Home  |  News  |  Reviews  | About Search :  HardWare.fr 

MiscellaneousStorageGraphics CardsMotherboardsProcessors
Advertise on BeHardware.com

News of the day

  • Imbroglio around the SandForce SF-1200
  • LGA1155 and LGA2011 for Intel
  • Archives

    Avril 2010
    1 2 3 4
    5 6 7 8 9 10 11
    12 13 14 15 16 17 18
    19 20 21 22 23 24 25
    26 27 28 29 30

    You can read previous news using our archives.

    << Previous day's news
    April 16, 2010
    Next day of news >>
    April 19, 2010

     Imbroglio around the SandForce SF-1200
      Posted on 17/04/2010 at 23:59 by Marc
    Imprimer cette news Envoyer cette news par e-mail

    AnandTech has given us something to chew on on the subject of the SandForce SF-1200 controllers and their firmware. As you may know, one of the main differences between the SF-1200, for the general consumer market, and the entrerprise SF-1500 is a purely software limitation on performance. The SF-1200 can reach 10,000 IO/s in random writes, against 30,000 IO/s for the SF-1500.

    Where things become a little strange is that following an agreement between SandForce and OCZ, OCZ has been given the privilege of having a specific firmware for its Vertex 2 that’s based on the SF-1200 which allows it to bypass the software limitation and attain 30,000 IO/s. The 10,000 IO/s limitation on all other SF-1200 based SSDs is effective in the SandForce firmware 3.0.5 which has been allocated to SSDs on sale commercially.

    Where things are a bit funny however in terms of this agreement, is that firmware 3.0.1, the “release candidate” SandForce version, doesn’t have this limitation. Yet, even though this firmware wasn’t designed for sale commercially, several manaufacturers, including Corsair, have delivered SF-1200 SSDs that use this version. The manufactuer had moreover to disactivate an energy economy feature on the SF-1200 that endanagered the SSD with version 3.0.1. of this firmware.

    So then, quite an imbroglio, with, on the one hand, SSD manufacturers delivering SSDs with a non-definitive firmware, which is in itself quite problematic for such a young technology, and on the other an update of this firmware which will limit the performance of the SF-1200 to planned levels, putting a spanner in the works of the initial tests. Decidedly, all SSD controllers have a difficult beginning and SandForce is no exception!

     LGA1155 and LGA2011 for Intel
      Posted on 17/04/2010 at 23:39 by Marc
    Imprimer cette news Envoyer cette news par e-mail

    VR-Zone has managed to obtain some info on the next generation Intel platforms designed for Sandy Bridge and planned for end 2010 / start 2011. This information should of course be taken with a pinch of salt but does nevertheless seem credible. Intel is reported then to be continuing with two distinct platforms as is the case currently with LGA1156 and LGA1366.

    For the mid-range, the LGA1156 will be replaced by LGA1155. Designed for 2 and 4 core forthcoming Sandy Bridge versions, this Socket will be supported by P67 and H67 chipsets which will replace the current P55s and H55s. P67 chipsets will not support the IGP built onto Sandy Bridge but will will enable the use of its PCI-Express controller at 1x16 or 2x8 for SLI/CrossFire configurations, while the H67 will support the IGP but be limited to 1x16. In terms of functionality, SATA 6 Gbit/s will be available on 2 of the 6 SATAs supported by the chipset but USB 3 is unfortunately not on the agenda. The PCI will no longer be integrated in these chipsets.

    At the high end is the LGA2011 platform (!) designed for Sandy Bridge-Es. These processors are supposed to be coming out in 4 and 6 core versions and there is reported to be a 15 MB cache in this last version (12 MB of L3 and 6x512 KB of L2, against 6x256 KB currently). For memory, DDR3-1600 is to be supported on 4 channels, which in part explains the increase in the number of contact points. There are to be 40 PCI-Express 3.0 lanes on this platform, against 36 2.0 lanes on the X58 platorm.

    In a big change, the PCI-Express is said to be supported directly by the CPU, as is already the case with LGA1156, and not by the chipset. The complete integration of the northbridge sounds the death nell for IOH and QPI and there will therefore only be a PCH linked to the Sandy Bridge-E via a DMI bus at 2 GB/s and a PCI-E 3.0 x4 bus at 4 GB/s. This new PCH still won’t however support USB 3, but it will support 6 SATAs in AHCI of which 2 SATA 6 Gbit/s, as well as 8 other SATA3/SAS ports.

    << Previous day's news
    April 16, 2010
    Next day of news >>
    April 19, 2010

    Copyright © 1997- Hardware.fr SARL. All rights reserved.
    Read our privacy guidelines.