NVIDIA GeForce GT 645M vs NVIDIA GeForce GTX 1050 Ti Comparison

NVIDIA
GEFORCE

NVIDIA GeForce GT 645M

CORE STATE GK107
VRAM 2 GB
CLOCK SPEED 780 MHz
TDP 32 W
BUS WIDTH 128 bit
ARCHITECTURE Kepler
nm
PROCESS 28 nm
LAUNCH DATE 2012
VS
NVIDIA
GEFORCE

GeForce GTX 1050 Ti

CORE STATE GP107
VRAM 4 GB
CLOCK SPEED 1392 MHz
TDP 75 W
BUS WIDTH 128 bit
ARCHITECTURE Pascal
nm
PROCESS 14 nm
LAUNCH DATE 2016

PERFORMANCE BENCHMARKS

geekbench_metal
5,679
7,834
geekbench_opencl
2,680
18,129
geekbench_vulkan
4,875
10,001
3dmark_3dmark_steel_nomad_dx12
N/A
305
passmark_directx_10
N/A
31
passmark_directx_11
N/A
45
passmark_directx_12
N/A
26
passmark_directx_9
N/A
104
passmark_g2d
N/A
651
passmark_g3d
N/A
6,340
passmark_gpu_compute
N/A
2,652

Analysis: NVIDIA GeForce GT 645M vs NVIDIA GeForce GTX 1050 Ti

The NVIDIA GeForce GT 645M and the NVIDIA GeForce GTX 1050 Ti represent two very different eras of mobile graphics, and the benchmark data reflects a complete generational shift. The GTX 1050 Ti is the decisive winner in every head-to-head test, but the GT 645M holds its own in the context of its own time. The data shows the GTX 1050 Ti is the clear choice for any modern workload, while the GT 645M is a legacy part best suited for basic, older tasks. The GTX 1050 Ti wins 3 out of 3 head-to-head benchmarks, with a perfect 100% win rate, making the verdict straightforward for anyone choosing between the two today.

The Verdict

The data is unambiguous: the NVIDIA GeForce GTX 1050 Ti is the superior product by every measurable metric. In the three shared benchmarks, it wins decisively, with the smallest gap being a 27.5% lead in Geekbench Metal. The GT 645M, on the other hand, is a product of an older era, with its performance clustered near the 26th percentile of all GPUs, while the GTX 1050 Ti sits at the 25th percentile. This near-identical percentile ranking is notable, as it suggests that while the GTX 1050 Ti is far faster in absolute terms, both cards occupy a similar relative position in the modern GPU landscape—entry-level. The GTX 1050 Ti is for anyone who needs a functional, modern graphics card for light gaming or general desktop use. The GT 645M is for legacy systems or basic display output, where its low power draw and integrated form factor are its only advantages. Its 32 W TDP versus the 75 W of the GTX 1050 Ti makes it a logical choice for ultra-portable or older laptops where power efficiency is critical. However, for any task involving 3D rendering or compute, the GTX 1050 Ti’s 2.138 TFLOPS of FP32 performance dwarfs the GT 645M’s 599.0 GFLOPS, making the choice a foregone conclusion.

FAQ

Q: Which GPU is faster in the Geekbench OpenCL benchmark?

A: The NVIDIA GeForce GTX 1050 Ti is drastically faster, scoring 18129 compared to the GT 645M's 2680. This represents an 85.2% lead for the GTX 1050 Ti, indicating a massive advantage in general-purpose compute workloads.

Q: How do the two cards compare in terms of memory bandwidth?

A: The GTX 1050 Ti has a significant advantage with 112.1 GB/s of bandwidth from its 4 GB of GDDR5 memory. The GT 645M, with its 2 GB of DDR3 memory, offers only 28.80 GB/s, which is a 74.3% deficit and a major bottleneck for modern textures and data.

Q: Are these GPUs similar in their overall performance ranking?

A: Yes, surprisingly so. The GT 645M is in the 26th percentile of all GPUs, while the GTX 1050 Ti is in the 25th percentile. Despite the massive performance gap, both are positioned at the very bottom of the performance spectrum, indicating that the GTX 1050 Ti is an entry-level card in the modern context, just as the GT 645M was in its time.

Q: What is the difference in their manufacturing processes?

