Nvidia RTX 4080 vs Nvidia RTX 3080
The NVIDIA RTX 4080 is a next-generation flagship graphics card released by NVIDIA in November 2022. Built on the Ada Lovelace architecture and manufactured using TSMC's 4N process, it boasts 45.9 billion transistors and comes equipped with 16GB of GDDR6X memory, delivering an impressive memory bandwidth of 760 GB/s. Compared to its predecessor, the RTX 3080, the RTX 4080 offers a significant performance boost, with FP32 floating-point performance jumping from 29.8 TFLOPS to 40 TFLOPS, an improvement of 34%. Additionally, the RTX 4080 features third-generation RT Cores and fourth-generation Tensor Cores, providing enhanced ray tracing and AI capabilities.
Nvidia RTX 4080
Pros:
Significantly faster performance, offering a 30-50% uplift over the RTX 3080
More features, such as DLSS 3 and improved ray tracing
Higher VRAM capacity (16GB vs 10GB)
Cons:
Very expensive
High power consumption (320W vs 320W)
Large physical size
Nvidia RTX 3080
Pros:
Still a very powerful graphics card that will handle most games with ease
Significantly cheaper than the RTX 4080
Lower power consumption
Smaller physical size
Cons:
Not as fast as the RTX 4080
Less VRAM capacity
Lacks some newer features, such as DLSS 3 and improved ray tracing
Graphics core
| Release Date | November 16, 2022 | September 17, 2020 |
| Generation | Ada Lovelace | Ampere |
| Applicable Type | Desktop | Desktop |
| Processor Number | AD103-300 | GA102-200 |
| Bus Interface | PCI Express 4.0 x16 | PCI Express 4.0 x16 |
Memory
| Type | GDDR6X | GDDR6X |
| Capacity | 16GB | 10GB |
| Bus Width | 256-bit | 320-bit |
| Bandwidth | 760.3 GB/s | 760 GB/s |
Clock speed
| Base Clock | 2205 MHz | 1440 MHz |
| Boost Clock | 2505 MHz | 1710 MHz |
Render Config
| CUDA Cores | 9728 | 8704 |
| Tensor Cores | 3072 | 272 |
| RT Cores | 112 | 68 |
Theoretical Performance
| FP32 | 49.6 TFLOPS | 30.1 TFLOPS |
| FP64 | 7.8 TFLOPS | 4.7 TFLOPS |
| Tensor | 1000 TFLOPS | 238 TFLOPS |
Graphics Processor
| Die Size | 378.7 mm² | 628 mm² |
| Transistors | 45.9 billion | 28.3 billion |
| Manufacturing Process | TSMC 4N | Samsung 8N |
Board Design
| Form Factor | ATX | ATX |
| Length | 304 mm | 285 mm |
| Width | 137 mm | 112 mm |
| Height | 41 mm | 40 mm |
| Power Connectors | 1x 16-pin PCIe Gen5 | 2x 8-pin PCIe |
| TDP | 320W | 320W |