NVIDIA GeForce GTX 870M vs NVIDIA GeForce GTX TITAN BLACK Comparison

NVIDIA
GEFORCE

NVIDIA GeForce GTX 870M

CORE STATE GK104
VRAM 3 GB
CLOCK SPEED 967 MHz
TDP 100 W
BUS WIDTH 192 bit
ARCHITECTURE Kepler
nm
PROCESS 28 nm
LAUNCH DATE 2014
VS
NVIDIA
GEFORCE

GeForce GTX TITAN BLACK

CORE STATE GK110B
VRAM 6 GB
CLOCK SPEED 980 MHz
TDP 250 W
BUS WIDTH 384 bit
ARCHITECTURE Kepler
nm
PROCESS 28 nm
LAUNCH DATE 2014

PERFORMANCE BENCHMARKS

geekbench_metal
7,288
24,172
geekbench_opencl
12,630
27,218
geekbench_vulkan
N/A
29,561
passmark_directx_10
N/A
47
passmark_directx_11
N/A
73
passmark_directx_12
N/A
34
passmark_directx_9
N/A
133
passmark_g2d
N/A
595
passmark_g3d
N/A
9,184
passmark_gpu_compute
N/A
3,726

Analysis: NVIDIA GeForce GTX 870M vs NVIDIA GeForce GTX TITAN BLACK

The NVIDIA GeForce GTX 870M and the NVIDIA GeForce GTX TITAN BLACK are both Kepler-architecture parts from NVIDIA, but they occupy vastly different segments of the market. The data shows a decisive performance gap, yet the 870M’s mobile positioning offers a distinct set of trade-offs. This analysis relies strictly on the provided benchmark results and specifications to quantify the differences.

Head-to-Head Benchmarks

The benchmark data is unambiguous: the GTX TITAN BLACK wins both recorded head-to-head tests. In the Geekbench Metal test, the TITAN BLACK scores 24,172 points against the 870M’s 7,288 points. This represents a delta of -69.8% for the 870M, meaning the TITAN BLACK is roughly 3.3 times faster in this specific workload. The margin is similarly lopsided in Geekbench OpenCL, where the TITAN BLACK posts 27,218 points versus the 870M’s 12,630 points, a -53.6% delta for the mobile part. In practical terms, the TITAN BLACK more than doubles the 870M’s OpenCL output.

The aggregate scores reinforce this hierarchy. The 870M’s average benchmark score is 9,959, while the TITAN BLACK’s is 9,474. This is a counterintuitive result—the lower-performing part in direct comparisons has a higher average score. The explanation lies in the composition of the benchmark suites. The 870M’s average is pulled up by its two Geekbench scores, while the TITAN BLACK’s average is dragged down by its Passmark results, which include a DirectX 10 score of 47, a DirectX 11 score of 73, and a DirectX 9 score of 133. These legacy API tests are not favorable to the TITAN BLACK, yet its Geekbench scores remain dominant. The TITAN BLACK’s Passmark G3D score of 9,184 and GPU compute score of 3,726 are strong, but they are weighted alongside older API tests that penalize its modern architecture.

Looking at the nearest rivals provides context for each part’s standing. The 870M’s average score of 9,959 places it within 1.1% of the NVIDIA Quadro 6000 (9,846) and 0.5% below the AMD Radeon Pro 5300M (10,013). The TITAN BLACK’s average of 9,474 puts it 0.2% ahead of the NVIDIA GeForce GTX 760 (9,458) and 1.7% ahead of the AMD Radeon R7 M380 (9,313). Notably, the TITAN BLACK’s percentile rank among all GPUs is 46, slightly lower than the 870M’s 48. This suggests that while the TITAN BLACK is overwhelmingly faster in compute-heavy tests, the 870M’s overall profile is more balanced across the full range of benchmark types.

The wins column is decisive: the TITAN BLACK wins 2 out of 2 head-to-head tests, while the 870M wins none. There is no workload in the provided data where the 870M comes out ahead. The only question is the magnitude of the loss, which varies from roughly half to two-thirds depending on the test.

Architecture Differences

Both GPUs use the Kepler architecture and are fabricated on TSMC’s 28 nm process, but the silicon underneath is fundamentally different. The 870M uses the GK104 chip, which contains 3,540 million transistors on a 294 mm² die, yielding a transistor density of 12.0 million per mm². The TITAN BLACK uses the GK110B chip, which is a much larger part: 7,080 million transistors on a 561 mm² die, with a slightly higher density of 12.6 million per mm². The TITAN BLACK’s die is nearly double the size and contains exactly double the transistor count.

The compute resources scale accordingly. The 870M has 1,344 shading units, 112 texture mapping units, and 24 ROPs. The TITAN BLACK has 2,880 shading units, 240 TMUs, and 48 ROPs. Each of these counts is exactly double (or near-double) the 870M’s, which explains the large performance delta in raw throughput. The pixel rate for the 870M is 27.08 GPixel/s, while the TITAN BLACK achieves 58.80 GPixel/s. The texture rate is 108.3 GTexel/s versus 235.2 GTexel/s. FP32 performance, a key measure of compute capability, is 2.599 TFLOPS for the 870M and 5.645 TFLOPS for the TITAN BLACK—again, more than double.

Clock speeds are closer than the compute resources. The 870M has a base clock of 941 MHz and a boost clock of 967 MHz. The TITAN BLACK has a lower base clock of 889 MHz but a higher boost clock of 980 MHz. This means the TITAN BLACK’s advantage comes entirely from its wider architecture, not from running faster. The memory subsystem also differs sharply. The 870M uses 3 GB of GDDR5 on a 192-bit bus, delivering 120.0 GB/s of bandwidth. The TITAN BLACK has 6 GB of GDDR5 on a 384-bit bus, delivering 336.0 GB/s—2.8 times the bandwidth. The memory clock is 1250 MHz (5 Gbps effective) for the 870M and 1750 MHz (7 Gbps effective) for the TITAN BLACK.

Power and physical design are where the 870M’s mobile nature becomes clear. The 870M has a TDP of 100 W and uses an MXM Module slot width with no power connectors, reflecting its laptop-oriented design. The TITAN BLACK has a 250 W TDP, requires a dual-slot cooler, and needs both a 6-pin and an 8-pin power connector, with a suggested 600 W power supply. The TITAN BLACK is a desktop card measuring 267 mm in length, while the 870M’s dimensions are not listed, but its MXM form factor implies a compact module. The bus interface differs as well: the 870M uses MXM-B (3.0), while the TITAN BLACK uses PCIe 3.0 x16.

API support shows a minor divergence. Both support OpenGL 4.6 and Vulkan 1.2.175, but the 870M supports DirectX 12 (11_0), while the TITAN BLACK supports DirectX 12 (11_1). This means the TITAN BLACK has a slightly higher DirectX feature level, though both are limited to the 11.x era of features.

The Verdict

From the data alone, the choice is clear for raw performance: the GTX TITAN BLACK wins every recorded benchmark by a wide margin. Its double shading units, double TMUs, double ROPs, and nearly triple memory bandwidth translate into a 53.6% to 69.8% lead over the 870M in the head-to-head tests. For any workload that stresses FP32 compute, texture fill, or memory bandwidth, the TITAN BLACK is the superior part by a factor of two or more.

However, the 870M’s higher average benchmark score (9,959 vs. 9,474) and higher percentile rank (48 vs. 46) indicate that it is not universally inferior in aggregate metrics. This is likely because the 870M’s benchmark profile is narrowly focused on Geekbench tests, which are compute-oriented, while the TITAN BLACK’s broader suite includes legacy DirectX tests that penalize it. A user who only cares about modern compute workloads would prefer the TITAN BLACK without hesitation. A user who values a balanced score across a wide range of APIs might find the 870M’s profile less volatile, but this is a statistical artifact rather than a real performance advantage.

