backtop


Print 11 comment(s) - last by retrospooty.. on Jul 24 at 10:20 AM

Intel "Nehalem" to drop into "Tylersburg" platforms

Intel has released early details of its upcoming server and workstation platforms for Nehalem processorsTylersburg. Tylersburg has three variants in the pipeline – EP for efficient performance servers, EN for entry servers and WS for workstations.

Details of Tylersburg are scarce; however, Intel plans to have different variants of a single chipset for each platform. Intel currently names the chipsets Tylersburg-36D, Tylersburg-36S and Tylersburg-24D. Tylersburg-36D is the chipset of choice for dual-socket Tylersburg EP and WS platforms. However, Tylersburg WS makes use of dual Tylersburg-36D chipsets for enhanced expansion. Intel plans to replace existing Bensley and Stoakley platforms with Tylersburg EP and WS platforms.

Tylersburg-24D will find its way into the Tylersburg EN platform. Tylersburg EN is the entry-level server platform that will join Bensley VS and Cranberry Lake platforms. Intel’s single-processor Garlow WS platform will receive a Nehalem makeover with a Tylersburg WS platform. Intel plans to equip single processor Tylersburg WS platforms with the Tylersburg-36S chipset.

Intel’s upcoming Nehalem processor will drop into Tylersburg platforms. Nehalem introduces QuickPath to Intel’s Xeon-based platforms. QuickPath is Intel’s new bus infrastructure, formerly known as common-system interface, or CSI. Intel’s Nehalem architecture features quad cores with simultaneous multi-threading technology, allowing up to sixteen threads on an eight-core platform. Nehalem remains on a 45nm hi-K silicon process, as with Penryn. However, Nehalem will have multi-level shared cache architecture.

Intel plans to debut Tylersburg platforms and Nehalem in the second half of 2008.


Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Tylersburg-36-Double D's
By nerdye on 7/23/2007 6:02:27 PM , Rating: 4
I think I'm gonna hold out tell even cheaper quad cores and the upcoming Tylersburg-36-Double D's edition motherboards for super well endowed over clocking ability! Sorry I couldn't resist =)




RE: Tylersburg-36-Double D's
By dubldwn on 7/23/2007 6:19:43 PM , Rating: 2
I can only conclude that an American didn't come up with or approve this naming scheme. If that isn't the first thing you think of, you must have just walked out of the woods. Well, unless they did that on purpose, but that would be weird.


RE: Tylersburg-36-Double D's
By MonkeyPaw on 7/23/2007 7:24:16 PM , Rating: 2
Intel must be having their office lunches catered by Hooters.


RE: Tylersburg-36-Double D's
By jpeyton on 7/23/2007 7:41:19 PM , Rating: 1
Yet another chipset/platform change for Intel?

At least they're keeping their title for the shortest upgrade paths ever.

I thought they'd learn something from AMD and design chipsets/platforms to be good for 2-3 years at minimum, like Socket 940 was and Socket F is.


RE: Tylersburg-36-Double D's
By Pandaren on 7/23/2007 8:42:21 PM , Rating: 2
This isn't surprising at all. Nehalem has an IMC and CSI links. It would be impossible for Intel to make it compatible with FSB architecture chipsets.

Complaining about platform change here makes about as much sense as complaining that Athlon 64 didn't work with the EV6 bus that the K7s used.


RE: Tylersburg-36-Double D's
By Dactyl on 7/23/2007 10:11:00 PM , Rating: 2
Yet another?

Socket 775 has held P4, Pentium D, Core Duo, Core 2 Duo, Core 2 Quad, and it will hold Penryn.

Plus, this article is only about Intel's XEON architecture. Intel has been talking about including a memory controller on some but not all versions of its Nehalem chips (perhaps only Xeons and Extreme Edition desktop chips).

If so, that could mean that we have Nehalem (and even the 32nm refresh of Nehalem) on Socket 775.


RE: Tylersburg-36-Double D's
By DeepThought86 on 7/23/07, Rating: -1
RE: Tylersburg-36-Double D's
By retrospooty on 7/24/2007 10:20:46 AM , Rating: 1
got a bit too much anger do we?

Settle down, and really think about who is being ignorant in this thread.


RE: Tylersburg-36-Double D's
By Andrwken on 7/24/2007 2:30:44 AM , Rating: 2
The original socket 775 for prescott came out in 2004. The revised socket 775 for core2 came out in 2006. The new socket for Nehalem will be out in 2008.

Looks like they are giving you your 2 years minimum, and if it wasn't for the voltage issue with core2, you would have had 4. What's the problem?


RE: Tylersburg-36-Double D's
By kingpotnoodle on 7/24/2007 6:07:45 AM , Rating: 3
AMD haven't exactly been short on changes, while Intel have gone from 478 to 775, AMD have gone 754 - 939 (some 940) and then onto AM2, now coming is AM2+ and the backwards compatibility to AM2 is not total (due to some voltage differences) and they road mapped in AM3 in the same time frame as this Intel change.

Chipsets for both companies change all the time, plenty of single core Athlon chipsets and motherboards never supported the X2, and likewise for Intel where older boards didn't support Pentium-D and then Core2, though thats not just Intel chipsets, same is true for VIA/SIS ones.

Its often changes in the voltage regulations for new chips which cause issues, and in that respect maybe the motherboard makers who are making boards with these bare-minimum VRMs should also shoulder some blame.

In short - platform changes ruining upgrade paths are an industry problem. Although quite why you would want to team your brand new high end processor with a 2 year old motherboard/platform which is likely missing the latest storage and memory technologies is debatable...


RE: Tylersburg-36-Double D's
By FNG on 7/24/2007 8:57:40 AM , Rating: 2
Well, I believe they more or less tout the platform stabilization for driver compatibility only (i.e. imaging). Thus this is targeted at enterprises where the cost of a motherboard, processor, ram, re-installation is multiplied by 1000+. Joe consumer is not so much of a worry because he only has to buy new system from Dell or pick up some parts from Newegg and re-install his OS a handful of times.


"There's no chance that the iPhone is going to get any significant market share. No chance." -- Microsoft CEO Steve Ballmer

Related Articles
Intel Prices "Penryn" Xeons
July 17, 2007, 4:06 PM
Intel Goes Lead-Free for 45nm, Again
May 23, 2007, 4:34 AM
Intel "Bearlake" Plans Unveiled
October 17, 2006, 1:24 PM













botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki