost clock of 1124 MHz. The R5 M320 runs at 780 MHz base and 855 MHz boost.

Where Each One Wins

The data paints a lopsided picture, but there are specific areas where each GPU holds an advantage.

The NVIDIA GeForce 840M wins in every benchmark category the database records. It is 14.1% ahead in OpenCL and 14.5% ahead in Vulkan. Its compute throughput is far higher, with 863.2 GFLOPS of FP32 performance versus 547.2 GFLOPS for the AMD part. It also has a higher pixel rate (8.992 GPixel/s vs 6.840 GPixel/s) and texture rate (17.98 GTexel/s vs 17.10 GTexel/s). The 840M's clock advantage, 1029 MHz base versus 780 MHz base, is a major factor in these results. Its shading unit count of 384 versus 320 gives it more parallel execution capacity. The 840M also supports a newer Vulkan version (1.4 vs 1.2.170), which could matter for future software compatibility.

The AMD Radeon R5 M320 has far fewer wins in the data, but it does have one clear specification advantage: memory capacity. With 4 GB of DDR3 versus the 840M's 2 GB, the R5 M320 can hold larger datasets in the frame buffer. For workloads that are memory-capacity-bound, such as high-resolution textures or certain compute tasks, that extra capacity could be useful. The R5 M320 also has more texture mapping units (20 vs 16), though its lower clock speed means the texture rate still falls behind. Its DirectX feature level is slightly higher (11_1 vs 11_0), which could enable certain features in older DirectX 11 titles.

The R5 M320 also has a lower transistor density (12.3M / mm² vs 13.2M / mm²) and a smaller die (56 mm² vs 77 mm²), which suggests it may be a simpler, less power-hungry part, though the database does not list a TDP for it.

In practical terms, the 840M is the clear winner for compute-heavy tasks, gaming, and any workload that stresses raw GPU throughput. The R5 M320's 4 GB frame buffer could be its only saving grace for specific memory-hungry scenarios, but the bandwidth is essentially equal, so that capacity advantage is limited.

For most users, the choice is straightforward: the data favors the NVIDIA GeForce 840M. The R5 M320 is a viable option only if the larger memory pool is an absolute requirement, and even then, the performance gap is significant. The database records 2 wins for the 840M, 0 for the R5 M320, and the average scores and percentiles both confirm the same story.