The NVIDIA RTX 3090 is a consumer graphics card based on the Ampere architecture. Released on 2020-09-24 with a launch price of $1,499.00, it features 10496 CUDA cores and 24 GB of GDDR6X memory.
Architecture | Ampere (GA102) |
---|---|
Market Segment | Consumer |
Release Date | 2020-09-24 |
Launch Price | $1,499.00 |
Manufacturing Process | 8nm |
CUDA Cores | 10496 |
---|---|
Base Clock Speed | 1395 MHz |
Boost Clock Speed | 1695 MHz |
Transistor Count | 28.3B |
VRAM Capacity | 24 GB GDDR6X |
Memory Bus Width | 384 bits |
Memory Bandwidth | 936.0 GB/s |
TDP | 350 W |
Key technical parameters of the NVIDIA RTX 3090 include its 10496 CUDA cores, 1395 MHz base clock, and 1695 MHz boost clock, delivering high performance for consumer applications.
FP32 (float) | 35.6 TFLOPS |
---|---|
FP16 (half) | 71.2 TFLOPS |
The NVIDIA RTX 3090 achieves 35.6 TFLOPS in FP32 performance, making it a top contender in consumer workloads.
1080p | 220 FPS |
---|---|
4K | 100 FPS |
In gaming, the NVIDIA RTX 3090 delivers 220 FPS at 1080p and 100 FPS at 4K, showcasing its capability for high-resolution gaming.
Tensor_cores | Gen 3 |
---|---|
Ports | HDMI 2.1, DisplayPort 1.4a x3 |
Length | 12.3in |
---|---|
Width | 5.4in |
Height | 3-slot |
Ranking Position | #8 |
---|---|
Popularity Ranking | #2 |
Cost Effectiveness | 2.2/5 |
Power Efficiency | 1.2/5 |
Category | Rank 1 | Rank 2 | Rank 3 |
---|---|---|---|
Best for Training | NVIDIA H200 | NVIDIA H100 | NVIDIA B200 |
Best for Inference | NVIDIA A40 | NVIDIA A100 | NVIDIA A10 |