Intel’s Latest CPU creation, code named Sandy Bridge is here and it is alleged to be very fast. One thing for sure, it is not as expensive as the other top of the line Intel CPU’s are. So, recently I decided to give the i7 2600K a test drive and so far, it is faster than anything else I have used. As a comparison, a kernel compile on a i7 870 (2.9 GHz and fully optimized) took 13 minutes while on a i7 2600K (3.4 GHz totally stock setup) it took only 8 minutes. It is working under openSUSE 11.3 with kernel 2.6.37 with no problems. I have not yet tried to use the built-in graphics (stuck in an nVidia card) as I think it is too soon, but perhaps in a few months, you can use it. Here are some articles I have found on the subject. Please feel free to add your own:
So the real thing that got me interested was the cost to switch to Sandy Bridge. If you build your own computers and thus use a standard clone of recent manufacturer, you can just swap out the motherboard, CPU, Fan & memory and reuse the rest of your setup. I found the CPU for $300 (Intel i7 2600K) at Fry’s, motherboard (MSI H67MA-E45) was $120 & memory (Patriot G Series Sector 5 DDR3 2x4GB) was $110 for 8 GB DDR3 for a total cost of $ 530 US plus tax. It cost less than an upgrade to an Intel i7 870 this summer and is at least 40% faster. It is hard to beat such performance. But what do you think? Please post your comments here…
This is about computer hardware, it is not pre-release and is not your general chit-chat about the Rain in Spain, so as far as I am concerned it is in the right place, but I am sure if it needs to be moved, it will be.
But the Hardware forum is in the Help section. It appears to be general chit-chat about your latest acquisition and prices. You don’t appear to have a problem concerning openSUSE, just soliciting for comments i.e. chit-chat.
But the Hardware forum is in the Help section. It appears to be general chit-chat about your latest acquisition and prices. You don’t appear to have a problem concerning openSUSE, just soliciting for comments i.e. chit-chat.
consused, do you have a specify comment you would like to make about the Intel Sandy Bridge CPU’s you would like to post here in following with the subject of this thread?
Well, one thing is that it isn’t a top of the line CPU. Technically (that is, pedantically) Sandy Bridge is a range of CPUs, or an architecture, depending on how fine-grained your definition of architecture happens to be.
Presumably, it won’t be all that long before a Sandy Bridge variant that is top of the line is released. Maybe Intel is waiting to time their counterblast against Bulldozer?
That was interesting, but largely because it doesn’t apply here. Usually, the ‘early adopter premium’ is higher here…as well as general prices being that bit higher, as is the taxation, all of which are vexations.
To be honest, that (on-chip graphics) was the bit that I was most interested in, so please do post back when you have actual results, even if the initial results are just ‘this doesn’t work with that kernel version’, wherever this thread ends up.
I do agree that Sandy Bridge in its present form may be slow compared to what it may do in the future. The present version is just dual channel for instance, which is a savings over the triple channel thing. Sandy Bridge actually undercuts their present CPU’s in almost every way. You can not ignore the new Intel CPU’s if you are AMD or nVidia.
It is said that you could spend an extra $75 and beat the Intel built in graphics of Sandy Bridge and we all know that graphics is not something that Intel has done well. But wait, how many people are already on the fence on buying a add-in video card? What if the Intel Graphics do not suck? It means more people will stick with Intel and not buy an add-in video card. What about Laptops and video? Intel, if the graphics are way better than what they had before, this could kill anyone else putting out their own video for Laptops.
What is Intel really trying to do, beside making their more expensive CPU’s obsolete? They have pointed their big guns at both AMD and nVidia. It is doubtful nVidia can be beat on the desktop by Intel, but they could be irrelevant in Laptops. AMD has got a double barrel shotgun pointing at their heads. They must fend off Intel with its much cheaper higher power CPU’s while Intel takes up the cheap video and nVidia takes up the high powered video. Now AMD has proved it can compete in both worlds on price and performance. What happens if they can not compete on price any more?
And what about the PCI bus? What about dual video cards. The new Intel CPU Chipset allows one video card but not two. How will nVidia get those SLI cards installed when Sandy Bridge Takes over? Will anyone be willing to pay a lot extra to get the two to work together? I am not sure, but none of this looks too good right now for either nVidia or AMD. I would not bet against either one and we need the competition from both, but has Intel finally produced the A-bomb that blows the competition away for good?
The neat thing about Intel graphics, is when they work well, they are GREAT ! But when they do not work well, one is left with the FBDEV driver which is very poor in performance (albeit it works with almost anything).
With ATI if the proprietary fglrx driver does not work well, one has the radeon, and radeonhd, and vesa and only then if necessary the very slow performance FBDEV driver.
With nVidia, if the proprietary nvidia driver does not work well, one has the nouvea, and nv, and vesa and only then if necessary the very slow performance FBDEV driver.
I like the Intel only WHEN it works, because the latest Intel hardware versions provide HD acceleration to offload video decoding from CPU to GPU and I think it works with all Linux video players. It typically easily survives a kernel re-install (as opposed to the ATI/nVidia proprietary drivers).
To get that HD video CPU to GPU offload capability in nVidia in Linux, one needs to use vdpau (which currently is only supported by mplayer, kdenlive and mythtv and NOT yet by vlc nor xine nor totem nor other players). And to get that capability in ATI in Linux is even worse. Intel has a DEFINITE lead here IF they can get their Linux drivers working properly, which unfortunately to date they have had a very spotty record.
But when Intel drivers work, they work NICE . REALLY REALLY NICE.
