Similar to the RTX 4070 specifications from yesterday, we now have a specifications update for the NVIDIA GeForce RTX 4080 graphics card too.
NVIDIA GeForce RTX 4080 Allegedly Gets Spec Bump: 9728 Cores, 16 GB GDDR6X Memory, 420W TBP, Around 30% Faster Than 3090 Ti
The same leaker, Kopite7kimi, had previously posted a slightly better core configuration for the RTX 4080 but it should be pointed that NVIDIA makes several revision changes over a product's pre-launch cycle & as such, the specifications change a lot which Kopite7kimi is simply reporting. The card now has the same amount of cores and memory albeit at a much lower TGP than what was reported earlier.
I'm not a chatterbox, but I have to make some updates. I hope you don't mind.
a possible RTX 4080,
PG136/139-SKU360
AD103-300-A1
9728FP32
256bit 16G 21Gbps GDDR6X
total power ~420W
TSE ~15000
Now I have completed the latest update for 4090, 4080 and 4070.— kopite7kimi (@kopite7kimi) August 5, 2022
NVIDIA GeForce RTX 4080 'Expected' Specifications
The NVIDIA GeForce RTX 4080 is expected to utilize a cut-down AD103-300 GPU configuration with 9,728 cores or 76 SMs enabled of the total 84 units whereas the previous configuration offered 80 SMs or 10,240 cores. While the full GPU comes packed with 64 MB of L2 cache and up to 224 ROPs, the RTX 4080 might end up with 48 MB of L2 cache and lower ROPs too due to its cut-down design.
The clock speeds are not confirmed yet but considering that the TSMC 4N process is being used, we are expecting clocks between the 2.0-3.0 GHz range. The higher than usual clock speed bump comes from the fact that NVIDIA is making a two-node jump considering the Ampere GPUs with Samsung 8nm node was in reality a 10nm process node with some optimizations. NVIDIA is skipping 7nm and going straight for a 5nm node and not even the vanilla variant but an optimized version of it. With Pascal on the TSMC 16nm node, NVIDIA delivered a huge frequency leap and we can expect a similar jump this time around too.
As for memory specs, the GeForce RTX 4080 is expected to rock 16 GB GDDR6X capacities that are said to be adjusted at 21 Gbps speeds across a 256-bit bus interface. This will provide up to 672 GB/s of bandwidth. For power, the TBP is said to be rated at around 420W which is close to the RTX 4090 and since the memory clocks aren't that aggressive, it is likely that the core clocks could be tuned for maximum graphics performance.
In terms of performance, the NVIDIA GeForce RTX 4080 is expected to offer 15,000 points in the 3DMark Time Spy benchmark which the leaker had previously stated as a conservative figure so one can expect around a 35% performance uplift over the RTX 3090 Ti in this specific benchmark.
3DMark Time Spy Extreme Graphics
As for its feature set, the NVIDIA GeForce RTX 4080 graphics cards will rock all the modern NV feature sets such as the latest 4th Gen Tensor Cores, 3rd gen RT cores, the latest NVENC Encoder, and NVCDEC Decoder, and support for the latest APIs. They will pack all the modern RTX features such as DLSS, Reflex, Broadcast, Resizable-BAR, Freestyle, Ansel, Highlights, Shadowplay, and G-SYNC support too. The graphics card should launch around Q4 2022 or may even slip too Q1 '23 but it should come priced at around $699-$799 US.
NVIDIA GeForce RTX 40 Series Official Specs:
| Graphics Card Name | NVIDIA GeForce RTX 4090 | NVIDIA GeForce RTX 4080 | NVIDIA GeForce RTX 4070 Ti | NVIDIA GeForce RTX 4070 |
|---|---|---|---|---|
| GPU Name | Ada Lovelace AD102-300 | Ada Lovelace AD103-300 | Ada Lovelace AD104-400 | Ada Lovelace AD104-250 |
| Process Node | TSMC 4N | TSMC 4N | TSMC 4N | TSMC 4N |
| Die Size | 608mm2 | 378.6mm2 | 294.5mm2 | 294.5mm2 |
| Transistors | 76 Billion | 45.9 Billion | 35.8 Billion | 35.8 Billion |
| CUDA Cores | 16384 | 9728 | 7680 | 5888 |
| TMUs / ROPs | 512 / 176 | 320 / 112 | 240 / 80 | 184 / 64 |
| Tensor / RT Cores | 512 / 128 | 304 / 76 | 240 / 60 | 184 / 46 |
| Base Clock | 2230 MHz | 2210 MHz | 2310 MHz | 1920 MHz |
| Boost Clock | 2520 MHz | 2510 MHz | 2610 MHz | 2475 MHz |
| FP32 Compute | 83 TFLOPs | 49 TFLOPs | 40 TFLOPs | 29 TFLOPs |
| RT TFLOPs | 191 TFLOPs | 113 TFLOPs | 82 TFLOPs | 67 TFLOPs |
| Tensor-TOPs | 1321 TOPs | 780 TOPs | 641 TOPs | 466 TOPs |
| Memory Capacity | 24 GB GDDR6X | 16 GB GDDR6X | 12 GB GDDR6X | 12 GB GDDR6X |
| Memory Bus | 384-bit | 256-bit | 192-bit | 192-bit |
| Memory Speed | 21.0 Gbps | 23.0 Gbps | 21.0 Gbps | 21.0 Gbps |
| Bandwidth | 1008 GB/s | 736 GB/s | 504 GB/s | 504 GB/s |
| TBP | 450W | 320W | 285W | 200W |
| Price (MSRP / FE) | $1599 US / 1949 EU | $1199 US / 1469 EU | $799 US | $599 US |
| Price (Current) | $1599 US / 1859 EU | $1199 US / 1399 EU | $799 US | $599 US |
| Launch (Availability) | 12th October 2022 | 16th November 2022 | 5th January 2023 | 13th April 2023 |
