Real World Technologies - Forums - Thread: New article - The Common System Interface: Intel's Future Interconnect

Article: The Common System Interface: Intel's Future Interconnect

By: David Kanter (dkanter.delete@this.realworldtech.com), September 15, 2007 7:50 am

unknown (un@known.net) on 9/15/07 wrote:
---------------------------
>David Kanter (dkanter@realworldtech.com) on 9/14/07 wrote:
>>It's very straight forward - if different regions of memory have different latencies
>>then the system is NUMA. If all memory has the same latency, then the system is UMA. It cannot be both.
>
>By that logic, turning on Node Interleave on the Opteron >would make it UMA (?)

Yes, that's right. I believe the idea behind node interleave is to have a mode that works effectively for operating systems with no, or poor NUMA support.

DK

< Previous Post in Thread

Next Post in Thread >

Topic	Posted By	Date
New article - The Common System Interface: Intel's Future Interconnect	David Kanter	2007/08/28 01:16 AM
New article - The Common System Interface: Intel's Future Interconnect	Vincent Diepeveen	2007/08/28 05:02 AM
New article - The Common System Interface: Intel's Future Interconnect	Richard Cownie	2007/08/28 10:28 AM
New article - The Common System Interface: Intel's Future Interconnect	Vincent Diepeveen	2007/08/31 11:44 AM
New article - The Common System Interface: Intel's Future Interconnect	Richard Cownie	2007/08/31 08:53 PM
New article - The Common System Interface: Intel's Future Interconnect	Vincent Diepeveen	2007/09/01 02:21 AM
Adding layers can simplify design	Paul A. Clayton	2007/09/01 07:39 AM
New article - The Common System Interface: Intel's Future Interconnect	Michael S	2007/09/02 02:25 AM
New article - The Common System Interface: Intel's Future Interconnect	Jonathan Kang	2007/09/14 12:47 PM
New article - The Common System Interface: Intel's Future Interconnect	David Kanter	2007/09/14 08:47 PM
New article - The Common System Interface: Intel's Future Interconnect	Paul	2007/08/28 11:04 AM
New article - The Common System Interface: Intel's Future Interconnect	David Kanter	2007/08/28 12:43 PM
New article - The Common System Interface: Intel's Future Interconnect	Joe Chang	2007/08/28 06:17 PM
New article - The Common System Interface: Intel's Future Interconnect	Joe Chang	2007/08/29 04:27 PM
Thanks for the work	Wouter Tinus	2007/08/28 12:33 PM
Thanks for the work	mac	2007/08/29 12:44 PM
New article - The Common System Interface: Intel's Future Interconnect	Herbert Hum	2007/08/28 01:22 PM
Thanks	David Kanter	2007/08/28 04:13 PM
Many thanks, very, very interesting! (NT)	Cameron Jack	2007/08/29 01:51 AM
very nice article + memory ctrl integration	Marcin Niewiadomski	2007/08/29 11:46 AM
very nice article + memory ctrl integration	David Kanter	2007/09/14 08:50 PM
very nice article + memory ctrl integration	Marcin Niewiadomski	2007/09/16 08:48 PM
Coherency: Forwarding and Owned	Peter Gerdes	2007/08/29 02:11 PM
Coherency: Forwarding and Owned	David Kanter	2007/08/29 06:29 PM
Coherency: Forwarding and Owned	nick	2007/08/29 07:03 PM
Coherency: Forwarding and Owned	David Kanter	2007/08/29 11:08 PM
Coherency: Forwarding and Owned	Michael S	2007/08/30 01:17 AM
Coherency: Forwarding and Owned	David Kanter	2007/08/30 07:31 AM
Coherency: Forwarding and Owned	Peter Gerdes	2007/08/30 11:46 AM
Coherency: Forwarding and Owned	David Kanter	2007/08/30 01:46 PM
Coherency: Forwarding and Owned	Peter Gerdes	2007/08/30 07:03 PM
Coherency: Forwarding and Owned	David Kanter	2007/09/14 08:44 PM
