I havent been excited for a processor for a long time.. but it reminds me of being little and waiting for christmas haha. ironically, the release will be around that time :P
2.93 + OC goodness, here i come :D
Printable View
Guys don't forget that the current DDR3 kits aren't what you should be using with Nehalem as it is tri-channel so expect to see 3Gb and 6Gb kits soon. I wonder what the price difference will be like and whether there is a significant advantage in everyday applications and games. Don't be too surprised when the prices are a great deal higher than the numbers we are seeing now. New tech is always at a price premium so depending whether you are a budget conscious [aka cheap] buyer like me then you must play your cards at the right time. Remember when the B3 Q6600 was $999? Luckily for us most of the people here will be embracing this new architecture rather early so we can see what are the ups and downs. I also wonder what the next chipset could offer and maybe X68 could offer a jump like the P35 did.
Remember that at XS you have gamers, ocers and dc members. The dc guys are drooling over Bloomfield and the sever versions.
Don't cont on much between Tylersburg and it's successor, overclocking is all done through multies and reference clock and really the only data thats being sent over the QPI on bloomfield is graphics and I/O. In both cases, the bandwidth and latency of the interconnect are way faster then anything we could need in the next few years. The only place where any improvement is really noticed is inter-socket commmunication on the DP platform.
Since the interconnect now can use adjustable multipliers rather then being stuck at "quad-pumped" there is little reason to spend time designing in scaleability on the reference clock, only the max link speed. My belief is that the main purpose of the X68 (or whatever we call it) will be PCIe 3.0 support. It might support some increases to the interconnect speed and who knows, mayby add a bit more bandwidth to the DMI link to the SB, but it probably won't help out with OC'ing locked processors.
Since it is the same die in the package, i believe support for reg/unreg is all through BIOS. The big server board guys (Tyan/Supermicro/ect.) dont have all that much reason to code in support for unreg, but without knowing much about the stuff that goes into coding for ECC, i would think that writing BIOS for ECC is way harder the unreg. So who knows, maybe they will all leave the option in for customers who don't need the memory relyability and don't want the added cost. The same people that bought the San Clemintine platform that does DP Harpertowns with regular DDR2.
better save hard! 'cos mobo will cost you like the CPU it self ;)
http://www.fudzilla.com/index.php?op...=8434&Itemid=1
Has anyone confirmed that Bloomfields can be used in 2P Tylersberg mobos? Is Intel going to release Gainstown before Bloomsfield (or at least the same time)? Looking forward to the next SkullTrail or xTrail now that nVidia has licensed SLI for Nehalem.
Next quarter is going to be interesting.
Would these new motherboards be standard ATX or will they be like EATX?
From the incident before, I'm not posted any benches results anymore ... :shrug: This SS just exclusively shows only on XS :rofl:
Just a little preview of Gainestown hope enough :clap:
Just reminder, the CINEBENCH R10 bench is impressive :up:
This SS below with VISTA 64bit + 20Gb DDR3 ECC + Gainestown 3.06GHz*2
CpuZ codename has wrong detection to Bloomfield ...
http://i245.photobucket.com/albums/g...inestown_0.jpg
...
wow. CPU-Z is of by a bit.
That's going to be a beast lol.
it's scary to think that like 6 years so, such power was only offered in a $8k+ box.
That's going to have some insane memory bandwidth. I really want one.
Now I wet my pants...
DAVE! New pants please!
NICE. Too bad we won't get any benches. . .yet :D
It helps if you just bring a laptop to the bathroom with you if you're reading this in the morning. When you go an mess yourself, you are sitting and it's all contained, or at the very least the shower isn't far away:rofl:
EDIT:
oh crap, I just noticed the 2 instances of BOINC that appear to be on JC's systray... and I just took a shower:shakes:
Just a dampness problem so the Depends are taking care of the issue.:yepp:
As to the two boinc, I think a error as BOINC will not run twice on one system but you can show 2 in the tray if you hit the exe twice.
Just a error I think with someone not that familar with the app.
Dave
A lot of talk on the X58 chipset for this, but any news on the chipset for the dual-processor version? Really interested if there are any block diagrams of it that were leaked to the web yet. Looking for a good server/workstation board.
