AMD FirePro W5170M vs NVIDIA GeForce GTX 870M Comparison

FAQ

Q: How do the two GPUs compare in overall benchmark performance?

A: The NVIDIA GeForce GTX 870M has an average benchmark score of 9959, while the AMD FirePro W5170M scores 8595. The GTX 870M sits at the 48th percentile of all GPUs, whereas the FirePro W5170M sits at the 44th percentile.

Q: Which GPU wins in the OpenCL benchmark?

A: The NVIDIA GeForce GTX 870M wins the only direct head-to-head benchmark, Geekbench OpenCL, with a score of 12630 versus 8140 for the AMD FirePro W5170M. That represents a 55.2% advantage for the NVIDIA part.

Q: What are the closest rivals to each GPU according to the database?

A: The GTX 870M's nearest rival is the AMD Radeon Pro 5300M, which scores 10013, just 0.5% higher. The FirePro W5170M's nearest rival is the Intel Arc A380, which scores 8558, only 0.4% lower than the AMD part.

Q: How do the memory subsystems differ between the two?

A: The GTX 870M has 3 GB of GDDR5 memory on a 192-bit bus, delivering 120.0 GB/s of bandwidth. The FirePro W5170M has 2 GB of GDDR5 on a 128-bit bus, providing 72.00 GB/s. The NVIDIA part offers 66.7% more bandwidth.

Q: What are the API compatibility differences?

A: Both support DirectX 12 and OpenGL 4.6. The GTX 870M supports Vulkan 1.2.175, while the FirePro W5170M supports Vulkan 1.2.170. The DirectX versions differ as well: the NVIDIA part is listed as 12 (11_0), while the AMD part is 12 (11_1).

Q: Which GPU was released later?

A: The AMD FirePro W5170M was released on 2014-08-24, while the NVIDIA GeForce GTX 870M was released earlier on 2014-03-11. Both are now end-of-life products.

The Verdict

The benchmark data points to a clear overall winner: the NVIDIA GeForce GTX 870M. Its average score of 9959 surpasses the FirePro W5170M's 8595 by roughly 15.9%. In the single shared benchmark, the OpenCL test, the NVIDIA part dominates with a 55.2% margin. The GTX 870M also claims a higher percentile ranking (48th versus 44th), indicating it outperforms a larger share of the GPU population.

For users prioritizing raw compute throughput in OpenCL workloads, the GTX 870M is the obvious choice. Its performance profile places it near the AMD Radeon Pro 5300M (10013, only 0.5% higher) and the NVIDIA Quadro K5100M (10043, 0.8% higher), meaning it competes in a performance class above the FirePro W5170M's neighborhood, where rivals include the Intel Arc A380 (8558) and the NVIDIA Quadro P2200 (8686).

The FirePro W5170M, however, should not be dismissed entirely. It offers a different feature set, including a newer DirectX version (12 (11_1) versus 12 (11_0)) and a smaller, denser chip. Its Vulkan score of 9050, while not directly comparable to the GTX 870M's OpenCL result, shows a distinct capability in that API. The GTX 870M has no recorded Vulkan benchmark, so the AMD part holds the only data point in that test category.

The verdict depends on workload. The GTX 870M is the performance leader across the board in recorded benchmarks. The FirePro W5170M is the alternative for those prioritizing the specific API support and architectural traits of AMD's GCN 1.0 design. For pure compute performance, the data favors NVIDIA decisively.

Head-to-Head Benchmarks

The database contains one direct head-to-head benchmark between these two GPUs: Geekbench OpenCL. The results are lopsided. The NVIDIA GeForce GTX 870M scores 12630, while the AMD FirePro W5170M scores 8140. The delta is 55.2% in favor of NVIDIA. This is not a marginal victory; it is a substantial performance gap that places the two GPUs in different tiers of compute capability.

Context from the nearest rivals reinforces this separation. The GTX 870M's average score of 9959 is within 1.1% of the AMD Radeon R9 M375 (10070) and within 0.8% of the NVIDIA Quadro K5100M (10043). It even edges out the NVIDIA Quadro 6000 (9846) by 1.1%. The FirePro W5170M, by contrast, sits in a lower band. Its average of 8595 is just 0.4% above the Intel Arc A380 (8558) and 1.6% above both the AMD Radeon HD 8870M (8462) and the NVIDIA GeForce MX330 (8458). It trails the NVIDIA Quadro P2200 (8686) by 1.1%.

The OpenCL result aligns with these broader averages. The GTX 870M's single-test score of 12630 is far above its own average, suggesting the OpenCL workload plays to its strengths. The FirePro W5170M's OpenCL score of 8140 is slightly below its average, indicating this particular test is not its strongest area. The AMD part does have a recorded Vulkan score of 9050, a test category where the GTX 870M has no data, so direct comparison in that API is impossible.

The wins tally reflects the data: the GTX 870M has 1 win in head-to-head benchmarks, the FirePro W5170M has 0. The 55.2% delta is the largest recorded margin between the two across any shared test. For users whose workloads resemble the OpenCL benchmark, the GTX 870M is the superior hardware by a wide margin.

Specification Differences

The two GPUs diverge on nearly every core specification. The NVIDIA GeForce GTX 870M uses the GK104 chip with 3,540 million transistors on a 294 mm² die. The AMD FirePro W5170M uses the Tropo chip with 1,500 million transistors on a 123 mm² die. The transistor density is nearly identical: 12.0M / mm² for NVIDIA versus 12.2M / mm² for AMD.