Node Interleave	unknown	2007/09/15 03:14 AM
Node Interleave	David Kanter	2007/09/15 07:50 AM
Node Interleave	Howard Chu	2007/09/16 12:14 PM
Coherency: Forwarding and Owned	Peter Gerdes	2007/09/16 12:50 PM
Coherency: Forwarding and Owned	David Kanter	2007/09/16 04:34 PM
Coherency: Forwarding and Owned	EduardoS	2007/09/16 04:52 PM
Coherency: Forwarding and Owned	Jonathan Kang	2007/09/17 05:16 AM
Coherency: Forwarding and Owned	Matthias	2007/09/17 06:59 AM
Coherency: Forwarding and Owned - addition	Matthias	2007/09/17 07:01 AM
Coherency: Forwarding and Owned	anonymous	2007/09/17 09:15 AM
Coherency: Forwarding and Owned	Peter Gerdes	2007/09/17 12:44 PM
New article - The Common System Interface: Intel's Future Interconnect	Mr. Camel	2007/08/30 03:16 PM
New article - The Common System Interface: Intel's Future Interconnect	Michael S	2007/08/31 01:11 AM
New article - The Common System Interface: Intel's Future Interconnect	Mr. Camel	2007/08/31 03:13 AM
New article - The Common System Interface: Intel's Future Interconnect	Michael S	2007/08/31 03:24 AM
New article - The Common System Interface: Intel's Future Interconnect	David Kanter	2007/08/31 05:39 AM
New article - The Common System Interface: Intel's Future Interconnect	Michael S	2007/08/31 06:53 AM
New article - The Common System Interface: Intel's Future Interconnect	David Kanter	2007/08/31 05:41 AM
New article - The Common System Interface: Intel's Future Interconnect	Michael S	2007/08/31 06:36 AM
New article - The Common System Interface: Intel's Future Interconnect	Mr. Camel	2007/08/31 08:36 AM
Thanks and excellent work!	Jack A.	2007/08/30 07:41 PM
Lamport's TLA	Konrad Schwarz	2007/09/02 01:57 AM
Lamport's TLA	David Kanter	2007/09/02 07:55 PM
Lamport's TLA	Brannon	2007/09/03 07:12 AM
Lamport's TLA	Konrad Schwarz	2007/09/18 10:21 AM
Lamport's TLA	Brannon	2007/09/18 01:58 PM
New article - The Common System Interface: Intel's Future Interconnect	José Javier Zarate	2007/09/09 04:01 PM
New article - The Common System Interface: Intel's Future Interconnect	Jonathan Kang	2007/09/16 05:42 PM
Remote prefetch	David Kanter	2007/09/17 07:51 AM
New article - The Common System Interface: Intel's Future Interconnect	Jigal	2007/09/22 02:39 PM
New article - The Common System Interface: Intel's Future Interconnect	David Kanter	2007/09/22 04:35 PM
New article - The Common System Interface: Intel's Future Interconnect	8B/10B Latency	2007/09/22 06:16 PM
New article - The Common System Interface: Intel's Future Interconnect	anon	2007/09/22 08:05 PM
New article - The Common System Interface: Intel's Future Interconnect	David W. Hess	2007/09/22 08:50 PM
Clocking in CSI	David Kanter	2007/09/23 08:46 AM
Hypertransport 3 AC Coupling	David W. Hess	2007/09/23 09:32 AM
Hypertransport 3 AC Coupling	anon	2007/09/23 09:53 AM
Clocking lanes	David Kanter	2007/09/23 10:51 AM
New article - The Common System Interface: Intel's Future Interconnect	David Kanter	2007/09/22 08:34 PM
New article - The Common System Interface: Intel's Future Interconnect	David W. Hess	2007/09/22 09:10 PM
New article - The Common System Interface: Intel's Future Interconnect	Jonathan Kang	2007/09/25 07:15 AM
New article - The Common System Interface: Intel's Future Interconnect	Michael S	2007/09/23 01:06 AM
New article - The Common System Interface: Intel's Future Interconnect	David W. Hess	2007/09/23 03:41 AM
Serialization delay	David Kanter	2007/09/23 08:57 AM
Serialization delay	Michael S	2007/09/23 11:20 AM