So that would imply that for the pci-e slots they would be directly off a particular cpu (say 3 slots for cpu 0 and 3 for cpu 1) and if you had to do data transfers between those you would have to cross over from one cpu to the other. Not that bad in itself actually, but would require to fully populate the board with cpu's to use all the slots (in my case no big deal as I am planning that anyway). Still knowing the number of pci lanes for each cpu dedicated to the pcie slots would be useful, and what the speed is for the ibexpeak which would need to be tied to each cpu as well in addition to the bandwidth between the cpu's to handle process migration in addition to i/o from attached periphs.
All in all, more information would be real nice. ;)
Tylersburg is for both single and dual socket.
http://pc.watch.impress.co.jp/docs/2...gai388_09l.gifQuote:
* Tylersburg-24S – 24 PCIe lanes, 1x QuickPath Link
* Tylersburg-24D – 24 PCIe lanes, 2x QuickPath Links
* Tylersburg-36S – 36 PCIe lanes, 1x QuickPath Link
* Tylersburg-36D – 36 PCIe lanes, 2x QuickPath Links
http://babelfish.yahoo.com/translate...rUrl=Translate
Just one more QPI link.
Pretty pictures
The Flextronics board looks very 'fancy' if you will, blue pcb, yellow pci-e slots.. Maybe an ST2 competitor?
He's on about the dual-core Havendale and quad Lynnfield solutions. The dual socket version will of course have QPI and the chipset Tylersburg.
Thanks! That's exactly what I was looking for. So I have to wait for Tylersburg-36D it looks like as well as to get some performance specs on the implementation of tylserburg itself to see if it can actually handle the traffic of all those lanes unlike the current server/workstation boards.
The ones with Ibexpeak will be Lynnsfiels and Havendale. Not Bloomsfield.
Bloomsfield is a quadcore with tripple channel memory and a QPI link to a PCie switch and again with DMI from there to the ICH10.
Lynnsfield is a quadcore with dualchannel memory and an ondie 16 or 20 lane PCIe controller (I cant remember). It will use a DMI interface (2GB/sec each way) to the Ibexpeak that is essentially the same as ICH10.
Havendale is like Lynnsfield. But a dualcore and with GPU (IGP Class) on the CPU aswell.
So for Lynnsfield and Havendale its impossible to run more than 1 x16 or 2x x8 for GFX. And I doubt many would offer more than 1 x16 slot on the MBs.
I thought that Gainstown was the dual socket version and bloomsfield was the single socket version has that changed? I still don't like only having 36 PCIe lanes (really want something like 48-56 or so which is why AMD is still in the running but can't ignore the processor performance of intel).
Bloomfield is the single-socket QPI-supporting processor, Gainestown is the dual-socket QPI-supporting processor, Lynnfield and Havendale are the DMI-connected processors with onboard PCI-E. Going by your current system's specs those are not really something to be interested in, those are more for Dells and laptops from the specs they've currently released.. Besides, only Bloomfield and Gainestown is coming this year. Those other ones are due later next year.
If you look at the diagrams in my link it's also possible to have two Tylersburg chips on the board connected by QPI and to the two CPUs, for 64 PCI-E 2.0 lanes.
This should clear up some confusion...
http://members.shaw.ca/virtualrain/n...alem-chart.png
More here, including diagrams of the chipsets...
http://www.nehalemnews.com/2008/04/nehalem-faq.html
BEEEECKTONNN
Would anyone be able to buy that mofo?
Look at the prices of Tigerton and Dunnington for your answer to that..
Besides, dual socket systems will finally be able to run regular DDR, why ditch that by buying immensely expensive processors and then running FB-DIMMs? For a desktop/workstation it doesn't make sense no matter how much money you have. :rofl:
Thanks, and to the above specifically, that would be bloody awesome if some manufacturer would actually produce one. The biggest problem I've seen so far (server & workstation boards) is a complete lack of I/O performance. Having 1 or 2 slots run well is good for desktop but not when you really want to push data.
So has it been confirmed whether or not the mainstream as well as performance and server segments can be overclocked yet?
