Linked by Thom Holwerda on Wed 31st Oct 2012 00:46 UTC
Permalink for comment 540805
To read all comments associated with this story, please click here.
To read all comments associated with this story, please click here.
Features
Linked by Thom Holwerda on 05/18/13 21:33 UTC
Linked by David Adams on 05/16/13 4:23 UTC
Linked by Thom Holwerda on 05/11/13 21:41 UTC
Linked by Thom Holwerda on 05/08/13 14:22 UTC
Linked by Thom Holwerda on 05/02/13 15:28 UTC
Linked by Thom Holwerda on 04/29/13 21:06 UTC
Linked by Thom Holwerda on 04/24/13 22:24 UTC
Linked by Thom Holwerda on 04/18/13 11:21 UTC
Linked by Thom Holwerda on 04/16/13 9:29 UTC
Linked by Thom Holwerda on 04/15/13 22:44 UTC
More Features »
Sponsored Links



Member since:
2011-01-28
bassbeast,
"The problem with ARM is it just doesn't scale well and it has lousy IPC. Companies like Nvidia have spent crazy money on ARM R&D and still can't overcome the weak IPC and the scalability problem, which is why Tegra is up to 5 cores now."
I don't know, I think most engineers would claim that being able to support many cores IS an example of good scalability. There's been a long debate over having very deep complex logic gates required to run serial logic in parallel pipelines versus having more cores on the same die space. Also, ARM has some advantages on the opcode level which can significantly reduce branching for free.
Not to dis intel though, I'm impressed with how they can essentially recompiling code on the fly to get more parallelism out of sequential instructions, it's truly an incredible feat. But this complexity has a cost, it requires more gates and power. This may be somewhat mitigated by the fact that intel has die fabrication advantages. But these pipeline gates could have gone to parallel cores instead, which theoretically should perform even better with software that can take advantage of it.
Serial programming is not scalable and deep pipelines cannot increase single thread performance any longer. The more prevalent concurrent programming becomes, the better ARM processors will perform per gate compared to x86 counterparts. More so if they use the same manufacturing technology (since ARM lags behind).
Here are some benchmarks comparing various mobile x86 and ARM processors, unfortunately they haven't included any x86 server processors for reference. Never the less, even with a faster clock, the atom does not win on all performance tests, and it consumes much more power!
http://www.brightsideofnews.com/news/2011/5/19/the-coming-war-arm-v...
http://www.brightsideofnews.com/news/2011/5/19/the-coming-war-arm-v...
Edited 2012-11-01 14:47 UTC