47score
#48 of 131
Overall rank
NVIDIAMid-range

TITAN V

1.9 · 370 votes
Best for 1080p / 1440p gaming

As an Amazon Associate we may earn from qualifying purchases.

VRAM
12 GB
CUDA
5,120
FP32
14.9 TF
Bandwidth
651.3 GB/s
TDP
250W
Boost
1455 MHz
Strengths at a Glance

How it stacks up to the flagship

Each metric is shown as a percentage of the GeForce RTX 5090 D, the strongest card we track.

FP32 compute14.9 TFLOPS14%
vs RTX 5090 D: 104.8 TFLOPS
Memory bandwidth651.3 GB/s36%
vs RTX 5090 D: 1790 GB/s
VRAM capacity12 GB38%
vs RTX 5090 D: 32 GB
Shading units5,12024%
vs RTX 5090 D: 21,760
Power efficiency19/10033%
vs RTX 5090 D: 58/100
Synthetic Benchmarks

Estimated benchmark results

Each result is shown as a share of the RTX 5090 D's score in the same test.

3DMark Time Spy16,920pts
vs RTX 5090 D: 36,000 pts
3DMark Port Royal (RT)8,640pts
vs RTX 5090 D: 18,000 pts
Blender (samples/min)2,444spm
vs RTX 5090 D: 5,200 spm
Geekbench Compute117,600pts
vs RTX 5090 D: 240,000 pts

Performance breakdown

Gaming47
Ray tracing48
AI / Compute49
Creator / 3D47
Power efficiency19
Real-World Gaming

FPS Across Resolutions

1080p · Ultra preset1440p · Ultra preset4K · Ultra preset
Cyberpunk 2077avg 41 fps
1080p
55
1440p
41
4K
26
Call of Duty: MW IIIavg 57 fps
1080p
78
1440p
57
4K
36
Alan Wake 2avg 33 fps
1080p
45
1440p
33
4K
21
Forza Horizon 5avg 59 fps
1080p
81
1440p
60
4K
37
Baldur's Gate 3avg 48 fps
1080p
66
1440p
49
4K
30
Community Feedback

What Owners Say

Owners love the raw compute power for deep learning and rendering. The usual gripe is the high power draw and noise under load.

Pros
  • Eats professional 3D workloads for breakfast
  • Silent under heavy rendering loads
  • Handles massive VRAM datasets easily
  • Stays cool with its vapor chamber
Cons
  • No consumer gaming software support
  • Huge power draw for its era
  • Cooling fan noise under load

Supported technologies

DLSSNVENCAV1 Encode

Full specifications

Graphics processor
Architecture
Volta
Process node
12 nm
Transistors
21.1 B
SM Count
40
Release date
2017
Launch price
$2999
Core configuration
CUDA Cores
5,120
Tensor Cores
640
TMUs
320
ROPs
96
L2 cache
4.5 MB
Memory
Size
12 GB
Type
HBM2
Bus width
3072-bit
Bandwidth
651.3 GB/s
Memory clock
848 MHz
Clocks & throughput
Base clock
1200 MHz
Boost clock
1455 MHz
FP32 (float)
14.9 TFLOPS
FP16 (half)
14.9 TFLOPS
Pixel rate
140 GPixel/s
Texture rate
465.6 GTexel/s
Board & power
TDP
250W
Suggested PSU
500W
Power connectors
1x 6-pin + 1x 8-pin
Bus interface
PCIe 3.0 x16
Length
267 mm
Slot width
2-slot
Display & outputs
Max resolution
7680×4320
Outputs
1x HDMI, 3x DisplayPort

API and SDK support

DirectX
12 (12_1)
Shader Model
6.4
OpenGL
4.6
OpenCL
1.2
Vulkan
+
CUDA
7.0
Verdict

Our verdict on the TITAN V

NVIDIA's TITAN V is a high-end Volta card built for compute work, hampered by its massive price and lack of gaming-focused drivers.

Get it if you need a GPU for serious scientific computing or AI research without the business restrictions of a Quadro. Skip it if you’re gaming or rendering—a newer card gives you better performance for far less cash.

Buy it if…

  • You need top-tier compute for scientific or AI workloads.
  • You want the best single-GPU for VR development.
  • You are a pro creator who must have maximum VRAM.
1.9

370 votes

Rate this GPU

Add your verdict