NVIDIA has just confirmed that its GeForce RTX 5080 GPU will utilize the latest GDDR7 memory modules from multiple partners.
NVIDIA GeForce RTX 50 GPUs, Including The RTX 5080, Make Use of GDDR7 Memory From Multiple Partners
We have just confirmed with NVIDIA that its GeForce RTX 5080 GPU will make use of GDDR7 VRAM dies from multiple vendors. This is important information, since yesterday, multiple photos of the RTX 5090's PCB were shown off using the Samsung GDDR7 memory.
GeForce RTX 50 series features high speed GDDR7 memory manufactured by multiple partners, starting with Samsung
Jensen Huang - NVIDIA CEO
The NVIDIA GeForce RTX 5080 is a unique example in the RTX 50 "Blackwell" lineup. It is the only card to feature memory specifications that are faster than the rest of the lineup. While the lineup uses 28 Gbps GDDR7 memory, the RTX 5080 is equipped with 30 Gbps GDDR7 memory. To fulfill the supply requirement, NVIDIA is sourcing its 30 Gbps GDDR7 memory dies from multiple DRAM partners, including Samsung and Micron.
This multi-sourcing helps a lot since GDDR7 is still in its infancy stage and each memory vendor is going to take time to get their yields up in producing higher-spec modules. A few weeks ago, we reported that the NVIDIA GeForce RTX 5080 would end up with 32 Gbps memory modules & while this was the initial plan, it looks like yields prompted NVIDIA to switch to slightly lower spec.
There is a likelihood that we will see these faster speeds in refreshes, and we are also expecting to see higher VRAM offerings in the future with 3 GB modules (24 GB RTX 5080 anyone?).
GDDR Graphics Memory Evolution:
| GRAPHICS MEMORY | GDDR7 | GDDR6X | GDDR6 | GDDR5X |
|---|---|---|---|---|
| Workload | Gaming / AI | Gaming / AI | Gaming / AI | Gaming |
| Platform (Example) | 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 | 28-48 | 19-24 | 14-16 | 11.4 |
| GB/s/placement | 112-192 | 76-96 | 56-64 | 45 |
| GB/s/system | 1536-2304 | 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) |
With that said, NVIDIA is the first graphics card and GPU manufacturer to adopt the new GDDR7 memory standard. AMD and Intel have gone with the traditional and older GDDR6 standard for cost-effective purposes, but NVIDIA's mainstream lineups are also expected to leverage GDDR7, so that's going to give them an edge in terms of raw bandwidth.
The updated memory interface allows NVIDIA's GeForce RTX 50 GPUs to push bandwidth figures to new heights, with the flagship RTX 5090 boosted up to 1.8 TB/s versus the 1.0 TB/s of bandwidth for the RTX 4090 while the RTX 5080 jumps to almost 1 TB/s bandwidth (960.0 GB/s), a 30% uplift over the RTX 4080 SUPER.
We still don't know the efficiency of the new GDDR7 memory interface, how much power it pulls versus GDDR6/X dies, and what their overclocking potential is. The last-gen GDDR6 memory was killer in terms of overclocking, so whether that's the case or not with GDDR7, we will know for sure soon enough.
NVIDIA GeForce RTX 50 GPU Specs (Official):
| Graphics Card Name | NVIDIA GeForce RTX 5090 | NVIDIA GeForce RTX 5080 | NVIDIA GeForce RTX 5070 Ti | NVIDIA GeForce RTX 5070 | NVIDIA GeForce RTX 5060 Ti | NVIDIA GeForce RTX 5060 | NVIDIA GeForce RTX 5050 |
|---|---|---|---|---|---|---|---|
| GPU Name | Blackwell GB202-300 | Blackwell GB203-400 | Blackwell GB203-300 | Blackwell GB205-300 | Blackwell GB206-300 | Blackwell GB206-250 | Blackwell GB207-300 |
| GPU SMs | 170 | 84 | 70 | 48 | 36 | 30 | 20 |
| GPU Cores | 21760 | 10752 | 8960 | 6144 | 4608 | 3840 | 2560 |
| Clock Speeds | 2017 MHz (Base) 2407 MHz (Boost) | 2295 MHz (Base) 2617 MHz (Boost) | 2295 MHz (Base) 2452 MHz (Boost) | 2325 MHz (Base) 2512 MHz (Boost) | 2407 MHz (Base) 2572 MHz (Boost) | 2280 MHz (Base) 2497 MHz (Boost) | 2317 MHz (Base) 2572 MHz (Boost) |
| Memory Capacity | 32 GB GDDR7 | 16 GB GDDR7 | 16 GB GDDR7 | 12 GB GDDR7 | 8-16 GB GDDR7 | 8 GB GDDR7 | 8 GB GDDR6 |
| Memory Bus | 512-bit | 256-bit | 256-bit | 192-bit | 128-bit | 128-bit | 128-bit |
| Memory Speed | 28 Gbps | 30 Gbps | 28 Gbps | 28 Gbps | 28 Gbps | 28 Gbps | 20 Gbps |
| Bandwidth | 1792 GB/s | 960 GB/s | 896 GB/s | 672 GB/s | 448 GB/s | 448 GB/s | 320 GB/s |
| Power Interface | 1 12V-2x6 (16-Pin) | 1 12V-2x6 (16-Pin) | 1 12V-2x6 (16-Pin) | 1 12V-2x6 (16-Pin) | 16-Pin or 8-Pin | 8-Pin | 8-Pin |
| Launch | 30th Jan, 2025 | 30th Jan, 2025 | 20th Feb, 2025 | 4th March, 2025 | 16th April, 2025 | 19th April, 2025 | 1st July, 2025 |
| TBP | 575W | 360W | 300W | 250W | 180W | 145W | 130W |
| Price (MSRP) | $1999 | $999 | $749 | $549 | $379-$429 | $299 | $249 |
About the author: A Software Engineer by training and a PC enthusiast by passion, Hassan Mujtaba serves as Wccftech's Senior Editor for hardware section. With years of experience in the industry, he specializes in deep-dive technical analysis of next-generation CPU and GPU architectures, motherboards, and cooling solutions. His work involves not only breaking news on upcoming technologies but also extensive hands-on reviews and benchmarking.
Follow Wccftech on Google to get more of our news coverage in your feeds.
