Revealed Samsung unveiled a new RAM module that shows the potential of DDR5 memory in terms of speed and capacity.
DDR5, the new standard in DRAM, aims to meet the hungry demands of supercomputing, artificial intelligence and machine learning, as well as data analysis applications.
Samsung said the 512GB DDR5 unit is the first to use HKMG processing technology, providing speeds up to 7200 Mbps – more than twice the speed of DDR4.
Currently this unit targets data-hungry supercomputing, artificial intelligence and machine learning functions, but DDR5 will find its way onto PCs, boosting games and other applications.
Samsung first used HKMG in 2018 with the GDDR6 chips used in GPUs.
By expanding its use of DDR5, Samsung is establishing its leadership in the next generation of DRAM technology.
The HKMG process technology was developed by Intel, and it uses hafnium instead of silicon, with the metal replacing the regular polysilicon gate electrodes, allowing With a higher chip density, while reducing leakage current.
Each chip uses eight layers of 16GB of DRAM chips to deliver the largest capacity of 512GB, and as such, Samsung needs 32 of these to create a 512GB RAM module.
In addition to higher speeds and capacity, Samsung said: The chip uses 13 percent less energy than units without HKMG technology, and is ideal for data centers.
With speeds of 7200 Mbps, the latest unit from Samsung provides transfer speeds of around 57.6 GB per second over a single channel.
In a Samsung press release, Intel indicated that the memory will be compatible with Xeon Scalable processors of the next generation Sapphire Rapids.
This architecture uses an eight-channel DDR5 memory controller, so we can see multi-terabyte memory configurations with memory transfer speeds of up to 460 GB per second.
Meanwhile, the first consumer PCs could arrive in 2022 when AMD unveils its Zen 4 platform, which is rumored to support DDR5.
Samsung’s DDR5 also features ECC technology to enhance reliability, where the hardware is expected to assist in powering the computing devices needed, among others, for medical research, financial markets, autonomous leadership and smart cities.