The specifications of NVIDIA’s upcoming RTX 4080 graphics card have been shared by Kopite7kimi. It looks like the defacto Lovelace gaming GPU will be a smidge slower than initially anticipated rocking 9,728 FP32 cores alongside 16GB of GDDR6X memory across a 256-bit bus. NVIDIA has disabled 1,024 FP32 cores or 8 SMs on the AD103 die, leaving room for an RTX 4080 Ti on the same possibly with a wider bus.
| GPU | GA102 | AD102 | RTX 4090 | AD103 | RTX 4080 | RTX 4070 Ti (AD104) | RTX 4070 |
|---|---|---|---|---|---|---|---|
| Arch | Ampere | Ada Lovelace | Ada Lovelace | Ada Lovelace | |||
| Process | Sam 8nm LPP | TSMC 5nm | TSMC 5nm | TSMC 5nm | |||
| GPC | 7 | 12 | 11 | 7 | 7 | 5 | 5 |
| TPC | 42 | 72 | 64 | 42 | 40 | 30 | 30 |
| SMs | 84 | 144 | 128 | 84 | 80 | 60 | 60 |
| Shaders | 10,752 | 18,432 | 16,384 | 10,752 | 9,728 | 7,680 | 7,680 |
| TP | 37.6 | ~100 TFLOPs? | 83 TFLOPs | ~50 TFLOPs | 47 TFLOPs? | ~35 TFLOPs | 35 TFLOPs? |
| Memory | 24GB GDDR6X | 48GB GDDR6X | 24GB GDDR6X | 16GB GDDR6X | 12GB GDDR6X | ||
| L2 Cache | 6MB | 96MB | 72MB | 64MB | 48MB | ||
| Bus Width | 384-bit | 384-bit | 256-bit | 192-bit | |||
| TGP | 350W | 600W | 450W | 450W | 285-340W | 300W | 285W |
| Launch | Sep 2020 | Sept 22? | Sept 22? | Q1 2023? |
The GPU core will likely be running close to 3GHz with the memory clocks set to 21Gbps. The RTX 4080 is expected to score 15,000 points in 3DMark TimeSpy Extreme, putting it approximately 70% faster than the RTX 3080. It will have a TGP of 420W, exactly 100W more than its predecessor. It’s worth noting that the RTX 3080 features 8,704 shaders and a 320-bit bus alongside a 10GB GDDR6X memory buffer, so most of the performance gains this generation are coming from the increased IPC, clocks, and cache (and power!). The RTX 4080 is slated to feature as much as 48MB of L2 cache.
The same source shared the specifications of the RTX 4070 the other day. You can go through them below:
