The GeForce RTX 4080 successor, NVIDIA’s GeForce RTX 5080 “Blackwell” GPU, is expected to launch soon. The RTX 5080 “Blackwell” GPU will be an enthusiast-grade option from NVIDIA.
NVIDIA GeForce RTX 5080 Graphics Card Performance, Specs, Price – Everything We Know As Of Now
The next generation GPU for NVIDIA will be the GeForce RTX 5080 that is based on Blackwell architecture. As the next line of the Nvidia graphic cards after the flagship RTX 5090, it becomes the second most powerful card of the RTX 50 series, replacing the present RTX 4080 and RTX 4080 SUPER.
The RTX 50 series once again brings in a new architecture from NVIDIA dubbed Blackwell, intended to support both gaming and data center platforms. Being built upon a new custom 5nm process node called N5B, these GPUs are said to bring important improvements in both performance and efficiency figures.
The GeForce RTX 5080 aims to surpass the Ada-based RTX 4080 in terms of standard graphical performance. NVIDIA’s 80-class GPUs have always been considered enthusiast-grade, traditionally offering top performance per dollar until the introduction of the 90-class with the RTX 30 series. With the RTX 5080, NVIDIA continues to demonstrate its ability to set the benchmark for performance in gaming and creative applications.
NVIDIA GeForce RTX 5080 Graphics Card Internal DIE Specification
As for the specs, the NVIDIA GeForce RTX 5080 is expected to use the GB203-300-A1 GPU die and is known to replace the AD103-300-A1 GPU of RTX 4080. It is expected to come with the PG144/147-SKU45 PCB layout and contain 84 Streaming Multiprocessors (SMs), which is really a significant boost in core configuration from the previous model.
NVIDIA GeForce RTX 5080 Graphics Card Rumored Key Specific List
- CUDA Cores: 10,752 (10.5% increase from RTX 4080’s 9,728 CUDA cores).
- Streaming Multiprocessors (SMs): 84 (up from 76 on the RTX 4080).
- Memory Type: GDDR7 (upgraded from GDDR6X on RTX 4080).
- Memory Size: 16 GB VRAM.
- Memory Bus: 256-bit.
- Memory Speed: Up to 32 Gbps.
- Total Memory Bandwidth: 1 TB/s (ahead of the RTX 4090’s 1007 GB/s).
Graphic Card Memory Evolution From GDDR5X To GDDR7
| GRAPHICS MEMORY | GDDR7 | GDDR6X | GDDR6 | GDDR5X |
|---|---|---|---|---|
| Workload Type | Gaming / AI | Gaming / AI | Gaming / AI | Gaming |
| Platform | GeForce RTX 5090? | GeForce RTX 4090 | GeForce RTX 2080 Ti | GeForce GTX 1080 Ti |
| Die Capacity (Gb) | 16-64 | 8-32 | 8-32 | 8-16 |
| Number of Placements | 12? | 12 | 12 | 12 |
| Gb/s/pin | 32-42.5 | 19-24 | 14-16 | 11.4 |
| GB/s/placement | 128-144 | 76-96 | 56-64 | 45 |
| GB/s/system | 1536-1728 | 912-1152 | 672-768 | 547 |
| Configuration (Example) | 384 IO (12pcs x 32 IO package)? | 384 IO (12pcs x 32 IO package) | 384 IO (12pcs x 32 IO package) | 384 IO (12pcs x 32 IO package) |
| Frame Buffer of Typical System | 24 GB (16 Gb) 36 GB (24 Gb) | 24 GB | 12GB | 12GB |
| Module Package | 266 (BGA) | 180 (BGA) | 180 (BGA) | 190 (BGA) |
| Average Device Power (pJ/bit) | TBD | 7.25 | 7.5 | 8.0 |
| Typical IO Channel | PCB (P2P SM) | PCB (P2P SM) | PCB (P2P SM) | PCB (P2P SM) |
The NVIDIA GeForce RTX 5080 is expected to be much faster in memory than the RTX 5090 with a memory bandwidth of about 43% greater than that of the RTX 4080. There are also rumors that two more versions of the card could be on the way soon with 24GB of vRAM (3GB per module). However, the first iteration of the launch model is thought to offer 16GB of VRAM only. A comparison of the power, then, indicates that the RTX 5080 will be more powerful than the RTX 4080 in terms of Total Board Power (TBP). At 400W it takes 80W more from the power supply than the 320W of the RTX 4080. This basically means that TBP which is the maximum power the board can deliver, but generally is lower in real-wold usage as we saw with the RTX 40 series graphics card. RTX 5080 will be using a single 12V-2×6 power connector which can have maximum power of 600W.
