NVIDIA GeForce GTX 850M vs NVIDIA GeForce GTX 870M Comparison

The NVIDIA GeForce GTX 870M and GTX 850M are two mobile graphics solutions from the same GeForce 800M generation, but they are built on fundamentally different architectures and target distinct performance tiers. The data shows the GTX 870M holds a commanding lead in raw compute performance, while the GTX 850M counters with a significantly more efficient design and a newer feature set. The average benchmark score for the GTX 870M is 9959, placing it in the 48th percentile of all GPUs, while the GTX 850M averages 9302, landing in the 46th percentile.

Where Each One Wins

The performance split between these two GPUs is stark, primarily defined by a single head-to-head benchmark and their respective architectural priorities. The GTX 870M is the clear winner in raw compute throughput. In the Geekbench OpenCL test, the GTX 870M scores 12630 against the GTX 850M's 9821, a decisive 28.6% advantage. This makes the older Kepler-based chip the undisputed choice for any workload that stresses parallel processing, such as GPU-accelerated content creation, scientific simulation, or any OpenCL-compatible application.

Conversely, the GTX 850M wins in the efficiency and integration domain, though this is not represented by a benchmark score in the data. Its 45W TDP is less than half of the GTX 870M's 100W, making it the superior option for thinner, lighter laptops where thermal management and battery life are paramount. The GTX 850M also supports the Vulkan 1.4 API, a newer version than the 1.2.175 found on the GTX 870M, suggesting better long-term compatibility with modern games and applications that leverage this API. While the GTX 870M wins the only direct performance comparison, the GTX 850M's advantage lies in its suitability for a different class of portable devices. The data implies a trade-off: maximum performance against maximum portability and efficiency.

Architecture Differences

The two GPUs represent a generational shift within NVIDIA's own lineup. The GTX 870M is based on the GK104 chip using the Kepler architecture, while the GTX 850M utilizes the GM107 chip with the newer Maxwell architecture. This is a fundamental difference that explains their distinct performance and efficiency profiles. The Kepler chip is a large, complex design, with a die size of 294 mm² and a transistor count of 3,540 million, built on a 28 nm process at TSMC. In contrast, the Maxwell chip is much smaller at 148 mm², with 1,870 million transistors on the same 28 nm node. This results in a slightly higher transistor density for the Maxwell chip (12.6M / mm² vs 12.0M / mm²), indicating a more efficient packing of logic.

The core configurations diverge significantly. The GTX 870M has 1344 shading units, 112 texture mapping units (TMUs), and 24 render output units (ROPs). The GTX 850M, by comparison, has a much leaner setup with 640 shading units, 40 TMUs, and 16 ROPs. This difference in core count directly translates to the GTX 870M's higher theoretical compute rates: its FP32 performance is 2.599 TFLOPS and its pixel rate is 27.08 GPixel/s. The GTX 850M, with its Maxwell architecture, achieves 1,154.6 GFLOPS and 14.43 GPixel/s. Even with fewer cores, the Maxwell architecture allows the GTX 850M to deliver respectable performance at a fraction of the power draw, as evidenced by its 45W TDP. The GTX 870M, with more than double the shading units and a higher clock speed, requires 100W. This architectural difference is the core reason for the performance-per-watt disparity.

FAQ

Q: Which GPU is faster in compute-heavy tasks?

A: The GTX 870M is significantly faster. In the Geekbench OpenCL test, it scores 12630 compared to the GTX 850M's 9821, a 28.6% difference in favor of the GTX 870M.

Q: How do their average benchmark scores compare?

A: The GTX 870M has an average benchmark score of 9959, while the GTX 850M has an average of 9302. This puts the GTX 870M in the 48th percentile of all GPUs, slightly ahead of the GTX 850M's 46th percentile.

Q: Is the newer Maxwell architecture in the GTX 850M better than the older Kepler in the GTX 870M?

A: The data suggests a trade-off. The Maxwell architecture in the GTX 850M is more efficient, with a 45W TDP versus the GTX 870M's 100W. However, the Kepler-based GTX 870M still delivers substantially higher raw performance, winning the only head-to-head benchmark by a wide margin.

Q: What are the memory specifications for each card?

A: The GTX 870M comes with 3 GB of GDDR5 memory on a 192-bit bus, providing 120.0 GB/s of bandwidth. The GTX 850M uses 2 GB of DDR3 memory on a 128-bit bus, offering a much lower 32.03 GB/s of bandwidth.

Q: Do they support different graphics APIs?

A: Yes. While both support DirectX 12 (11_0) and OpenGL 4.6, the GTX 850M supports Vulkan 1.4, whereas the GTX 870M only supports Vulkan 1.2.175.

Q: Which GPU is more likely to be found in a thin and light laptop?

A: The GTX 850M. Its low 45W TDP and IGP slot width make it suitable for compact, power-efficient designs, whereas the GTX 870M's 100W TDP and MXM Module form factor are typical of larger, high-performance gaming laptops.

