NVIDIA GeForce 930M vs NVIDIA GeForce 940MX Comparison

Head-to-Head Benchmarks

The GeForce 940MX and GeForce 930M are both Maxwell-based mobile parts from the GeForce 900M generation, but the benchmark data reveals a clear split personality between the two. In the two recorded tests, each card takes one win, and the margins tell the real story.

The 930M claims victory in Geekbench OpenCL with a score of 5046 against the 940MX's 4939, a delta of 2.1% in favor of the older chip. That is a narrow margin, well within the range of driver-level variation or thermal state differences in a mobile chassis. The 940MX is effectively at parity here, sitting just 107 points behind. The database's nearest-rival data supports this: the 940MX's average score of 4844 places it within 0.2% of the GeForce GTX 560M (4855) and 0.5% of the AMD Radeon R6 M255DX (4867), while the 930M's average of 4388 sits 0.5% below the GeForce GT 645M (4411). Both cards are firmly in entry-level mobile territory, but the 940MX's OpenCL result is less impressive than its hardware specifications would suggest.

The Vulkan test is where the 940MX separates itself decisively. It scores 4749 against the 930M's 3729, a 27.4% advantage. That is a substantial gap, and it aligns with the architectural differences between the two chips. The 940MX's higher shading unit count and faster memory subsystem give it a clear edge in API-bound workloads that exercise geometry throughput and bandwidth. The 930M's Vulkan result of 3729 sits well below its own OpenCL score, indicating that the older, slower memory configuration becomes a bottleneck when the API demands more from the memory bus. The 940MX's 40.10 GB/s of bandwidth versus the 930M's 12.80 GB/s is the single largest hardware discrepancy between the two, and the Vulkan delta reflects it.

Across the two tests, the average benchmark scores summarize the overall picture: the 940MX averages 4844, the 930M averages 4388. That is a 456-point gap, roughly 10.4%, in favor of the 940MX. The percentile rankings, however, are closer than the averages suggest: the 940MX sits at the 28th percentile of all GPUs, the 930M at the 26th. Both are near the bottom of the performance distribution, which is expected for entry-level mobile parts.

The nearest-rival lists add context. The 940MX's closest competitor by average score is the GeForce RTX 3080 12 GB, which scores 4791, a 1.1% delta. This is a bizarre pairing, but it reflects how close the raw averages can be across very different architectures when the benchmark pool is small. The 930M's nearest rivals include the Intel Iris Pro Graphics 5200 (4360, 0.7% delta) and the AMD FirePro W2100 (4295, 2.2% delta), confirming that the 930M competes with integrated-class and low-end workstation parts.

The practical takeaway from the head-to-head data is that the 940MX is the stronger overall performer, but not by a landslide in compute-oriented OpenCL tasks. In Vulkan, the 940MX is in a different class. If a workload leans on modern graphics APIs, the 940MX is the clear choice; if the workload is purely compute-based OpenCL, the difference is small enough that other factors, such as thermal design or driver maturity, could tip the scales.

FAQ

Q: Which GPU wins more benchmarks in the head-to-head comparison?

A: Each GPU wins one test. The 930M wins Geekbench OpenCL (5046 vs. 4939, a 2.1% delta), and the 940MX wins Geekbench Vulkan (4749 vs. 3729, a 27.4% delta).

Q: How much faster is the 940MX in Vulkan?

A: The 940MX scores 4749 in Geekbench Vulkan, which is 27.4% higher than the 930M's 3729. This is the largest performance gap between the two in any recorded test.

Q: Is the 930M ever the better choice?

A: In OpenCL compute tasks, the 930M is marginally ahead by 2.1%. If a user's workload is exclusively OpenCL-based and does not use Vulkan, the 930M's slight edge could be relevant, though the difference is small.

Q: How do the two compare in average benchmark score?

A: The 940MX has an average benchmark score of 4844, while the 930M averages 4388. The 940MX leads by 456 points, approximately 10.4%.

Q: What are the percentile rankings for each GPU?

A: The 940MX is at the 28th percentile of all GPUs, and the 930M is at the 26th percentile. Both are entry-level performers.

Q: Which GPUs are the nearest rivals for each card?

A: The 940MX's nearest rivals include the GeForce GTX 560M (4855, -0.2%), AMD Radeon R6 M255DX (4867, -0.5%), and GeForce GTS 450 (4893, -1%). The 930M's nearest rivals include the GeForce GT 645M (4411, -0.5%), Intel Iris Pro Graphics 5200 (4360, 0.7%), and AMD FirePro W2100 (4295, 2.2%).

The Verdict

The data points to the GeForce 940MX as the better overall GPU for modern workloads. Its 27.4% Vulkan advantage over the 930M is the kind of margin that matters in current games and graphics applications, which increasingly rely on low-overhead APIs. The 940MX also leads in average benchmark score, 4844 to 4388, and holds a higher percentile ranking at 28 versus 26.

The 930M's only win is the OpenCL test, and even there the margin is a thin 2.1%. That result is unlikely to sway a buyer toward the older chip unless their specific application suite is OpenCL-only and they are willing to accept lower performance everywhere else. The 930M does have a lower TDP at 23 W versus the 940MX's 33 W, which could matter in ultra-thin laptops where thermal headroom is scarce, but that is a power consideration, not a performance one.

For a user choosing between these two in a used or budget laptop market, the 940MX is the safer pick. It delivers meaningfully better graphics API performance, a higher average score, and a better percentile standing. The 930M remains a competent entry-level part, but its hardware limitations, particularly the DDR3 memory and lower shader count, cap its potential in ways that the 940MX avoids. The verdict is straightforward: the 940MX for performance, the 930M only if power draw is the overriding concern.

Specification Differences

The two GPUs differ in nearly every compute-related specification. The 940MX uses the GM107 chip with 1,870 million transistors on a 148 mm² die, while the 930M uses the GM108S chip with 1,020 million transistors on a 77 mm² die. Transistor density is slightly higher on the 930M at 13.2M per mm² versus 12.6M per mm², but the 940MX's larger die houses more functional units.

The 940MX has 512 shading units, 32 texture mapping units, and 8 ROPs. The 930M has 384 shading units, 24 TMUs, and the same 8 ROPs. This gives the 940MX a 33% advantage in shader count and a 33% advantage in TMUs. The pixel rate reflects this: the 940MX outputs 6.888 GPixel/s versus 4.392 GPixel/s for the 930M. Texture rate is 27.55 GTexel/s on the 940MX versus 13.18 GTexel/s on the 930M, more than a 2x difference. FP32 compute is 881.7 GFLOPS for the 940MX versus 421.6 GFLOPS for the 930M, again roughly 2x.

Clock speeds differ substantially. The 940MX runs at a base clock of 795 MHz and a boost clock of 861 MHz. The 930M runs at a fixed 549 MHz for both base and boost. Memory clocks are also far apart: the 940MX uses 1253 MHz (5 Gbps effective) GDDR5, while the 930M uses 800 MHz (1600 Mbps effective) DDR3. Both have 2 GB of memory on a 64-bit bus, but the memory type difference produces a massive bandwidth gap: 40.10 GB/s for the 940MX versus 12.80 GB/s for the 930M.

Power and physical characteristics diverge as well. The 940MX is rated at 23 W TDP and uses an MXM Module slot width. The 930M is rated at 33 W TDP and uses an IGP slot width. Neither requires power connectors. Both use PCIe 3.0 x8 and have portable-device-dependent display outputs. The 940MX was released on 2016-06-27, while the 930M came earlier on 2015-03-12. Both are end-of-life, both share the same GeForce 800M predecessor and GeForce 10 Mobile successor, and neither has a launch MSRP in the database.

Architecture Differences

Both chips are built on NVIDIA's Maxwell architecture, so the core instruction set and feature baseline are identical. They are fabricated by TSMC on the same 28 nm process node. The API support is also identical: DirectX 12 (11_0), OpenGL 4.6, and Vulkan 1.4 for both. Neither chip has ray tracing cores or tensor cores.

The architectural differences are quantitative rather than qualitative. The 940MX's GM107 is a larger, more fully featured implementation of Maxwell with 512 shading units and 32 TMUs, whereas the 930M's GM108S is a cut-down variant with 384 shading units and 24 TMUs. The ROP count is the same at 8, which means the two chips have identical pixel output capability per clock, but the 940MX's higher clocks push its pixel rate to 6.888 GPixel/s versus 4.392 GPixel/s.

The most significant architectural divergence is the memory subsystem. The 940MX pairs its 64-bit bus with GDDR5 memory running at 5 Gbps effective, yielding 40.10 GB/s. The 930M pairs the same 64-bit bus with DDR3 at 1600 Mbps effective, yielding 12.80 GB/s. This 3.1x bandwidth advantage is the single largest architectural difference and explains the 940MX's Vulkan dominance. Memory bandwidth is a critical resource for modern graphics APIs, which tend to be more bandwidth-hungry than older implicit APIs.

The transistor counts reflect the scale difference: 1,870 million on the 940MX versus 1,020 million on the 930M. The die size difference, 148 mm² versus 77 mm², follows the transistor count. The 930M's slightly higher transistor density, 13.2M per mm² versus 12.6M, is a minor curiosity, likely due to the smaller die's overhead being amortized differently, but it does not translate into a performance advantage.

Chip design choices also differ in clock strategy. The 940MX has a boost clock 66 MHz above its base, allowing brief performance headroom under favorable thermal conditions. The 930M has no boost capability, locked at 549 MHz. This makes the 930M's performance predictable but also limits its ability to respond to thermally cool moments. The 940MX's TDP of 23 W, lower than the 930M's 33 W, is counterintuitive given its larger die, but it likely reflects the efficiency of the GM107 design and the lower power draw of GDDR5 versus the older DDR3 implementation on the 930M.

In essence, the 930M is a smaller, slower Maxwell chip with a crippled memory interface, while the 940MX is a fuller expression of the same architecture with a proper memory subsystem. The architectural gap is why the Vulkan benchmark shows a 27.4% delta in the 940MX's favor, and why the average scores place it nearly 10% higher overall. The 930M's only redeeming architectural feature is its lower TDP, which makes it suitable for power-constrained designs, but from a pure performance standpoint, the 940MX is the superior Maxwell implementation in every measurable way.