I think this is to combat Bulldozers release that should be soon. I find it toooo convenient that this was "leaked" so close to the original update a few days ago. It's possible it was missed, but I'm betting that wasn't the case.
Yeah, I think Intel is saving this for Bulldozer launch. They have had over 9 months to work on these CPUs (which are essentially the same chips with an extra multiplier) so missing the launch sounds laughable.
What's laughable is your common sense. Consider this.
The top Bulldozer consumer chip will be AMD's top chip for the consumer retail market. Sandy Bridge and Socket 1155 are Intel's mainstream lineup. Intel doesn't need to position Sandy Bridge entirely against Bulldozer when it still has Sandy Bridge E and Socket 2011. Sure they will be more expensive butt hey will be faster, Intel can charge what it wants for the fastest chips and value hunters can settle for AMD.
I don't think this in any way relates to the impending launch of Bulldozer, it's just Intel working their usual quarterly refresh.
The process has matured enough to allow equal or better yields that handle the higher frequency so they'll introduce a new chip with that as the default.
Correct. AMD traditionally does a couple of speed bumps per year, coinciding with improvements in the manufacturing process. Intel has moved to a slower release schedule during the year but we still do see at least one speed bump per year.
I know it's nothing extraordinary, Intel does this every year. I just don't think it's a coincidence that Intel left out 2700K from the update a week ago and that Bulldozer is right around the corner. Sure it could be due to some internal issues but obviously, we know nothing about them.
Totally agree. And it's interesting, because today Intel knows Bulldozer's performance: I am sure of that (Bulldozer is shipping already to OEMs, and I am certain that many OEMs, basically, belong to Intel). It might be good sign that Bulldozer's performance is somewhat competitive. This said, let's not forget that the Bulldozer currently out there are for servers, while the i7 2700k is for desktop ...
May be just a tweak, but its often been said there is no advantage for most people to the 2600K over the 2500K. It may be unlikely but it is possible there will be an effort to surpass 2500K performance by more at stock levels. Speculation only.
Depends on the usage. For gaming, 2500K is the king since games are poorly multithreaded and the bottleneck is elsewhere. For usage which takes advantage of multithreading, Hyper-Threading can boost the performance by up to 30%.
Indeed, I went with the 2600K over the 2500K for the sole reason of reducing compilation times.
Turns out that was a great choice, as I were working on a minor web application for most of the spring term and I measured a compilation time drop from 40s to 28s in the early stages of development.
While 30% doesn't sound like a lot when you're measuring it in seconds, or single-digit minutes, it does add up when you're waiting on it for large parts of the day.
I very much doubt the 3820 will be worth the money. Quad-channel memory and more cache aren't likely to do nearly as much for performance as the unlocked status of the 2500K/2600K.
And it's going to be notably more expensive as well, when considering both the processor, motherboard and memory modules.
Recently found out that the 2600K doesn't have Intel's Virtualization Technology for Directed I/O (VT-d), yet the vanilla 2600 has it. I sure hope Intel sees the light and makes sure VT-d is included in the 2700K, because it is utterly inane the 2600K lacks it.
Same situation with the 2500 vs 2500K, as well. It is disingenuous of Intel to not include the same functionality in the K series parts.
It very probably will be missing VT-d. It's pretty logical. VT-d is used by enterprise users and others who have financial resources. Intel does not want those users to overclock, Intel wants them to buy more CPUs.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
21 Comments
Back to Article
Lunyone - Saturday, September 10, 2011 - link
I think this is to combat Bulldozers release that should be soon. I find it toooo convenient that this was "leaked" so close to the original update a few days ago. It's possible it was missed, but I'm betting that wasn't the case.Just my $0.02
Kristian Vättö - Saturday, September 10, 2011 - link
Yeah, I think Intel is saving this for Bulldozer launch. They have had over 9 months to work on these CPUs (which are essentially the same chips with an extra multiplier) so missing the launch sounds laughable.TypeS - Saturday, September 10, 2011 - link
What's laughable is your common sense. Consider this.The top Bulldozer consumer chip will be AMD's top chip for the consumer retail market. Sandy Bridge and Socket 1155 are Intel's mainstream lineup. Intel doesn't need to position Sandy Bridge entirely against Bulldozer when it still has Sandy Bridge E and Socket 2011. Sure they will be more expensive butt hey will be faster, Intel can charge what it wants for the fastest chips and value hunters can settle for AMD.
