Intel Z68 Express Chipset

[onethreehill]

Update: Intel Z68 Chipset & Smart Response Technology
http://www.anandtech.com/show/4329/i...caching-review
http://www.hardwarecanucks.com/forum...m-expands.html
http://hothardware.com/Reviews/Intel...se-Technology/
http://www.hexus.net/content/item.php?item=30346
http://www.legitreviews.com/article/1612/1/
http://www.pcper.com/reviews/Motherb...g-and-Overcloc
http://www.tomshardware.com/reviews/...hing,2938.html

Intel Smart Response Technology
http://benchmarkreviews.com/index.ph...=745&Itemid=60
http://www.hardocp.com/article/2011/...technology_srt
http://www.pcper.com/reviews/Storage...ing-Z68-Tested
http://www.pcstats.com/articleview.cfm?articleID=2593

Intel 311 Solid State Drive SSDSA2VP020G2E
http://benchmarkreviews.com/index.ph...1&limitstart=0

Enthusiasts were forced to hit the brakes on Sandy Bridge when motherboard vendors massively recalled platforms based on Cougar Point. We take a Z68 Express-based board for a spin to see if you should wait for Intel's true LGA 1155 enthusiast chipset.

At the risk of sounding like a broken record, Intel’s Sandy Bridge-based CPUs are great. But the first round of chipsets enabling the new LGA 1155 interface (H67 and P67 Express) is not....

http://www.tomshardware.com/reviews/...hing,2888.html

[badboy18187]

Well, however you look at it, I think Intel is trying to rip us off by paying extra for smthg we should already have/(like it says all throughout the article). But then again, If AMD's Bulldozer comes in less than a month after this chipset hits retail, what will the majority of buyers do?

[Cooper]

Well, however you look at it, I think Intel is trying to rip us off by paying extra for smthg we should already have/(like it says all throughout the article). But then again, If AMD's Bulldozer comes in less than a month after this chipset hits retail, what will the majority of buyers do?

+1. Intel is too late to the show with caching thing.

[Hornet331]

Well, however you look at it, I think Intel is trying to rip us off by paying extra for smthg we should already have/(like it says all throughout the article). But then again, If AMD's Bulldozer comes in less than a month after this chipset hits retail, what will the majority of buyers do?

Lol why do people always think they are entitled to anything?

If you don't like what you see.. don't buy it. Your a mature customer who can make its own decions.

As for the article, quite interesting and good to see Lucid has listen and allows now to use the IGP as virtual cpu, which menas no performance impact on games but still the use of quicksync, though you wont see any energy savings.

As for the SSD caching... hmm I wonder how it would work out for a gaming drive, it would be interesstenig to see 1-2TB caviar black with a 64gb C300, which will put it at 170-230€, give you superrior space and if caching works somewhat ok you also get better performance.

[PatRaceTin]

interesting feature

no bclk oc?

socket 2011 only

[Glow9]

Lol why do people always think they are entitled to anything?

Like a working motherboard? Bada ching!

[Johnny87au]

Lol why do people always think they are entitled to anything?

If you don't like what you see.. don't buy it. Your a mature customer who can make its own decions.

As for the article, quite interesting and good to see Lucid has listen and allows now to use the IGP as virtual cpu, which menas no performance impact on games but still the use of quicksync, though you wont see any energy savings.

As for the SSD caching... hmm I wonder how it would work out for a gaming drive, it would be interesstenig to see 1-2TB caviar black with a 64gb C300, which will put it at 170-230€, give you superrior space and if caching works somewhat ok you also get better performance.

If its the same price then I can only say positive things but honestly this should of been released with the current SB chips.. I'm sure as hell not going to get one of these unless the igp can clock along with the CPU quite decent and would be suitable for a htpc, as lately that's all I've been interested in..

[Andrew LB]

Well, however you look at it, I think Intel is trying to rip us of). f by paying extra for smthg we should already have/(like it says all throughout the article

Kinda like how AMD did the same as Intel and changed only one pin on the motherboard socket for their new chip? The only difference is that AMD advertised back then that thost AM3 boards would be compatible with the new processors. Intel never did.

But then again, If AMD's Bulldozer comes in less than a month after this chipset hits retail, what will the majority of buyers do?

If Bulldozer was such a monster and "coming out any day now" (i've heard that for a year now) like so many of the peanut gallery are claiming, then AMD's choice to not flood the market with Bulldozer chips while Intel was having to return ALL of their P67 chipset boards was about as idiotic as i've ever seen. No wonder why many of AMD's top brass jumped ship?

