GPU Comparison

# NVIDIA GeForce 940M vs AMD FirePro W5130M

The Verdict

The benchmark data separates these two mobile GPUs clearly, with the NVIDIA GeForce 940M winning the only head-to-head test by a substantial margin. In Geekbench OpenCL, the 940M scores 6,018 points against the FirePro W5130M’s 4,904 points, a 22.7% advantage. The 940M also holds a higher average benchmark score of 5,284 versus 4,904 for the AMD part, and it sits at the 31st percentile among all GPUs compared to the 29th percentile for the FirePro. The 940M’s nearest rivals include the GeForce GTX 980M (average score 5,308, delta -0.4%), the GeForce 930A (5,317, -0.6%), and the GeForce 840M (5,322, -0.7%), all of which are statistically indistinguishable from the 940M. The FirePro W5130M, meanwhile, is clustered with the GeForce RTX 5060 Ti 8 GB (4,901, +0.1%), the GeForce GTS 450 (4,893, +0.2%), and the Radeon R7 M265 (4,929, -0.5%), placing it in a lower performance tier entirely.

For users choosing between these two, the data points firmly toward the 940M for raw compute performance. The 940M also offers a Vulkan score of 4,549, a benchmark the FirePro does not have a recorded result for. The FirePro W5130M does have one architectural advantage: its GDDR5 memory with a 128-bit bus delivers 64.00 GB/s of bandwidth, compared to the 940M’s DDR3 memory on a 64-bit bus at 14.40 GB/s. This bandwidth advantage could matter in memory-bound workloads, but the 940M’s higher clock speeds and superior compute throughput make it the default pick from the data. The FirePro’s only path to preference is if the workload is heavily memory-bandwidth dependent and the software is optimized for GCN 1.0 architecture. Otherwise, the 940M wins on measured performance.

Architecture Differences

The two GPUs come from different architectural generations. The NVIDIA GeForce 940M uses the GM107 chip built on the Maxwell architecture, part of the GeForce 900M generation. It is fabricated on a 28 nm process at TSMC, with 1,870 million transistors on a 148 mm² die, yielding a transistor density of 12.6M per mm². The AMD FirePro W5130M uses the Tropo chip based on GCN 1.0, belonging to the FirePro Mobile (Wx100M) generation. It is also fabricated on 28 nm at TSMC, but packs 1,500 million transistors on a 123 mm² die, for a density of 12.2M per mm². The 940M therefore has 370 million more transistors and a 25 mm² larger die.

Clock speeds differ notably. The 940M runs at a base clock of 1,020 MHz with a boost clock of 1,098 MHz, while the FirePro W5130M runs at 900 MHz base and 925 MHz boost. This gives the 940M a 120 MHz base advantage and a 173 MHz boost advantage. Both chips have identical core counts: 512 shading units, 32 texture mapping units, and 16 raster operations units. Neither has dedicated ray tracing cores or tensor cores.

The memory subsystems diverge sharply. The 940M uses 2 GB of DDR3 memory with a 64-bit bus, running at 900 MHz (1,800 Mbps effective), delivering 14.40 GB/s of bandwidth. The FirePro W5130M also has 2 GB, but it is GDDR5 on a 128-bit bus, running at 1,000 MHz (4 Gbps effective), for 64.00 GB/s of bandwidth, more than 4.4 times the bandwidth of the 940M. The API support differs slightly: both support DirectX 12 and OpenGL 4.6, but the 940M supports Vulkan 1.4 while the FirePro supports Vulkan 1.2.170. The 940M also lists DirectX 12 (11_0) while the FirePro lists DirectX 12 (11_1). The 940M uses an MXM-B (3.0) bus interface with a 75 W TDP, while the FirePro uses PCIe 3.0 x16 and has no listed TDP.

Where Each One Wins

The NVIDIA GeForce 940M wins in compute throughput benchmarks. Its Geekbench OpenCL score of 6,018 versus the FirePro’s 4,904 represents a 22.7% victory. The 940M also posts a Geekbench Vulkan score of 4,549, a test the FirePro W5130M has no recorded result for, meaning the data shows the 940M as the only one of the two with a validated Vulkan performance metric. The 940M also has higher pixel rate (17.57 GPixel/s versus 14.80 GPixel/s), higher texture rate (35.14 GTexel/s versus 29.60 GTexel/s), and higher FP32 throughput (1,124.4 GFLOPS versus 947.2 GFLOPS). These figures follow from the 940M’s higher clock speeds, since both chips have identical core counts.

The AMD FirePro W5130M wins where memory bandwidth dominates. Its 64.00 GB/s bandwidth is more than four times the 940M’s 14.40 GB/s, and it has a 128-bit bus versus 64-bit. In workloads that are bandwidth-limited rather than compute-limited, for example, large data transfers, certain image processing tasks, or applications that stream textures heavily, the FirePro’s GDDR5 memory could close the gap seen in synthetic compute benchmarks. The FirePro also has a higher memory clock in terms of effective data rate: 4 Gbps versus 1,800 Mbps. However, the data does not include a benchmark that isolates memory bandwidth, so this advantage is inferred from specifications rather than measured head-to-head results.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce 940M, with an average benchmark score of 5,284, compared to the AMD FirePro W5130M’s 4,904.

