The High Bandwidth Memory (HBM) processors have been on the drawing board for more than seven years, and it has now outlined the benefits they will offer over conventional GDDR5 RAM.
AMD partnered up with a number of companies to help define a new specification and design new type of memory chip with low power consumption and an ultra-wide bus width. Hynix developed the HBM standard and memory, which was eventually adopted by JEDEC in October 2013. AMD also worked to develop a key component of the technology, called the interposer, along with ASE, Amkor, and UMC.
The interposer allows DRAM to be brought into close proximity with the GPU and simplifies communication and clocking, even at very wide bus widths.
AMD claims that HBM chips offer a significant performance boost with a fraction of the power consumption, and are able to support 1024-bit architecture and transfer speeds of 100GB per second.
The design allows memory cells to be stacked vertically, rather than crammed next to each other like in GDDR5 RAM.
The HBM components have a lower clock speed of around 500Mhz, but AMD chief technical officer Joe Macri said that the processors are easier to overclock.
He said that with HBM you don’t have to turn the voltage knob as much, there were huge andwidth increases so I think it’s going to be very overclockable, maybe even easier than G5.
AMD is yet to reveal a firm release date for its HBM processors, but the first iterations are expected to appear in the coming months.