I really hope AMD proves me wrong and Bulldozer is a monster chip. But I don't see any signs pointing toward being anything more than another product that the average "joe consumer" sees as being a slightly slower, cheaper alternative to Intel which he would have bought if he only had a few more bucks.

[onethreehill]

Intel Z68 Chipset Confirmed To Arrive On May 8th
http://en.expreview.com/2011/04/11/i...tml#more-16069

[Mats]

Kinda like how AMD did the same as Intel and changed only one pin on the motherboard socket for their new chip? The only difference is that AMD advertised back then that thost AM3 boards would be compatible with the new processors.

Not really.

- Socket AM3 will be 2.5 years old when BD shows up, 1156 was half as old when 1155 came.

- That extra pin doesn't stop board manufacturers from making current AM3 boards compatible. Sure, not all models, but it's still better than none.

[Brahmzy]

The SSD caching thing is worthless to most enthusiasts. I'll keep my P67 until x79...

[[XC] hipno650]

some of the new features are kinda cool but considering I would never use both of the major changes I will pass. Honestly this B3 rev of my Maximus IV Extreme is hand downs the BEST board I have ever used in my life... I have no issue with it, it's cool, stable, overclocks well, has all the features I could ever want and a decent slot layout...

[Ket]

Like a working motherboard? Bada ching!

My Extreme4 is a B2 board, stable as a rock.. or Asrock, if you will Only thing that will make me buy a Z68 is if I could use the IGP on my 2500k to do Physics work, if not, I'll pass.

[STEvil]

Kinda like how AMD did the same as Intel and changed only one pin on the motherboard socket for their new chip? The only difference is that AMD advertised back then that thost AM3 boards would be compatible with the new processors. Intel never did.

Dont forget Intel did the 1-pin thing first many years ago with P4's and mobile chips... and 771/775..

[zanzabar]

so u cannot overclock the base and it turns the IGP on, how is this an enthusiast chipset.

[sdsdv10]

so u cannot overclock the base and it turns the IGP on, how is this an enthusiast chipset.

I far as I concerned, it not "an enthusiast chipset" and was never meant to be. That is reserved for socket 2011 SB-E systems.

Just look at this road map at another thread here is XS: http://www.xtremesystems.org/forums/...d.php?t=269415

Only one CPU, the 2600 (or the 2600K) even make it into the premium preformance sector. The rest are mainstream or below. Even then, the 2600 will be overlapping with SB-E when it's finally launched.

[Hornet331]

so u cannot overclock the base and it turns the IGP on, how is this an enthusiast chipset.

The whole SB line up till now is not in the enthuisast segment to beginn with... its mainstream... or upper mainstream.

As sdsdv10 said, enthusiasts will get there fix with SB-E.. Octocores here I come.

[saaya]

How is this an enthusiast chipset?
Looks more like workstation?

[[XC] Synthetickiller]

I was under the impression a while back z68 was going to be something like dual x16 PCIe slots for 1155. I was horribly wrong.

What's the real point of this? If you need performance, get a raid card. At least you can move it from build to build. After having a raidchip go on a supermicro board, I'll never do raid onboard again.

[CrazyNutz]

Quick sync, meh, I'll pass. My 2500k + GTX460's transcode a full DVD in 5min's or less using DVDFab.

SSD Caching, not interested either

I could not be happier, the 2500k + p67 have far exceeded my expectations in both value, and performance

[zanzabar]

I far as I concerned, it not "an enthusiast chipset" and was never meant to be. That is reserved for socket 2011 SB-E systems.

Just look at this road map at another thread here is XS: http://www.xtremesystems.org/forums/...d.php?t=269415

Only one CPU, the 2600 (or the 2600K) even make it into the premium preformance sector. The rest are mainstream or below. Even then, the 2600 will be overlapping with SB-E when it's finally launched.

what dose that have to do with the socket, and the IB will be out by the time the SB-e will and will most likely kill the sb-e unless u want more than 3 gpus like the 1156 did the 1366 and then the 1155 did to everything else on intel. but since they are throwing the "mainstream" 1st now there will be no reason for the high end. also way off topic the article says its the enthusiast platform for the 1155 but it has the IGP on and dose not change the no base clock. the Z was also for workstations in the past,

The whole SB line up till now is not in the enthuisast segment to beginn with... its mainstream... or upper mainstream.