The physical and power differences are decisive for system integration. The 870M is a 100 W MXM module with no external power connectors, designed for laptops where space and thermal limits are strict. The TITAN BLACK is a 250 W dual-slot desktop card requiring a 600 W power supply and dedicated power connectors. These are not interchangeable products. The 870M exists to bring Kepler-class performance to portable systems; the TITAN BLACK exists to maximize desktop throughput. The TITAN BLACK’s launch MSRP is 999 USD, a figure that reflects its flagship positioning, while the 870M has no listed launch price.

For a laptop user, the TITAN BLACK is physically impossible to install. For a desktop user, the 870M would be an unnecessary compromise in every measurable way except power consumption. The data supports the TITAN BLACK as the performance king, but the 870M remains a relevant part for its specific mobile niche. The verdict is not about which is "better" in absolute terms—it is about which fits the target platform. On raw numbers, the TITAN BLACK dominates. On portability, the 870M is the only option.

Specification Differences

  • Chip: GK104 (870M) vs. GK110B (TITAN BLACK)
  • Generation: GeForce 800M (870M) vs. GeForce 700 (TITAN BLACK)
  • Transistors: 3,540 million (870M) vs. 7,080 million (TITAN BLACK)
  • Die Size: 294 mm² (870M) vs. 561 mm² (TITAN BLACK)
  • Transistor Density: 12.0M / mm² (870M) vs. 12.6M / mm² (TITAN BLACK)
  • Base Clock: 941 MHz (870M) vs. 889 MHz (TITAN BLACK)
  • Boost Clock: 967 MHz (870M) vs. 980 MHz (TITAN BLACK)
  • Memory Clock: 1250 MHz / 5 Gbps effective (870M) vs. 1750 MHz / 7 Gbps effective (TITAN BLACK)
  • Memory Size: 3 GB (870M) vs. 6 GB (TITAN BLACK)
  • Memory Bus Width: 192 bit (870M) vs. 384 bit (TITAN BLACK)
  • Memory Bandwidth: 120.0 GB/s (870M) vs. 336.0 GB/s (TITAN BLACK)
  • Shading Units: 1,344 (870M) vs. 2,880 (TITAN BLACK)
  • TMUs: 112 (870M) vs. 240 (TITAN BLACK)
  • ROPs: 24 (870M) vs. 48 (TITAN BLACK)
  • Pixel Rate: 27.08 GPixel/s (870M) vs. 58.80 GPixel/s (TITAN BLACK)
  • Texture Rate: 108.3 GTexel/s (870M) vs. 235.2 GTexel/s (TITAN BLACK)
  • FP32: 2.599 TFLOPS (870M) vs. 5.645 TFLOPS (TITAN BLACK)
  • TDP: 100 W (870M) vs. 250 W (TITAN BLACK)
  • Slot Width: MXM Module (870M) vs. Dual-slot (TITAN BLACK)
  • Power Connectors: None (870M) vs. 1x 6-pin + 1x 8-pin (TITAN BLACK)
  • Suggested PSU: Not listed (870M) vs. 600 W (TITAN BLACK)
  • Bus Interface: MXM-B (3.0) (870M) vs. PCIe 3.0 x16 (TITAN BLACK)
  • Display Outputs: Portable Device Dependent (870M) vs. 2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 (TITAN BLACK)
  • DirectX Support: 12 (11_0) (870M) vs. 12 (11_1) (TITAN BLACK)
  • Dimensions: Not listed (870M) vs. 267 mm length, 111 mm height, 38 mm width (TITAN BLACK)
  • Release Date: 2014-03-11 (870M) vs. 2014-02-17 (TITAN BLACK)

FAQ

Q: Which GPU is faster in Geekbench Metal?

