Micron Collaborates With Intel to Enhance Knights Landing

Published by

Click here to post a comment for Micron Collaborates With Intel to Enhance Knights Landing on our message forum
https://forums.guru3d.com/data/avatars/m/228/228458.jpg
today announced an ongoing collaboration with Intel to deliver an on-package memory solution for Intel's next-generation Xeon Phi processor
Can someone elaborate further on what is in bold?
https://forums.guru3d.com/data/avatars/m/47/47825.jpg
Sounds like ram on die with the processor to be simple about it.
data/avatar/default/avatar19.webp
2.5D or 3D stack technology most likely.
https://forums.guru3d.com/data/avatars/m/228/228458.jpg
Sounds like ram on die with the processor to be simple about it.
Hmmm. Interesting.
data/avatar/default/avatar32.webp
Industry as a whole is moving towards 3D stacked memory chips, basically same thing Nvidia doing with Pascal
data/avatar/default/avatar30.webp
This is the next big battlefield - HPC/datacenters/supercomputers/workstations And the compute market is still rapidly growing. Intel vs Nvidia - FIGHT! :banana:
https://forums.guru3d.com/data/avatars/m/229/229509.jpg
This is the next big battlefield - HPC/datacenters/supercomputers/workstations And the compute market is still rapidly growing. Intel vs Nvidia - FIGHT! :banana:
Despite the fact that AMDs cards are far better at compute that nVidias...
data/avatar/default/avatar08.webp
Despite the fact that AMDs cards are far better at compute that nVidias...
HSA is still just a 3-letter word. Don't blame me. Tell that to academia,IBM, HP, Amazon. To these guys too: http://www.green500.org/lists/green201311&green500from=1&green500to=100 highlights from the 43rd list: A total of 62 systems on the list are using accelerator/co-processor technology, up from 53 from November 2013. Forty-four of these use NVIDIA chips, two use ATI Radeon, and there are now 17 systems with Intel MIC technology (Xeon Phi). The average number of accelerator cores for these 62 systems is 78,127 cores/system. http://www.top500.org/lists/2014/06/ AMD will always have it's place, but this is mostly between Intel/NV imho.
https://forums.guru3d.com/data/avatars/m/80/80129.jpg
HSA is still just a 3-letter word. Don't blame me. Tell that to academia,IBM, HP, Amazon. To these guys too: http://www.green500.org/lists/green201311&green500from=1&green500to=100 highlights from the 43rd list: A total of 62 systems on the list are using accelerator/co-processor technology, up from 53 from November 2013. Forty-four of these use NVIDIA chips, two use ATI Radeon, and there are now 17 systems with Intel MIC technology (Xeon Phi). The average number of accelerator cores for these 62 systems is 78,127 cores/system. AMD will always have it's place, but this is mostly between Intel/NV imho.
Well to be fair it's mostly because CUDA was significantly ahead of OpenCL in terms of development libraries. That gap has closed -- also AMD has since released a bunch of FirePro cards including the new 8100 which is significantly faster than Nvidia in double precision. That being said, there was an article on Arstechnica today about how there hasn't been any new machines in the top 10 super computer list in a while. They basically said that more and more companies are focusing on smaller systems with better power efficiency. Nvidia has been learning a lot in it's attempts to take it's design and shrink it for Tegra. Kepler is obviously a huge improvement in power design and Pascal will further improve that. I think Nvidia had a bit better foresight in that regard and I think they will continue to be leader in HPC.
https://forums.guru3d.com/data/avatars/m/229/229509.jpg
Well to be fair it's mostly because CUDA was significantly ahead of OpenCL in terms of development libraries. That gap has closed -- also AMD has since released a bunch of FirePro cards including the new 8100 which is significantly faster than Nvidia in double precision.
This is what I meant. Can also game on AMD's cards much better than nVidias, multipurpose 😀
https://forums.guru3d.com/data/avatars/m/149/149159.jpg
Sounds like ram on die with the processor to be simple about it.
On-package is different than on-die. On-package means it will sit next to the die under the heatsink.
https://forums.guru3d.com/data/avatars/m/56/56686.jpg
seeing as we been hearing about dd4 for years and years now and it not due out till later this year? How long out is "this" tech.
https://forums.guru3d.com/data/avatars/m/242/242471.jpg
Yes. Except there is 0 chance we will see these in GPUs in 2014. Q1 2015 maybe, AMD seems to be pushing a lot for it (co-developer, just like they did with GDDR)
I know, Q4 2014 was the initial and the best case scenario. Q1 2015 or even Q3 2015 looks more realistic.