As sdsdv10 said, enthusiasts will get there fix with SB-E.. Octocores here I come.

sb-e will not be for enthusiests its a server socket and i will bet will not clock as well as the 1155. and this is about chipsets being labeled not platforms or sockets.

[chispy]

This chipset its mehh :/ ... Not interesting at all.

[clayton]

Z68 itx!

[sdsdv10]

what dose that have to do with the socket

SKT1155 = mainstream (Z68, P67 and H67 are all chip sets for SKT1155, thus the relevance to this post and to the socket type)
SKT2011 = enthusiast/performance

or at least that is my understanding from reading the road map, plus everything else read on these forums and elsewhere. A CPU with IGP isn't meant for enthusiasts, just like Clarkdale isn't an enthusiast CPU, Nehalem is (i.e. SKT1156 = mainstream, SKT1366 = enthusiast/performance). Don't know how be any more clear than that. It certainly is great how well the 2600K clocks, but presumably SB-E on SKT2011 should be even better. I guess we will just have to wait and see if that is true.

and the IB will be out by the time the SB-e

Uh no, didn't you look at the roadmap? SB-E in 4Q11 and IB in 1Q12. I don't believe there is anyway we will see IB before 2012.

sb-e will not be for enthusiests its a server socket and i will bet will not clock as well as the 1155. and this is about chipsets being labeled not platforms or sockets.

SKT1366 was also a server based chip set and it clocked quite a bit better than SKT1156. I never really saw the 655K catch on while 920's, then 930's and now 950's were around. Why do you thing we won't see the same performance advantage we saw with Nehalem?

[zanzabar]

SKT1155 = mainstream (Z68, P67 and H67 are all chip sets for SKT1155, thus the relevance to this post and to the socket type)
SKT2011 = enthusiast/performance

or at least that is my understanding from reading the road map, plus everything else read on these forums and elsewhere. A CPU with IGP isn't meant for enthusiasts, just like Clarkdale isn't an enthusiast CPU, Nehalem is (i.e. SKT1156 = mainstream, SKT1366 = enthusiast/performance). Don't know how be any more clear than that. It certainly is great how well the 2600K clocks, but presumably SB-E on SKT2011 should be even better. I guess we will just have to wait and see if that is true.



Uh no, didn't you look at the roadmap? SB-E in 4Q11 and IB in 1Q12. I don't believe there is anyway we will see IB before 2012.



SKT1366 was also a server based chip set and it clocked quite a bit better than SKT1156. I never really saw the 655K catch on while 920's, then 930's and now 950's were around. Why do you thing we won't see the same performance advantage we saw with Nehalem?

the article is calling the p67 mainstream and the the z67 the enthusiast of the platform, the z67 is like the g series past though so thats my point.

the sb-e high end will be out q4, i doubt that we will see a sub $600 at launch.

on the 1366v1156, i dont think u looked at how the 1156 clocks for quads v the 1366 quads, the 1156 clocks higher base clocks but there was no unlocked chip for a long time and it only has 16x pci-e so its not something that people bench with (as u need 8x per gpu and u bench with 4), but for air and liquid the 1156 clocks better and i think that with ln2 it even clocks better than the 1366 quads. sure the hex cores clock better than the 1156 quads but only the extreme editions and then the 1155 clocks better than that. the 1366 only caught on since it came out before the 1156, but with the 1155 coming out before the 2011 and then the IB will be out at the same time that the upper mid 2011 parts out i dont see the SB-e being the performance leader unless the core count is really low on the IB.

also the 1156 has better ram speeds and performance than the 1366 quads even though the 1366 has tri channel, its a server chip base and so it has all of the server stuff like being optimized for ecc jdec ram speeds so the on chip latency is more than what it would be if it was optimized for quicker non ecc ram like the 1156 was. thats why the made the 1156 have less channels even though it shared a server platform it was made to have lower latency as u did not need the extra channels since the core bandwidth needs were what 1600mhz dual channel could bring. the 1366 also has a token style qpi buss that allows for 3 links per cpu (one to the imc one to NB and one for the cpu to cpu link) the 1366 desktop dose only have 2 but its still shared, while the 1156 and 1155 have the qpi linking the IMC only and the pci-e is linked to an SM buss from the l3 cashe, making the IMC function at a much lower latency and overhead.