A: The GTX TITAN BLACK scores 24,172 points, which is 69.8% higher than the 870M’s 7,288 points. The TITAN BLACK wins this test decisively.

Q: How much memory bandwidth does each card have?

A: The 870M has 120.0 GB/s over a 192-bit bus, while the TITAN BLACK has 336.0 GB/s over a 384-bit bus. The TITAN BLACK provides 2.8 times the bandwidth.

Q: Are these GPUs from the same architecture?

A: Yes, both use NVIDIA’s Kepler architecture and are fabricated on a 28 nm process by TSMC. However, they use different chips: GK104 for the 870M and GK110B for the TITAN BLACK.

Q: What is the average benchmark score difference?

A: The 870M has an average benchmark score of 9,959, which is higher than the TITAN BLACK’s 9,474. Despite this, the TITAN BLACK wins both direct head-to-head tests.

Q: Can the GTX 870M be installed in a desktop?

A: The 870M uses an MXM Module slot width with an MXM-B (3.0) bus interface and no power connectors. It is designed for portable devices, not standard desktop PCIe slots.

Q: What is the TITAN BLACK’s launch MSRP?

A: The GTX TITAN BLACK has a launch MSRP of 999 USD. The 870M does not have a listed launch MSRP in the data.

DETAILED SPECIFICATIONS

SPECIFICATION
GTX 870M
GTX TITAN BLACK
Core Specs
Shading Units
1,344
2,880 +114.3%
Shaders
1,344
2,880 +114.3%
TMUs
112
240 +114.3%
ROPs
24
48 +100.0%
Clocks
Base Clock
941 MHz
889 MHz
Boost Clock
967 MHz
980 MHz
Memory Clock
1250 MHz 5 Gbps effective
1750 MHz 7 Gbps effective
Memory
Memory Size
3 GB
6 GB
VRAM (MB)
3,072
6,144 +100.0%
Memory Type
GDDR5
GDDR5
Memory Bus
192 bit
384 bit
Bandwidth
120.0 GB/s
336.0 GB/s
Cache
L1 Cache
16 KB (per SMX)
16 KB (per SMX)
L2 Cache
384 KB
1536 KB
Performance
Pixel Rate
27.08 GPixel/s
58.80 GPixel/s
Texture Rate
108.3 GTexel/s
235.2 GTexel/s
FP32 (TFLOPS)
2.599 TFLOPS
5.645 TFLOPS
FP64 (TFLOPS)
108.3 GFLOPS (1:24)
1.882 TFLOPS (1:3)
Power
TDP
100 W
250 W
TDP (W)
100
250 +150.0%
Suggested PSU
—
600 W
Power Connectors
None
1x 6-pin + 1x 8-pin
Architecture
Architecture
Kepler
Kepler
GPU Name
GK104
GK110B
Generation
GeForce 800M
GeForce 700
Process Size
28 nm
28 nm
Transistors
3,540 million
7,080 million
Die Size
294 mm²
561 mm²
Foundry
TSMC
TSMC
Density
12.0M / mm²
12.6M / mm²
API Support
DirectX
12 (11_0)
12 (11_1)
OpenGL
4.6
4.6
Vulkan
1.2.175
1.2.175
OpenCL
3.0
3.0
CUDA
3.0
3.5
Shader Model
6.5 (5.1)
6.5 (5.1)
Physical
Slot Width
MXM Module
Dual-slot
Length
—
267 mm 10.5 inches
Height
—
111 mm 4.4 inches
Outputs
Portable Device Dependent
2x DVI1x HDMI 1.4a1x DisplayPort 1.2
Bus Interface
MXM-B (3.0)
PCIe 3.0 x16
Other
Launch Price
—
999 USD
Production
End-of-life
End-of-life
Predecessor
GeForce 700M
GeForce 600
Successor
GeForce 900M
GeForce 900
View GeForce GTX 870M Details View GeForce GTX TITAN BLACK Details