NVIDIA · Kepler
NVIDIA GeForce GTX 650 Ti: specs, benchmarks and comparison
- 1 GB GDDR5
- 110 W
- Launch price
- $149
- MSRP, US
- Release
- October 9, 2012
Performance
Rank among models tested in the same benchmark on the same render backend.
Blender Benchmark 2.93.1
GPU, bmw27 (CUDA)
559.68 s
lower is better
Median of 1 runs · small sample · Blender Open Data (Blender Foundation) · November 14, 2026
Other benchmark versions (5)
| Benchmark | Result | Rank |
|---|---|---|
| Blender Benchmark 2.92 | 537.1 s · small sample | — |
| Blender Benchmark 2.90 | 537.82 s · small sample | — |
| Blender Benchmark 2.83 | 555 s · small sample | — |
| Blender Benchmark 2.82 | 534.22 s · small sample | — |
| Blender Benchmark 2.81a | 552.53 s · small sample | — |
Each benchmark version has its own scale; results from different versions are not compared.
Blender Benchmark 2.92
GPU, classroom (CUDA)
1,917.2 s
lower is better
Median of 2 runs · small sample · Blender Open Data (Blender Foundation) · November 14, 2026
Other benchmark versions (3)
| Benchmark | Result | Rank |
|---|---|---|
| Blender Benchmark 2.90 | 1,949.35 s · small sample | — |
| Blender Benchmark 2.82 | 1,815.49 s · small sample | — |
| Blender Benchmark 2.81a | 1,929.48 s · small sample | — |
Each benchmark version has its own scale; results from different versions are not compared.
Blender Benchmark 2.92
GPU, fishy_cat (CUDA)
1,321.03 s
lower is better
Median of 1 runs · small sample · Blender Open Data (Blender Foundation) · November 14, 2026
Other benchmark versions (1)
| Benchmark | Result | Rank |
|---|---|---|
| Blender Benchmark 2.82 | 1,077.94 s · small sample | — |
Each benchmark version has its own scale; results from different versions are not compared.
Blender Benchmark 2.92
GPU, koro (CUDA)
1,995.73 s
lower is better
Median of 1 runs · small sample · Blender Open Data (Blender Foundation) · November 14, 2026
Other benchmark versions (1)
| Benchmark | Result | Rank |
|---|---|---|
| Blender Benchmark 2.82 | 1,945.96 s · small sample | — |
Each benchmark version has its own scale; results from different versions are not compared.
Blender Benchmark 2.92
GPU, pavillon_barcelona (CUDA)
4,047.48 s
lower is better
Median of 1 runs · small sample · Blender Open Data (Blender Foundation) · November 14, 2026
Other benchmark versions (1)
| Benchmark | Result | Rank |
|---|---|---|
| Blender Benchmark 2.82 | 4,025.66 s · small sample | — |
Each benchmark version has its own scale; results from different versions are not compared.
Blender Benchmark 3.4.0
GPU, sum of scenes monster + junkshop + classroom (CUDA)
42 samples/min
higher is better
Median of 2 runs · small sample · Blender Open Data (Blender Foundation) · November 14, 2026
Other benchmark versions (2)
| Benchmark | Result | Rank |
|---|---|---|
| Blender Benchmark 3.3.0 | 92 samples/min · small sample | — |
| Blender Benchmark 3.1.0 | 50 samples/min · small sample | — |
Each benchmark version has its own scale; results from different versions are not compared.
Specifications
Reference values for the chip. Clocks, length and connectors depend on the card model.
Chip
- GPU chip
- GK106-220-A1
- Architecture
- Kepler
- Shader units
- 768 (CUDA cores)
Clocks
- Base clock
- 928 MHz
Memory
- Video memory
- 1 GB
- Memory type
- GDDR5
- Memory bus
- 128 bit
- Memory bandwidth
- 86.4 GB/s
Power and dimensions
- Board power
- 110 W (TDP)
- Interface
- PCIe 3.0 x16
Specifications: Wikipedia (English), checked November 14, 2026. Release date and launch price: Wikipedia (CC BY-SA 4.0).
Will it fit your system
Checks the power supply, case length and PCIe slot.
- Needs checkingPower supply. The manufacturer did not list a recommended PSU wattage (or it is not in the database).
- Needs checkingCard length and case. Length depends on the specific partner model (ASUS, MSI, Palit and others), so pick the exact SKU.
- Needs checkingPCIe interface. Enter the PCIe version of the motherboard slot.