NVIDIA · Turing · 2018
T4 GPU accelerator (PCIe card)
T4 GPU accelerator (PCIe card): 16 GB GDDR6, 70 W, 2,560 CUDA cores. Launched September 12, 2018.
- Memory
- 16 GB
- Power
- 70 W
- Bus width
- 256 bit
Memory16 GB
13 of 32 in Data Center
Power70 W
40 of 40 by higher draw in Data Center
CUDA cores2,560
22 of 47 in Data Center
Current market
Affiliate links — we may earn a commission.
Specifications
Memory
- Capacity
- 16 GB
- Memory type
- GDDR6
- Bus width
- 256 bit
- Bandwidth
- 320 GB/s
Compute
- CUDA cores
- 2,560
- SM count
- 40
- TMUs
- 160
- ROPs
- 64
- FP32
- 8,100 GFLOPS
Clocks
- Base clock
- 585 MHz
- Boost clock
- 1,590 MHz
Board
- TDP
- 70 W
- Form factor
- datacenter
Physical
- Architecture
- Turing
- Series
- Data Center
- Launch date
- September 12, 2018
Not recorded by the source: generation, codename, process and transistors, or 8 further fields.
Die figure
Silicon, made visible
Die area is drawn against the largest recorded die in the catalog. Shader dots use a labeled, card-specific ratio so both early and modern GPUs remain legible.
Catalog context
Within its series, the T4 GPU accelerator (PCIe card) ranks 16th of 36 for FP32 throughput and 13th of 32 for memory capacity. Its 70 W TDP is within the 75 W PCIe slot-power ceiling.
Source: Wikipedia, revision 1369488585 · catalog loaded 2026-08-16