Well, it seems that Micron has officially started mass production of its next-generation GDDR6X memory chips that offer speeds of up to 24 Gbps.
Micron’s 24Gbps GDDR6X Memory Enters Mass Production and Will Come to NVIDIA’s Next-Generation GPUs Later This Year
Micron Announced Earlier this year they were preparing their new 24Gbps GDDR6X memory modules for future graphics cards. The announcement came at a time when NVIDIA’s GeForce RTX 3090 Ti had just entered the market for which Micron had produced 16Gb DRAM modules with speeds of up to 21Gbps.
Micron will now offer even faster pin speeds of 24Gbps at 16Gb density, which means we’ll once again be looking at 2GB VRAM capacities and capacities up to 24GB on a 384-bit bus interface. . Like the GeForce RTX 3090 Ti, which featured all of its memory modules on the front of the PCB, the Next Generation GeForce RTX 4090 would feature its 12 GDDR6X modules on the front, allowing for more cooling compared to solutions that feature memory modules on the back of the PCB, like the RTX 3090 (not Ti).

Although the NVIDIA GeForce RTX 4090 is not expected to use the full 24Gbps speeds (it is rumored to offer matrix speeds of up to 21Gbps), the 24Gbps GDDR6X memory modules can offer up to 1,152TB/s of bandwidth, which is a 14% increase over the 1,008TB/s offered by current 21Gbps GDDR6 memory modules.
The following are the bandwidth numbers you can expect with a 24 Gbps DRAM solution:
- 512 bit solution- 1.5TB/s
- 384 bit solution- 1.1TB/s
- 320 bit solution- 960GB/s
- 256 bit solution- 768GB/s
- 192 bit solution- 576GB/s
- 128 bit solution- 384GB/s
- 92 bit solution- 276GB/s
- 64 bit solution- 192GB/s
The NVIDIA GeForce RTX 4090 will be among the first next-generation graphics cards to use the latest GDDR6X memory modules when it launches later this year. Since these chips are likely to be scaled down to meet power targets, there will be plenty of overclocking potential for enthusiasts to take advantage of.
On the other hand, AMD is likely to rely on its partner, Samsung, for its next-gen RDNA 3 lineup. Samsung is also working on a 24 Gbps GDDR6 memory module which is expected to reach mass production soon.
GRAPHICS MEMORY | GDDR5X | GDDR6 | GDDR6X |
---|---|---|---|
Workload | Graphics | GraphicsAI Inference Accelerator | GraphicsAI Inference Accelerator |
Platform (Example) | TitanX | Titan RTX RX5700XT |
GeForce® RTX™ 3090Ti GeForce® RTX™ 3080Ti |
number of locations | 12 | 12 | 12 |
GB/s/pin | 11.4 | 14-16 | 19-24 |
GB/s/location | Four. Five | 56-64 | 76-96 |
GB/s/system | 547 | 672-768 | 912-1152 |
Configuration (Example) | 384IO (package of 12 pieces x 32IO) |
384IO (package of 12 pieces x 32IO) |
384IO (package of 12 pieces x 32IO) |
Typical System Frame Buffer | 12GB | 12GB | 24GB |
Device average power (pJ/bit) | 8.0 | 7.5 | 7.25 |
Typical I/O Channel | printed circuit board (P2PSM) |
printed circuit board (P2PSM) |
printed circuit board (P2PSM) |
News source: Harukaze5719
Products mentioned in this publication.