A: The GT 645M uses a 28 nm process at TSMC, while the GTX 1050 Ti is built on a more advanced 14 nm process at Samsung. This node difference allows the GTX 1050 Ti to pack 3,300 million transistors into a 132 mm² die, compared to the GT 645M's 1,270 million transistors on a 118 mm² die.

Q: Which card has a higher boost clock speed?

A: The GTX 1050 Ti operates at a much higher frequency, with a boost clock of 1392 MHz. The GT 645M's boost clock is only 780 MHz, which is 44% lower and contributes significantly to the performance disparity.

Q: Does the GTX 1050 Ti support newer DirectX features?

A: Yes. The GTX 1050 Ti supports DirectX 12 (12_1), while the GT 645M is limited to DirectX 12 (11_0). This means the GTX 1050 Ti is compatible with more advanced rendering features and effects in modern games.

Architecture Differences

The architectural gap between the GK107 chip in the GT 645M and the GP107 chip in the GTX 1050 Ti is vast. The GT 645M is built on the Kepler architecture, which debuted in the GeForce 600M generation. It is fabricated on a 28 nm process at TSMC, with a transistor count of 1,270 million on a die size of 118 mm². This results in a transistor density of 10.8M per mm². In contrast, the GTX 1050 Ti uses the Pascal architecture, a much newer design from the GeForce 10-series. It is built on a 14 nm process at Samsung, packing 3,300 million transistors into a 132 mm² die, for a density of 25.0M per mm². This more than doubling of transistor density is the foundation of the GTX 1050 Ti's superiority.

The core configurations also differ dramatically. The GT 645M has 384 shading units, 32 TMUs, and 16 ROPs. The GTX 1050 Ti doubles the shading units to 768, increases TMUs to 48, and doubles ROPs to 32. This directly translates to higher pixel and texture rates. The GT 645M produces 6.240 GPixel/s and 24.96 GTexel/s, while the GTX 1050 Ti produces 44.54 GPixel/s and 66.82 GTexel/s. The FP32 performance tells a similar story: the GT 645M is rated at 599.0 GFLOPS, whereas the GTX 1050 Ti achieves 2.138 TFLOPS. Architecturally, the Pascal GPU also includes support for DirectX 12 (12_1) and Vulkan 1.4, whereas the Kepler part is limited to DirectX 12 (11_0) and Vulkan 1.2.175. Both support OpenGL 4.6, but the Pascal card's feature set is more future-proof.

Specification Differences

The specification sheets reveal a generational leap in nearly every category. The most fundamental difference is the process node, where the GT 645M's 28 nm TSMC process is superseded by the GTX 1050 Ti's 14 nm Samsung process. This allows for a massive increase in transistor count: 1,270 million versus 3,300 million. Clock speeds are also significantly higher on the newer card, with the GTX 1050 Ti's base clock of 1291 MHz and boost of 1392 MHz dwarfing the GT 645M's 709 MHz base and 780 MHz boost.

Memory is another major point of divergence. The GT 645M is equipped with 2 GB of DDR3 memory on a 128-bit bus, yielding 28.80 GB/s of bandwidth. The GTX 1050 Ti features 4 GB of GDDR5 on the same 128-bit bus, but its much higher effective clock speed of 7 Gbps results in 112.1 GB/s of bandwidth. Power consumption also differs, with the GT 645M drawing only 32 W and the GTX 1050 Ti drawing 75 W. The GT 645M is an integrated graphics package (IGP) with no power connectors, while the GTX 1050 Ti is a dual-slot card that, although it also has no power connectors, requires a suggested PSU of 250 W. The GTX 1050 Ti is a physical card with dimensions of 145 mm in length and 111 mm in height, while the GT 645M is portable device dependent. Finally, the GTX 1050 Ti has a launch MSRP of 139 USD, while the GT 645M has no listed launch price.

Head-to-Head Benchmarks

The benchmark results are a one-sided affair, with the GTX 1050 Ti winning all three tests. The most extreme victory is in Geekbench OpenCL, where the GTX 1050 Ti scores 18129 versus the GT 645M's 2680. This is a deltaPct of -85.2%, meaning the GT 645M is a staggering 85.2% slower than its rival. This is a compute-heavy test, and the Pascal architecture's superior FP32 throughput and memory bandwidth are fully utilized here.

