Samsung and Micron are working together on the hyper memory cube (HMC) and they seem to have finally decided on a time frame for when the first specification appear.
The first hyper memory cube (HMC) products will be available in 2013,
according to the most recent update on Samsung's part.
This will be possible because the specification, or the standardization of multi-layer hyper memory cube DRAM, will be ready by late 2012.
Already demonstrated by Intel at IDF 2011, HMC, originally promoted by Intel and Micron, uses a stacked memory chip configuration.
The cube is efficient, about seven times more efficient in fact, compared to existing DDR3 memory.
Samsung and Micron
founded the hybrid memory cube consortium (HMCC) and want to use HMC in their quest to break the performance bottleneck, that “memory wall,” which refuses to allow multi-core, multi-threaded CPUs work at their fullest.
Given that, even with minimum energy per bit utilization, a “cube” can have data rates of 1 Tb/s (one trillion bits per second), this may very well happen.
This problem didn't exist prior to the appearance of this and last year's AMD and Intel CPUs, as well as advanced networking equipment which drive bandwidth requirements over the upper limit of memory architectures.
“We anticipate that the group will complete an industry-wide spec for the 1st phase of HMC by the end of 2012. Within a year or two after that, HMC products should begin to reach the market as an alternative to DRAM in HPC and network applications,” said Jim Elliott, vice president of memory marketing at Americas.
There is, of course, the possibility of research and development delays, or other factors, it may take a bit longer than 2012-2013 for palpable results to emerge.
Nonetheless, even if things drag on to 2014, HMC still has every assurance that it will be welcomed with open arms.