Specification Differences

The specifications reveal two very different design philosophies. The most critical differences are in the GPU core, memory subsystem, and power envelope.

• Chip & Architecture: The GTX 870M uses the GK104 chip based on Kepler architecture. The GTX 850M uses the GM107 chip based on Maxwell architecture.
• Transistors & Die Size: The GTX 870M has 3,540 million transistors on a 294 mm² die. The GTX 850M has 1,870 million transistors on a 148 mm² die.
• Shading Units: The GTX 870M has 1344 shading units, while the GTX 850M has 640.
• TMUs: The GTX 870M has 112 TMUs, compared to the GTX 850M's 40.
• ROPs: The GTX 870M has 24 ROPs, while the GTX 850M has 16.
• Clock Speeds: The GTX 870M has a base clock of 941 MHz and a boost clock of 967 MHz. The GTX 850M's base and boost clocks are not listed in the data.
• Memory Size & Type: The GTX 870M features 3 GB of GDDR5, while the GTX 850M features 2 GB of DDR3.
• Memory Bus Width & Bandwidth: The GTX 870M has a 192-bit bus with 120.0 GB/s bandwidth. The GTX 850M has a 128-bit bus with 32.03 GB/s bandwidth.
• Memory Clock: The GTX 870M's memory runs at 1250 MHz (5 Gbps effective), while the GTX 850M's memory runs at 1001 MHz (2 Gbps effective).
• Pixel & Texture Rates: The GTX 870M has a pixel rate of 27.08 GPixel/s and a texture rate of 108.3 GTexel/s. The GTX 850M has a pixel rate of 14.43 GPixel/s and a texture rate of 36.08 GTexel/s.
• FP32 Performance: The GTX 870M achieves 2.599 TFLOPS, while the GTX 850M achieves 1,154.6 GFLOPS.
• TDP: The GTX 870M has a TDP of 100 W, whereas the GTX 850M has a TDP of 45 W.
• Slot Width & Bus Interface: The GTX 870M is an MXM Module with an MXM-B (3.0) interface. The GTX 850M is an IGP with a PCIe 3.0 x16 interface.
• Vulkan Support: The GTX 850M supports Vulkan 1.4, a newer version than the GTX 870M's 1.2.175.

Head-to-Head Benchmarks

The only direct benchmark comparison available is Geekbench OpenCL, and it is a landslide victory for the GTX 870M. The GTX 870M scores 12630, while the GTX 850M scores 9821. This represents a 28.6% performance delta in favor of the GTX 870M. This score is a clear indicator of the raw compute advantage held by the larger Kepler chip.

Looking at the rival landscape for each card provides further context. The GTX 870M's average score of 9959 puts it within a tight cluster of competitors. It is a mere 0.5% behind the AMD Radeon Pro 5300M (10013), 0.8% behind the NVIDIA Quadro K5100M (10043), and 1.1% behind the AMD Radeon R9 M375 (10070). Interestingly, it sits 1.1% ahead of the NVIDIA Quadro 6000 (9846). This suggests the GTX 870M is positioned at the edge of a performance tier, trading blows with these professional and mid-range desktop parts.

The GTX 850M's average score of 9302 places it in a similar competitive bracket, but at a lower absolute level. It is virtually tied with the NVIDIA GeForce GTX 465 (9294), being 0.1% ahead. It also sits 0.1% behind the AMD Radeon R7 M380 (9313) and 0.3% ahead of the NVIDIA GeForce GTX 960 (9273). Its 0.9% lead over the AMD Radeon Vega 8 (9221) shows it is competitive with modern integrated graphics solutions. These numbers demonstrate that while the GTX 870M is competing at a higher absolute performance level, the GTX 850M is still a relevant performer against its own set of rivals, particularly given its much lower power consumption.

The Verdict

The choice between these two GPUs is a clear-cut decision based on user priorities, as the data leaves little room for ambiguity. For users who require maximum graphics performance from their laptop, the NVIDIA GeForce GTX 870M is the only choice. Its 28.6% lead in the OpenCL benchmark and its significantly higher average score (9959 vs 9302) make it the superior option for gaming, 3D rendering, and any GPU-accelerated workload. The larger memory pool of 3 GB and higher bandwidth of 120.0 GB/s also give it a substantial advantage in texture-heavy applications and higher resolution gaming. This is the GPU for the user who prioritizes frame rates and compute capability above all else.

For users who prioritize portability, battery life, and a cooler, quieter system, the GTX 850M is the more sensible pick. Its 45W TDP, less than half of the GTX 870M's 100W, makes it ideal for thinner laptops where a high-power MXM module is not feasible. The fact that it achieves a 46th percentile ranking while drawing so much less power reflects the efficiency of the Maxwell architecture. Furthermore, its support for Vulkan 1.4, a newer API version than the GTX 870M's 1.2.175, suggests it may be better prepared for future software. The GTX 850M is for the user who wants a capable gaming and multimedia laptop that doesn't sacrifice battery life or require a bulky chassis. The data shows two different values: the 870M is pure performance, and the 850M is efficient capability.