Micron officially reveals its GDDR6X memory and their future plans
Micron officially reveals its GDDR6X memory and their future plans
With GDDR6X, Micron is the first provider of GDDR memory to create a product that’s capable of over 1TB/s of system bandwidth, thanks to the 21Gbps performance of its fastest GDDR6X chips.Â
With GDDR6X, Micron has achieved a significant bandwidth boost over GDDR6 by utilising PAM4 multilevel signalling to move two bits at a time to double GDDR6X’s I/O data rate. This shift also allows GDDR6X to deliver its performance with lower power requirements per memory transaction, making GDDR6X memory more efficient than standard GDDR6 memory.Â
Modern graphics cards are useless without fast memory to keep their processors fed. With every GPU generation, manufacturers like AMD and Nvidia need to do what they can to increase the memory bandwidth that’s available to their systems. This bandwidth boost can be achieved through increases to effective bandwidth with architectural changes or increases to raw bandwidth through the use of faster memory technologies or a wider memory bus. With Ampere, Nvidia’s using faster memory modules to reach its bandwidth targets, thanks to Micron’s GDDR6X technology.Â
 Â
Micron’s GDDR6X will first be released to the public on September 17th with Nvidia’s RTX 3080 graphics card, which utilises 10GB of GDDR6X memory to exceed the memory bandwidth of all of Nvidia’s prior Geforce offerings.
With maximum speeds of 21Gbps, GDDR6X memory delivers a bandwidth increase of over 30% when compared to 16Gbps GDDR6 memory modules. This lead extends to 50% after you consider that most GDDR6-based consumer GPUs only use 14 Gbps GDDR6 memory modules. Â
Right now, Micron’s GDDR6X memory chips are only available in 8Gb densities, translating to 1GB of capacity. In 2021, Micron plans to offer 16Gb (2GB) GDDR6X memory chips, which will allow for the creation of Ampere graphics cards with larger memory densities.Â
This means that Nvidia’s RTX 3090 graphics card will use 24 8Gb GDDR6X memory chips at launch, a factor which explains why the graphics cards costs as much as it does. Nvidia’s RTX 3090 uses 140% more GDDR6X chips than the RTX 3080, adding significantly to the graphics card’s material cost and manufacturing complexity.Â
In the future, it is possible that Nvidia will release refreshed Ampere products with larger memory densities or faster GDDR6X memory chips. At this time, it looks like none of Nvidia’s RTX 30 series products utilises Micron’s 21Gbps GDDR6X memory modules. This could change if Nvidia eventually releases RTX Super series Ampere products.Â
The video below should give you an additional primer on Micron’s GDDR6X memory technology.Â
You can join the discussion on Micron’s GDDR6X memory technology on the OC3D Forums.Â