In Geekbench Vulkan, the GTX 1050 Ti also shows a dominant performance, scoring 10001 compared to the GT 645M's 4875. The deltaPct here is -51.3%, indicating the GT 645M is just over half as fast. This test highlights the newer card's ability to handle modern graphics APIs efficiently. The closest competition is in Geekbench Metal, where the GTX 1050 Ti scores 7834 and the GT 645M scores 5679. Even here, the deltaPct is -27.5%, a substantial gap that still places the GTX 1050 Ti firmly ahead. The data shows a clear pattern: the GTX 1050 Ti is not just faster, but often two to three times faster than the GT 645M in the same workloads.

Where Each One Wins

Based on the data, the GTX 1050 Ti wins everywhere that matters for performance. It is the definitive choice for any 3D application, including gaming, 3D modeling, or video editing. Its 2.138 TFLOPS of FP32 compute, 112.1 GB/s of memory bandwidth, and support for DirectX 12 (12_1) make it a capable performer for entry-level modern gaming. Its 4 GB of GDDR5 memory is also double the GT 645M's 2 GB, which allows it to handle larger textures and higher resolutions without running out of memory. In the head-to-head benchmarks, it wins 3 out of 3, making it the only rational choice for any workload that requires actual graphics processing.

The GT 645M's only advantages are its power efficiency and form factor. With a TDP of 32 W, it consumes less than half the power of the GTX 1050 Ti's 75 W. This makes it suitable for thin-and-light laptops where heat and battery life are primary concerns. As an IGP, it requires no extra space or cooling, making it a viable option for basic productivity tasks, video playback, and legacy applications. Its performance is adequate for the 26th percentile of GPUs, but the data shows that its performance is severely limited in modern compute benchmarks, with an 85.2% deficit in OpenCL. The GT 645M is a card for a bygone era, whereas the GTX 1050 Ti, despite its low percentile ranking today, offers a baseline level of performance that is still relevant for casual use.

DETAILED SPECIFICATIONS

SPECIFICATION
GT 645M
GTX 1050 Ti
Core Specs
Shading Units
384
768 +100.0%
Shaders
384
768 +100.0%
TMUs
32
48 +50.0%
ROPs
16
32 +100.0%
SM Count
—
6
Clocks
Base Clock
709 MHz
1291 MHz
Boost Clock
780 MHz
1392 MHz
Memory Clock
900 MHz 1800 Mbps effective
1752 MHz 7 Gbps effective
Memory
Memory Size
2 GB
4 GB
VRAM (MB)
2,048
4,096 +100.0%
Memory Type
DDR3
GDDR5
Memory Bus
128 bit
128 bit
Bandwidth
28.80 GB/s
112.1 GB/s
Cache
L1 Cache
16 KB (per SMX)
48 KB (per SM)
L2 Cache
256 KB
1024 KB
Performance
Pixel Rate
6.240 GPixel/s
44.54 GPixel/s
Texture Rate
24.96 GTexel/s
66.82 GTexel/s
FP32 (TFLOPS)
599.0 GFLOPS
2.138 TFLOPS
FP64 (TFLOPS)
24.96 GFLOPS (1:24)
66.82 GFLOPS (1:32)
FP16 (TFLOPS)
—
33.41 GFLOPS (1:64)
Power
TDP
32 W
75 W
TDP (W)
32
75 +134.4%
Suggested PSU
—
250 W
Power Connectors
None
None
Architecture
Architecture
Kepler
Pascal
GPU Name
GK107
GP107
Generation
GeForce 600M
GeForce 10
Process Size
28 nm
14 nm
Transistors
1,270 million
3,300 million
Die Size
118 mm²
132 mm²
Foundry
TSMC
Samsung
Density
10.8M / mm²
25.0M / mm²
API Support
DirectX
12 (11_0)
12 (12_1)
OpenGL
4.6
4.6
Vulkan
1.2.175
1.4
OpenCL
3.0
3.0
CUDA
3.0
6.1
Shader Model
6.5 (5.1)
6.8
Physical
Slot Width
IGP
Dual-slot
Length
—
145 mm 5.7 inches
Height
—
111 mm 4.4 inches
Outputs
Portable Device Dependent
1x DVI1x HDMI 2.01x DisplayPort 1.4a
Bus Interface
PCIe 3.0 x16
PCIe 3.0 x16
Other
Launch Price
—
139 USD
Production
End-of-life
End-of-life
Predecessor
GeForce 500M
GeForce 900
Successor
GeForce 700M
GeForce 20
View GeForce GT 645M Details View GeForce GTX 1050 Ti Details