Performance to price ratio. The higher, the better.
NVIDIA Tesla T4: specs and benchmarks
Summary
NVIDIA started Tesla T4 sales 13 September 2018. This is a Turing architecture desktop card based on 12 nm manufacturing process and primarily aimed at designers. 16 GB of GDDR6 memory clocked at 10 GHz are supplied, and together with 256 Bit memory interface this creates a bandwidth of 320.0 GB/s.
Compatibility-wise, this is single-slot card attached via PCIe 3.0 x16 interface. Its manufacturer default version has a length of 168 mm. No additional power connector is required, and power consumption is at 70 Watt.
It provides good gaming and benchmark performance at 27.55% of a leader's which is NVIDIA GeForce RTX 4090.
General info
Some basic facts about Tesla T4: architecture, market segment, release date etc.
Place in performance ranking | 164 | |
Value for money | 18.02 | |
Architecture | Turing (2018−2021) | |
GPU code name | TU104 | |
Market segment | Workstation | |
Release date | 13 September 2018 (5 years old) | |
Current price | $782 | of 168889 (A100 PCIe 80 GB) |
Technical specs
Tesla T4's specs such as number of shaders, GPU base clock, manufacturing process, texturing and calculation speed. These parameters indirectly speak of Tesla T4's performance, but for precise assessment you have to consider its benchmark and gaming test results.
Pipelines / CUDA cores | 2560 | of 20480 (Data Center GPU Max NEXT) |
Core clock speed | 585 MHz | of 2610 (Radeon RX 6500 XT) |
Boost clock speed | 1590 MHz | of 3599 (Radeon RX 7990 XTX) |
Number of transistors | 13,600 million | of 14400 (GeForce GTX 1080 SLI Mobile) |
Manufacturing process technology | 12 nm | of 4 (GeForce RTX 4080) |
Power consumption (TDP) | 70 Watt | of 2400 (Data Center GPU Max Subsystem) |
Texture fill rate | 254.4 | of 969.9 (H100 SXM5 96 GB) |
Size and compatibility
This section provides details about the physical dimensions of Tesla T4 and its compatibility with other computer components. This information is useful when selecting a computer configuration or upgrading an existing one. For desktop graphics cards, it includes details about the interface and bus (for motherboard compatibility) and additional power connectors (for power supply compatibility).
Interface | PCIe 3.0 x16 | |
Length | 168 mm | |
Width | 1-slot | |
Supplementary power connectors | None |
Memory
Parameters of memory installed on Tesla T4: its type, size, bus, clock and resulting bandwidth. Note that GPUs integrated into processors have no dedicated memory and use a shared part of system RAM instead.
Memory type | GDDR6 | |
Maximum RAM amount | 16 GB | of 128 (Radeon Instinct MI250X) |
Memory bus width | 256 Bit | of 19000 (GeForce RTX 3080 12 GB) |
Memory clock speed | 10000 MHz | of 22400 (GeForce RTX 4080) |
Memory bandwidth | 320.0 GB/s | of 3276 (Aldebaran) |
Video outputs and ports
Types and number of video connectors present on Tesla T4. As a rule, this section is relevant only for desktop reference graphics cards, since for notebook ones the availability of certain video outputs depends on the laptop model, while non-reference desktop models can (though not necessarily will) bear a different set of video ports.
Display Connectors | No outputs |
API support
APIs supported by Tesla T4, sometimes including their particular versions.
DirectX | 12 Ultimate (12_1) | |
Shader Model | 6.5 | |
OpenGL | 4.6 | |
OpenCL | 1.2 | |
Vulkan | 1.2.131 | |
CUDA | 7.5 |
Benchmark performance
Synthetic benchmark performance of Tesla T4. The combined score is measured on a 0-100 point scale.
Combined synthetic benchmark score
This is our combined benchmark performance score. We are regularly improving our combining algorithms, but if you find some perceived inconsistencies, feel free to speak up in comments section, we usually fix problems quickly.
Passmark
This is probably the most ubiquitous benchmark, part of Passmark PerformanceTest suite. It gives the graphics card a thorough evaluation under various load, providing four separate benchmarks for Direct3D versions 9, 10, 11 and 12 (the last being done in 4K resolution if possible), and few more tests engaging DirectCompute capabilities.
Benchmark coverage: 25%
GeekBench 5 OpenCL
Geekbench 5 is a widespread graphics card benchmark combined from 11 different test scenarios. All these scenarios rely on direct usage of GPU's processing power, no 3D rendering is involved. This variation uses OpenCL API by Khronos Group.
Benchmark coverage: 9%
GeekBench 5 CUDA
Geekbench 5 is a widespread graphics card benchmark combined from 11 different test scenarios. All these scenarios rely on direct usage of GPU's processing power, no 3D rendering is involved. This variation uses CUDA API by NVIDIA.
Benchmark coverage: 4%
Send your test results of Tesla T4.
Gaming performance
Let's see how good Tesla T4 is for gaming. Particular gaming benchmark results are measured in frames per second. Comparisons with game system requirements are included, but remember that sometimes official requirements may reflect reality inaccurately.
Relative perfomance
Overall Tesla T4 performance compared to nearest competitors among server video cards.
AMD equivalent
According to our data, the closest AMD alternative to Tesla T4 is Radeon Pro SSG, which is slower by 2% and lower by 1 position in our ranking.
Here are some closest AMD rivals to Tesla T4:
Similar GPUs
Here is our recommendation of several graphics cards that are more or less close in performance to the one reviewed.
Recommended processors
These processors are most commonly used with Tesla T4 according to our statistics.