Wednesday, March 20th 2019

Samsung Electronics Introduces New Flashbolt HBM2E High Bandwidth Memory

Samsung Electronics Co., Ltd., the world leader in advanced semiconductor technology, today announced its new High Bandwidth Memory (HBM2E) product at NVIDIA's GPU Technology Conference (GTC) to deliver the highest DRAM performance levels for use in next-generation supercomputers, graphics systems, and artificial intelligence (AI).

The new solution, Flashbolt , is the industry's first HBM2E to deliver a 3.2 gigabits-per-second (Gbps) data transfer speed per pin, which is 33 percent faster than the previous-generation HBM2. Flashbolt has a density of 16Gb per die, double the capacity of the previous generation. With these improvements, a single Samsung HBM2E package will offer a 410 gigabytes-per-second (GBps) data bandwidth and 16 GB of memory.
"Flashbolt's industry-leading performance will enable enhanced solutions for next-generation data centers, artificial intelligence, machine learning, and graphics applications," said Jinman Han, senior vice president of Memory Product Planning and Application Engineering Team at Samsung Electronics. "We will continue to expand our premium DRAM offering, and improve our 'high-performance, high capacity, and low power' memory segment to meet market demand."
Add your own comment

6 Comments on Samsung Electronics Introduces New Flashbolt HBM2E High Bandwidth Memory

#1
AmioriK
1.6TB/s on a 4096 bit bus with 64GB capacity damn. Even just a single stack could feed a mid/high tier GPU !
Posted on Reply
#3
Zubasa
AmioriK said:
1.6TB/s on a 4096 bit bus with 64GB capacity damn. Even just a single stack could feed a mid/high tier GPU !
Hopefully that would mean less ram stacks required per GPU and maybe make it economical for consumer hardware.
Posted on Reply
#4
delshay
Lost for words on this one, other than we now need a GPU that can take advantage of 4 stacks.
Posted on Reply
#5
Xajel
Zubasa said:
Hopefully that would mean less ram stacks required per GPU and maybe make it economical for consumer hardware.
Still, it's very dense, denser than regular HBM2's, meaning it's a must to have a silicon interposer, which is the main reason HBM is expensive for the mainstream.

Samsung announced a low cost option for HBM few months back (maybe over a year) which uses 512bit rather than 1024bit bus at a higher clock, while providing less bandwidth than regular HBM. The narrower bus makes it possible to use much less expensive organic interposer, something like current MCM processors. But it seems Samsung is still working on it or they couldn't find much interest on it.
Posted on Reply
#6
Apocalypsee
Xajel said:

Samsung announced a low cost option for HBM few months back (maybe over a year) which uses 512bit rather than 1024bit bus at a higher clock, while providing less bandwidth than regular HBM. The narrower bus makes it possible to use much less expensive organic interposer, something like current MCM processors. But it seems Samsung is still working on it or they couldn't find much interest on it.
AMD should take that low cost HBM and slap them on an APU. Even a single stack 512-bit 2Gbps gives 128GB/s bandwidth, A LOT higher than mere 51.2GB/s DDR4-3200MHz offers. It would made a GREAT gaming notebook, and a decent NUC. But it will cannibalize lower end GPU market for desktop market.
Posted on Reply