The location of this thread may prompt discussion, but the substance is of (IMO) sufficiently greater significance than some of the threads in General Chit-chat. The Sandy Bridge arrival has a closer involvement to the general subjects here than Ballmer’s take on pirated software. That being said, I will still follow this thread.
As to Intel’s intentions, in the technology marketplace in general and regarding AMD and Nvidia specifically, are subjects that give stock-traders insomnia and ulcers, while honing the almost natural procrastination of PC-upgraders everywhere. Pricing information notwithstanding, Intel’s market direction seems focused away from netbooks/laptops and to more lucrative pastures of tablets and smart(er) phones.
One consistent approach with Intel has been “what are we missing” ? Several possibilities have been mentioned (the PCI bus, etc), and we are in for an experience. I, for one, have not yet forgotten Arrandale, and still remember earlier Intel surprises.
It is time to re-introduce Sun Tzu to my nightly reading selections.
Just because some threads in Chit-chat are of lesser significance, doesn’t mean they all are. In fact Soapbox could be a better choice as it’s for strong opinion (check out the description). If Soapbox contained more threads like this one, it might be better as a regular forum for sensible discussion.
Intel purportedly announced earlier this morning that it has discovered a flaw in the 6-series chipsets that accompany the new processor family. While Intel reassures users that they can “continue to use their systems with confidence,” they have still halted chipset shipments until a new, bug-free version of the silicon starts to ship out late next month.
Thanks so much for finding and posting this information oldcpu. Basically, the Intel Chipset that supports the Sandy Bridge CPU has a problem with the (4) total 3 GB speed SATA ports that come out of the Intel Chipset. They are said to degrade over time causing high data error rates and corrupting your data with at least a 5% failure rate over three years and a probable higher failure rate after that. The two 6 GB speed SATA ports are said to not affected. No word on how these motherboards will be replaced just yet.
It is suggested that you:
Connect your hard drives to the two 6 GB speed SATA ports.
Purchase an add-in PCIe adapter card for any over run (more than two) SATA hard drives and for your DVD drives.
Option two is just to make sure you have no problems with any other SATA devices. Any product replacement at this time has not been specified and so spend any extra money at your own risk, but the same thing applies to the continued using of the (4) total 3 GB speed SATA ports from the defective Intel chipset.
I think I will order another Vantec 4+1 SATA II 300 & PATA PCI-E Combo Host Card w/Raid from Amazon for $27.99. My motherboard is MSI, but who know when that might get replaced.
Since no one else reported it, sometime back (around 08-Feb-11) Intel reported they were shipping again their shipments of the Intel chipsets (Sandy bridge) for PC configurations that are not impacted by the design issue: Intel resumes shipping of faulty Sandy Bridge chipsets | bit-tech.net
“After causing chaos among motherboard makers by revealing a flaw in its 6-series motherboard chipsets, Intel has announced plans to recommence shipments of the faulty silicon, before the fixed chips have even started shipping. Intel claims it decided to start reshipping the chipsets after lengthy discussions with computer manufacturers. “As a result of these discussions and specific requests from computer makers,’ says the company, 'Intel is resuming shipments of the Intel 6-series chipset for use only in PC system configurations that are not impacted by the design issue.” The announcement follows Intel’s recent exposure of a well publicised design fault that affects the 3Gbps SATA ports (typically ports 2 to 5) in Intel’s P67 and H67 chipsets. As such, we assume that the new systems based on the faulty chipsets will either come with a separate SATA controller card, or that they will only use the two (unaffected) 6Gbps SATA ports provided by the chipset.”
Intel has officially announced that the first family of chipsets Cougar Point with the new B3 stepping will ship to customers (OEM manufacturers, developers, motherboards, etc.) as early as February 14. As the number of chipsets being transferred to the new value of stepping include Intel 6 Series models for desktop and mobile computers as well as chipsets line C200, targeted for use in desktop and server systems. http://it-chuiko.com/uploads/posts/2011-02/1297619980271281217.png
Among the major innovations implemented in chipsets Cougar Point stepping B3, can distinguish small changes in the metal layer, which is believed to help eliminate problems with lost productivity, ports, SATA 3.0 Gbps. Recall, it was this defect that led to the postponement of the release and even withdrawal from some PC manufacturers, and has caused so soon release an updated set of system logic.
In addition, the products of B3 stepping revision ID changed from 04h to 05h and will be provided update BIOS, with the construction of a new stepping chipsets are pin compatible with the previous version. Add that the number of chipsets, which should be transferred to the new stepping, the model includes Intel Q67, H67, H61, P67, Q65, QM67, HM67, HM65, B65, UM67, QS67, Z68, C202, C204 and C206.
Both are interesting … but the 13-line patch is especially eye catching. It helps address the rather annoying problem with the earlier situation where the MS-Windows driver was 2x to 3x faster than the Linux driver, and that lead has with this patch purportedly been narrowed to the MS-Windows driver being 27% faster than the Linux driver. That’s still a big lead, but its also a big improvement to see the lead cut by so much.
Both are interesting … but the 13-line patch is especially eye catching. It helps address the rather annoying problem with the earlier situation where the MS-Windows driver was 2x to 3x faster than the Linux driver, and that lead has with this patch purportedly been narrowed to the MS-Windows driver being 27% faster than the Linux driver. That’s still a big lead, but its also a big improvement to see the lead cut by so much.
So, right now I just put in an nVIDIA 450 and use their driver. However, someday in the future, support for the built in Intel graphics may be there and provide better support with an open source driver, as I read it. I will look to that day as the as the actual upgrading of what was not clear to me at the moment.