Q: How much faster is the 940M in the head-to-head OpenCL benchmark?

A: The 940M scores 6,018 versus 4,904 for the FirePro, a 22.7% advantage.

Q: Do both GPUs have the same number of shading units?

A: Yes, both have 512 shading units, along with 32 TMUs and 16 ROPs.

Q: Which GPU has higher memory bandwidth?

A: The AMD FirePro W5130M, with 64.00 GB/s from GDDR5 on a 128-bit bus, versus the 940M’s 14.40 GB/s from DDR3 on a 64-bit bus.

Q: Does the FirePro W5130M have a Vulkan benchmark score?

A: No, the data only includes a Geekbench OpenCL score for the FirePro, while the 940M has both OpenCL (6,018) and Vulkan (4,549) scores.

Q: Which GPU has a higher base clock speed?

A: The NVIDIA GeForce 940M, at 1,020 MHz base and 1,098 MHz boost, compared to the FirePro’s 900 MHz base and 925 MHz boost.

Head-to-Head Benchmarks

The single recorded head-to-head benchmark is Geekbench OpenCL. The NVIDIA GeForce 940M scores 6,018 points, while the AMD FirePro W5130M scores 4,904 points. The delta is 22.7% in favor of the 940M. This is a decisive margin in a compute-oriented test, and it aligns with the 940M’s higher clock speeds and larger transistor count (1,870 million versus 1,500 million). The 940M’s FP32 performance of 1,124.4 GFLOPS exceeds the FirePro’s 947.2 GFLOPS by roughly 18.7%, which is consistent with the benchmark gap when accounting for memory bandwidth differences.

The 940M also has a Geekbench Vulkan score of 4,549. There is no corresponding Vulkan score for the FirePro, so this cannot be compared directly. However, the existence of a Vulkan result for the 940M at all, with a score less than 25% below its own OpenCL score, suggests the Maxwell architecture handles Vulkan compute efficiently. The FirePro’s absence from the Vulkan test means no data supports its Vulkan performance.

Looking at the nearest rival clusters reinforces the separation. The 940M’s closest rivals, the GeForce GTX 980M (5,308, -0.4%), GeForce 930A (5,317, -0.6%), and GeForce 840M (5,322, -0.7%), all have average scores within 0.7% of the 940M. This places the 940M in a tight performance band where the GTX 980M, a much higher-tier product in name, is effectively tied with the 940M in average benchmark score. On the AMD side, the FirePro’s nearest rivals include the GeForce RTX 5060 Ti 8 GB (4,901, +0.1%) and the GeForce GTS 450 (4,893, +0.2%), both of which are older or lower-tier parts. The FirePro’s average score of 4,904 is 7.2% below the 940M’s average of 5,284.

In pixel rate, the 940M achieves 17.57 GPixel/s versus 14.80 GPixel/s for the FirePro, a 18.7% advantage. In texture rate, the 940M achieves 35.14 GTexel/s versus 29.60 GTexel/s, a 18.7% advantage as well. These consistent margins mirror the FP32 difference and the clock speed gap, since the core counts are identical. The FirePro’s only numerical victory is in memory bandwidth, where it has 64.00 GB/s versus 14.40 GB/s, a 4.44x advantage. But that advantage does not translate into a benchmark win in the data.

Specification Differences

The two GPUs differ in nearly every measurable specification except core counts and memory size. Process node and foundry are identical: 28 nm at TSMC. Transistors differ: the 940M has 1,870 million versus 1,500 million for the FirePro. Die size differs: 148 mm² versus 123 mm². Transistor density is close, 12.6M per mm² versus 12.2M per mm². Clock speeds: the 940M runs at 1,020 MHz base and 1,098 MHz boost; the FirePro runs at 900 MHz base and 925 MHz boost. Memory type differs: DDR3 versus GDDR5. Bus width differs: 64-bit versus 128-bit. Memory clock differs: 900 MHz (1,800 Mbps effective) versus 1,000 MHz (4 Gbps effective). Bandwidth differs: 14.40 GB/s versus 64.00 GB/s.

Pixel rate: 17.57 GPixel/s versus 14.80 GPixel/s. Texture rate: 35.14 GTexel/s versus 29.60 GTexel/s. FP32: 1,124.4 GFLOPS versus 947.2 GFLOPS. TDP: the 940M lists 75 W, while the FirePro has no TDP listed. Slot width: the 940M is an MXM Module, while the FirePro has no slot width listed. Power connectors: the 940M lists "None", the FirePro lists none. Bus interface: MXM-B (3.0) versus PCIe 3.0 x16. Display outputs: the 940M lists "Portable Device Dependent", the FirePro lists none. API support: DirectX 12 (11_0) versus DirectX 12 (11_1); OpenGL 4.6 for both; Vulkan 1.4 versus 1.2.170. Release dates differ: the 940M launched on 2015-03-12, the FirePro on 2015-10-01. Both are end-of-life. The 940M’s predecessor is the GeForce 800M and successor is the GeForce 10 Mobile; the FirePro’s predecessor is FirePro Mobility and successor is Radeon Pro Mobile. Neither has a launch MSRP in the data.