SK hynix confirms its plans to bring 24 Gb GDDR7 memory for higher VRAM capacities on GPUs, while also prepping up HBM4.
SK Hynix 24 Gb GDDR7 & HBM4 Memory Solutions Are Now Actively Being Worked On, Higher VRAM Capacities Coming To GPUs Later This Year
In its recent (2Q25) earnings call, SK Hynix revealed its current progress and upcoming solutions within the DRAM/Memory segment. The company highlights its expansion of sales for its 12-Hi HBM3e DRAM, while shipments for both DRAM and NAND flash were higher than expected, leading to the best quarterly results in a while.
The major announcement, besides the earnings, was that the company is now working on its 24 Gb GDDR7 memory modules which will be used to power next-generation graphics cards, and offer AI customers with even more room to expand VRAM capacities of their products. 24Gb (3 GB) VRAM is going to offer 50% more capacity than existing 16Gb (2 GB) dies.
GDDR7 is also expected to see uplifts within the speed department with 30+ Gbps becoming more mainstream as we move forward. It is still going to be a while before we see 40+ Gbps solutions enter the market, but higher VRAM capacities are definitely going to be great. GPU manufacturers such as NVIDIA have the flexibility to use GDDR7 memory from various partners, including SK Hynix, Samsung & Micron.
Samsung had already commenced the production of such memory modules, which were also spotted on sale in online listings. NVIDIA's upcoming "RTX 50 SUPER" series graphics cards are expected to leverage these higher-capacity memory modules, but they are expected to launch either later this year or early next year.
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) |
The company expects the solid performance of its products and mass-production capabilities to help double HBM, compared with a year earlier, to generate stable earnings. It will also ensure timely provision of HBM4 in accordance with customers’ requests to remain competitive.
SK hynix will start provision of an LPDDR-based module for servers within this year, and prepare for GDDR7 products for AI GPUs with an expanded capacity of 24Gb from 16Gb in a bid to enhance its leadership in the AI memory market with product diversification.
via SK hynix
Besides GDDR7, SK hynix is also working to ensure timely provision of HBM4 memory. The next-gen HBM4 standard is going to shape the HPC/AI segment. NVIDIA and AMD are expected to use these memory solutions on their next-generation products such as Rubin and MI400. Initial supply, which is going to be used for evaluation purposes, has reportedly commenced since last month. A recent HBM roadmap shows how the next-gen memory standards will stack up versus existing solutions, and what sort of architectures they will adopt.
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.