Clock speeds differ modestly. The GTX 870M has a base clock of 941 MHz and a boost clock of 967 MHz. The FirePro W5170M runs at 900 MHz base and 925 MHz boost. The NVIDIA part is roughly 4.4% faster at base and 4.5% faster at boost.

The memory configuration shows a larger gap. The GTX 870M offers 3 GB of GDDR5 on a 192-bit bus with 120.0 GB/s bandwidth and an effective speed of 5 Gbps. The FirePro W5170M offers 2 GB of GDDR5 on a 128-bit bus with 72.00 GB/s bandwidth and an effective speed of 4.5 Gbps. The NVIDIA part has 50% more memory capacity and 66.7% more bandwidth.

Compute resources are heavily skewed toward NVIDIA. The GTX 870M has 1344 shading units, 112 texture mapping units, and 24 ROPs. The FirePro W5170M has 640 shading units, 40 TMUs, and 16 ROPs. Pixel rates are 27.08 GPixel/s versus 14.80 GPixel/s, and texture rates are 108.3 GTexel/s versus 37.00 GTexel/s. FP32 throughput is 2.599 TFLOPS for NVIDIA versus 1,184.0 GFLOPS for AMD.

The bus interface differs: the GTX 870M uses MXM-B (3.0), while the FirePro W5170M uses MXM-A (3.0). The GTX 870M has a TDP of 100 W; the FirePro W5170M has no TDP listed. Both use MXM Module slot width, have no power connectors, and have portable-device-dependent display outputs.

Architecture Differences

The architectural divide is fundamental. The NVIDIA GeForce GTX 870M is built on the Kepler architecture, a design that prioritizes high shader counts and efficient geometry processing. The AMD FirePro W5170M uses GCN 1.0, AMD's Graphics Core Next architecture, which emphasizes compute throughput and a different execution model. Both are manufactured on a 28 nm process at TSMC, but the design philosophies diverge sharply.

The chip sizes tell the story. NVIDIA's GK104 is a large die at 294 mm² with 3,540 million transistors. AMD's Tropo is a much smaller chip at 123 mm² with 1,500 million transistors. The density figures are nearly identical (12.0M versus 12.2M per mm²), so the size difference reflects a deliberate choice: NVIDIA packed far more hardware into its GPU, while AMD opted for a leaner design.

Shader architecture differs significantly. The GTX 870M fields 1344 shading units, more than double the FirePro W5170M's 640. Texture units follow suit: 112 versus 40. The ROP count is 24 versus 16. These differences manifest directly in the pixel and texture rates: 27.08 GPixel/s versus 14.80 GPixel/s, and 108.3 GTexel/s versus 37.00 GTexel/s.

API support shows subtle but real differences. The FirePro W5170M supports DirectX 12 (11_1), a slightly newer feature level than the GTX 870M's DirectX 12 (11_0). Both support OpenGL 4.6. Vulkan versions are close: 1.2.175 for NVIDIA versus 1.2.170 for AMD. The AMD part has a recorded Vulkan benchmark score of 9050, while the NVIDIA part has no Vulkan data in the database.

The FirePro is part of AMD's FirePro Mobile (Wx100M) generation, while the GTX 870M belongs to NVIDIA's GeForce 800M generation. The GTX 870M's predecessor is the GeForce 700M and its successor is the GeForce 900M. The FirePro W5170M's predecessor is FirePro Mobility and its successor is Radeon Pro Mobile. These lineage differences reflect the distinct market segments each GPU was designed to serve.

Where Each One Wins

The NVIDIA GeForce GTX 870M wins in every recorded shared benchmark. Its OpenCL score of 12630 demolishes the FirePro W5170M's 8140 by 55.2%. The average benchmark score also favors NVIDIA: 9959 versus 8595, a 15.9% advantage. The percentile ranking confirms the pattern: 48th versus 44th. In terms of raw compute, memory bandwidth, and shader throughput, the GTX 870M is the clear victor. Its 120.0 GB/s bandwidth, 2.599 TFLOPS FP32, and 1344 shading units give it a decisive edge in data-heavy and parallel workloads. The nearest rival comparisons support this: the GTX 870M trades blows with the AMD Radeon Pro 5300M (10013, 0.5% higher) and the NVIDIA Quadro K5100M (10043, 0.8% higher), while the FirePro W5170M competes with the Intel Arc A380 (8558, 0.4% lower) and the NVIDIA Quadro P2200 (8686, 1.1% higher).

The AMD FirePro W5170M wins in the areas where the GTX 870M has no recorded data. It has a Vulkan score of 9050, a test category absent for the NVIDIA part. It also supports DirectX 12 (11_1), a newer feature level than the GTX 870M's DirectX 12 (11_0). For applications built around Vulkan or requiring that specific DirectX feature level, the FirePro W5170M holds a unique advantage. Its smaller die (123 mm² versus 294 mm²) and lower transistor count (1,500 million versus 3,540 million) suggest a more power-efficient design, though no TDP is listed for the AMD part to confirm this.

The use-case split is straightforward. The GTX 870M is the choice for OpenCL compute, high-resolution texture work, and any workload that demands memory bandwidth or shader throughput. The FirePro W5170M is the choice for Vulkan-centric applications or environments where the newer DirectX feature level matters. The data shows a 1-0 win tally in head-to-head benchmarks for NVIDIA, but the AMD part's untested capabilities in Vulkan leave room for scenario-specific consideration. Each GPU wins where its architectural strengths align with the workload.