Serialization delay	David Kanter	2007/09/23 12:43 PM
Serialization delay	Michael S	2007/09/24 12:40 AM
Serialization delay	Michael S	2007/09/24 04:28 AM
Serialization delay	Aaron Spink	2007/09/24 12:19 PM
Serialization delay	Michael S	2007/09/25 03:38 AM
Serialization delay	Jonathan Kang	2007/09/25 08:10 AM
Serialization delay	David W. Hess	2007/09/26 12:22 AM
Serialization delay	Aaron Spink	2007/09/25 12:13 PM
Thank you (NT)	Michael S	2007/09/25 12:53 PM
Serialization delay	Jonathan Kang	2007/09/25 07:26 AM
Serialization delay	Michael S	2007/09/25 01:57 PM
Serialization delay	Jonathan Kang	2007/09/26 05:24 AM
Serialization delay	David W. Hess	2007/09/26 06:39 AM
Serialization delay	Jonathan Kang	2007/09/26 09:56 AM
Serialization delay	David W. Hess	2007/09/27 02:21 AM
Serialization delay	Jonathan Kang	2007/09/27 04:36 AM
Serialization delay	David W. Hess	2007/09/27 05:31 PM
Serialization delay	rwessel	2007/09/26 01:26 PM
Serialization delay	Jonathan Kang	2007/09/27 07:16 AM
Serialization delay	rwessel	2007/09/27 12:20 PM
Serialization delay	Jonathan Kang	2007/09/28 04:38 AM
Serialization delay	rwessel	2007/09/28 01:00 PM
Serialization delay	Jonathan Kang	2007/10/01 07:07 AM
Cache coherent latency	David Kanter	2007/10/01 07:20 AM
Cache coherent latency	blaine	2007/10/01 10:36 AM
Critical word first on coherent interconnects	David Kanter	2007/10/01 11:10 AM
Does ccHT do critical word first?	blaine	2007/10/02 07:10 AM
Cache coherent latency	Jonathan Kang	2007/10/01 12:34 PM
Cache coherent latency	David Kanter	2007/10/01 01:13 PM
Serialization delay	Michael S	2007/09/28 04:32 AM
Serialization delay	anonymous	2007/09/28 10:25 AM
Serialization delay	Michael S	2007/09/29 09:06 AM
New article - The Common System Interface: Intel's Future Interconnect	Jonathan Kang	2007/09/25 07:05 AM
New article - The Common System Interface: Intel's Future Interconnect	jigal	2007/09/23 12:37 PM
CSI, PCI and HT	David Kanter	2007/09/23 12:46 PM
New article - The Common System Interface: Intel's Future Interconnect	Jonathan Kang	2007/09/25 07:39 AM
New article - The Common System Interface: Intel's Future Interconnect	jigal	2007/09/25 02:16 PM
New article - The Common System Interface: Intel's Future Interconnect	Michael S	2007/09/26 03:14 AM
New article - The Common System Interface: Intel's Future Interconnect	Anonymous	2007/09/26 09:41 AM
New article - The Common System Interface: Intel's Future Interconnect	Jonathan Kang	2007/09/26 09:59 AM
New article - The Common System Interface: Intel's Future Interconnect	Jonathan Kang	2007/09/26 05:48 AM

Reply to this Topic
Name:
Email:
Topic:
Body:	No Text David Kanter (dkanter.delete@this.realworldtech.com) on September 15, 2007 7:50 am wrote: > unknown (un@known.net) on 9/15/07 wrote: > --------------------------- > >David Kanter (dkanter@realworldtech.com) on 9/14/07 wrote: > >>It's very straight forward - if different regions of memory have different latencies > >>then the system is NUMA. If all memory has the same latency, then the system is UMA. It cannot be both. > > > >By that logic, turning on Node Interleave on the Opteron >would make it UMA (?) > > Yes, that's right. I believe the idea behind node interleave is to have a mode > that works effectively for operating systems with no, or poor NUMA support. > > DK
How do you spell avocado?

Node Interleave

Editor’s Picks

Intel’s Sandy Bridge Microarchitecture

A Historical Look at the VAX: Microprocessor Economics

Silvermont, Intel’s Low Power Architecture

RWT on Twitter