Mr Perfect - Monday, September 12, 2011 - link
Intel wants value hunters settling for a 2700K instead of AMD.Exodite - Saturday, September 10, 2011 - link
I don't think this in any way relates to the impending launch of Bulldozer, it's just Intel working their usual quarterly refresh.The process has matured enough to allow equal or better yields that handle the higher frequency so they'll introduce a new chip with that as the default.
Both Intel and AMD do this all the time.
Anand Lal Shimpi - Saturday, September 10, 2011 - link
Correct. AMD traditionally does a couple of speed bumps per year, coinciding with improvements in the manufacturing process. Intel has moved to a slower release schedule during the year but we still do see at least one speed bump per year.Take care,
Anand
Kristian Vättö - Sunday, September 11, 2011 - link
I know it's nothing extraordinary, Intel does this every year. I just don't think it's a coincidence that Intel left out 2700K from the update a week ago and that Bulldozer is right around the corner. Sure it could be due to some internal issues but obviously, we know nothing about them.yankeeDDL - Sunday, September 11, 2011 - link
Totally agree.And it's interesting, because today Intel knows Bulldozer's performance: I am sure of that (Bulldozer is shipping already to OEMs, and I am certain that many OEMs, basically, belong to Intel).
It might be good sign that Bulldozer's performance is somewhat competitive.
This said, let's not forget that the Bulldozer currently out there are for servers, while the i7 2700k is for desktop ...
Ushio01 - Saturday, September 10, 2011 - link
So it's a consumer version of either of these.http://ark.intel.com/products/52278
http://ark.intel.com/products/55452
infoilrator - Saturday, September 10, 2011 - link
May be just a tweak, but its often been said there is no advantage for most people to the 2600K over the 2500K.It may be unlikely but it is possible there will be an effort to surpass 2500K performance by more at stock levels.
Speculation only.
Kristian Vättö - Saturday, September 10, 2011 - link
Depends on the usage. For gaming, 2500K is the king since games are poorly multithreaded and the bottleneck is elsewhere. For usage which takes advantage of multithreading, Hyper-Threading can boost the performance by up to 30%.Exodite - Saturday, September 10, 2011 - link
Indeed, I went with the 2600K over the 2500K for the sole reason of reducing compilation times.Turns out that was a great choice, as I were working on a minor web application for most of the spring term and I measured a compilation time drop from 40s to 28s in the early stages of development.
While 30% doesn't sound like a lot when you're measuring it in seconds, or single-digit minutes, it does add up when you're waiting on it for large parts of the day.
gevorg - Saturday, September 10, 2011 - link
Meh. This is just to combat Bulldozer, pretty useless for current 2500K and 2600K owners.ilkhan - Saturday, September 10, 2011 - link
I'm more interested in the question of "will Intel release an i5-2600 and/or i5-2600k/s when these come out to replace the i5-2500(k).Doesn't matter to me, I bought my 2500K on release day, and not upgrading for a while. But its an interesting question.
Exodite - Saturday, September 10, 2011 - link
I very much doubt the 3820 will be worth the money. Quad-channel memory and more cache aren't likely to do nearly as much for performance as the unlocked status of the 2500K/2600K.And it's going to be notably more expensive as well, when considering both the processor, motherboard and memory modules.
Exodite - Sunday, September 11, 2011 - link
Not that I know, though you will need 4 modules to actually make use of the wider memory bus where S1155 chips can get away with two.Which naturally is going to cost more, even at the same total memory amount, unless you want at least 16GB memory in total.
Kougar - Sunday, September 11, 2011 - link
Recently found out that the 2600K doesn't have Intel's Virtualization Technology for Directed I/O (VT-d), yet the vanilla 2600 has it. I sure hope Intel sees the light and makes sure VT-d is included in the 2700K, because it is utterly inane the 2600K lacks it.Same situation with the 2500 vs 2500K, as well. It is disingenuous of Intel to not include the same functionality in the K series parts.
Mr Perfect - Monday, September 12, 2011 - link
Yes, I saw that article too. Pretty surprising, really. I thought the K series was all powerful.http://arstechnica.com/gadgets/news/2011/08/what-p...
I'd actually go with a vanilla i7-2700 if they make one, VT-d and TXT are things I'd actually want.
KaarlisK - Monday, September 12, 2011 - link
It very probably will be missing VT-d. It's pretty logical.VT-d is used by enterprise users and others who have financial resources. Intel does not want those users to overclock, Intel wants them to buy more CPUs.
tuklap - Monday, September 12, 2011 - link
trigate is the crown!!Shadowmaster625 - Monday, September 12, 2011 - link
Where is the i5-2600(k)? Or would it be a i5-2520k?