| Specifications | GeForce RTX 5080 | GeForce RTX 4080 |
|---|---|---|
| Architecture | Blackwell | Ada Lovelace |
| Process Node | 4NP(TSMC’s 5nm) | 4N(TSMC’s 5nm) |
| GPU | GB203-400-A1 | AD103-300-A1 |
| Cuda Cores | 10752 | 9728 |
| SM | TBD | 76 |
| Tensor Cores | TBD | 304 |
| RT Cores | TBD | 76 |
| Clock Speed(Base/Boost) | TBD | 2205/2505 MHz |
| VRAM Size and Type | 16 GB/24 GB GDDR7 | 16 GB GDDR6X |
| Memory Speed | 32 Gbps | 22.4 Gbps |
| Memory Bus Width | 256-bit | 256-bit |
| Memory Bandwidth | 1024 GB/s | 716.8 GB/s |
| TBP | 400W | 320W |
| Power Connector | 1x 12V-2×6 | 1x 12VHPWR |
The Performance Uplift We Can Expect From RTX 5080 Graphics Card
NVIDIA will launch the RTX 5080 Graphics Card as the successor to the RTX 4080, with rumors pointing that we can expect around a 10% improvement over the RTX 4090, especially in ray tracing. Although regarding the detail specifications, the RTX 4090 is reported to be slightly superior to the RTX 5080 in general performance if we only oversimplify the comparison, but the raw figures seem to suggest so, let alone the actual figures. For comparison, the performance between the RTX 4080 and the RTX 5080, it is believed the latter will be up to 2x as fast or offer 60+ fps at 4K in most gaming titles.
The RTX 4080 provided the 20-25 percent increase in the frame rates that it offered in comparison to the RTX 3090, and on the same note, higher frame rate gains are expected from the RTX 5080. The RTX 4080 and 4080 SUPER have already demonstrated acceptable performance in today’s games in terms of rasterization as well as ray tracing, and the RTX 5080 should maintain similar performance, especially in higher resolutions such as 4K, due to higher memory throughput.
Expected Pricing
As of now, there is no data on the potential prices of the RTX 5080 or any other GPUs in the RTX 50 series for now, including the RTX 5090 and RTX 5070. In some cases, the prices are established before the GPUs’ market launch and may change up to, and even after, their release, so one has to be very careful with predictions.
To put this into perspective, the RTX 4080 SUPER was released at $1,000 while the initial RTX 4080 was priced at $1,199. So this price slashing forces NVIDIA to consider the same approach with the RTX 5080, which is expected to cost between $1000. Nevertheless, if NVIDIA will be able to sell it for less money, then this oddly-named decision will make the RTX 5080 even more irresistible. As long as the RTX 5080 delivers a performance close to the current RTX 4090 ($1999), the $1000 price point would be perfect for a near-flagship card at half the price. Of course, the final pricing will vary depending on how competitive the market is going to be, but to meet the needs of enthusiasts, NVIDIA will set rather affordable prices for RTX 5080.
Availability
Nvidia finalized the design of the RTX 5080, RTX 5090, and RTX 5070 as of September 2024 but it is rumored that the RTX 5080 will not be released this year. However, NVIDIA is only expected to reveal the entirety of the 5000 series cards, which include the RTX 5080, RTX 5090 and RTX 5070, in the first quarter of 2025.
NVIDIA is likely to reveal RTX 5080 at CES 2025 in January, their usual stomping grounds for the unveiling of a new graphics technology. It is expected that this upcoming generation will show a leap in both ray tracing performance, alongside its integration with AI capabilities based on prior success of the RTX 4000 series. Anticipation is high for the increase in the performance and power efficiency, and with reports on the increase in memory bandwidth and addition of next gen architecture improvements.
Info Source: TECHPOWERUP
See More: China Has Initiated An Antitrust Investigation Into NVIDIA