There have been numerous references to overclocked Bloomfield CPU's at Computex and elsewhere.
I'm personally expecting the 3.2GHz part to be an Extreme branded processor with unlocked multi. The other Bloomfields will likely have upward locked multi's.
More info here on how the clock/power domains will likely work...
http://www.nehalemnews.com/2008/05/e...r-domains.html
No further word other than pure speculation on Lynnfield overclockability. Even if the clock is on package for Lynnfield, there's no reason it still can't be exposed in BIOS to allow adjustment. Just because they can make it unadjustable doesn't mean they will.
I believe they're they may be thinking DDR3 should be mainstream by the time the mainstream processors and boards are released and I sure hope so.
But at this point in time it seems to be moving along fairly slowly in my opinion.
Then again wouldn't the cheapest board/Processor be using the cheapest components air go DDR2?
But am I wrong in assumeing BloomField will most likely be extreme versions only?
I'm talking more about the Q9300, Q9450, Q6600, E8400 of this new generation of CPU's being overclockable.
So its LGA 1160 for P55 and LGA 1366 for X58 ??
Perkam
Weird so basically they'll be overclocked in a way similar to an Intel/Phenom mix based on that link's article.
Guess that's entirely possible...though one could fear for overclockability since phenom's don't overclock so well.
From the DDR3 prices I can see it's not long since DDR2 was at this price level.. We're just too used to the silly low DDR2 prices.
Yes that's wrong, this BS was spread by the Fud (as always), when we know now that the 2.66GHz version will actually hit the Q9300 price bracket (not Q9450 as previously thought), the 2.93GHz will be slightly more expensive and the 3.2 is an XE processor.
Anyone have info on models/pricing of Lynnfield CPUs?...
No, but they will most likely be from some 160$ up to maybe 316 or 540$.
And havendale will be 200$ and down approx.
I dont think it will be named P55 tho. The LGA1160 southbridge will also be used for the Havendales aswell. It might rather be called PCH1 or something. Its a new ICH class.
Definitely not a fan of having separate sockets for the performance and mainstream processors but I guess that's all required given the way the PCIe lanes are handled.
Sure and it is easy to forget PCI-E 2.0 8X = PCI-1.0 16X Bandwidth Wise. These still run through the Platform Controller Hub. I'am a Fan of Multiple Sockets because I don't like paying for something that is un-needed or wasted. I wish I could save the money spent on useless built-in Realtek Crap!
I think shintai is talking about PCH ...
Just let the pics talks :p:
http://pic.xfastest.com/CPU/Nehalem/nehalem-03.jpg
http://pic.xfastest.com/CPU/Nehalem/nehalem-02.jpg
...
I checked today on the price index here in Denmark. Sure DDR2 is alittle bit cheaper. But DDR3 is already ultra cheap. I was actually very surprised because I didnt pay much attention to DDR3 prices before. But you can get 2GB DDR3-1333 for 90-100$ and 2GB DDR2-800 for 50-60$.
Sure its still a premium, but its also alot faster. But the point is its not completely screwed in prices anymore.
Thanks JC!
http://www.nehalemnews.com/
There, looks like someone knows about the Havendale MCM now. To make the low end even cheaper, it isn't really a "system on a chip" but more of a "system in a package". What you see there is a Havendale(Auberndale) CPU and the shrink of Eaglelake named Ironlake combined together on a single package.
I'm just wondering if this CPU/GPU integration will ever turn into anything more exciting than glorified IGPs.. More 'talking' between the CPU and GPU, perhaps using the GPU as a generic floating point accelerator when not doing graphics, or using it for physics/etc and using a discrete add-in board for graphics..
Are they planning anything like that for Westmere+Larrabee, or will it be just like AMD's 'Fusion', boring old IGP on-die?
Thing is 4gb sets still are alot more expensive then DDR2 which if you wanna run vista and the newest games 4GB is needed.
I'm gonna wait on DDR3 till it's 4GB and at DDR2's current prices of around $100for 4GB, at technologies ageing speed it shouldn't be long.
Perhaps, but consider it was like $400 for 4GB not long ago and now you can get that much for $200.
Also with DDR3 becomeing more mainstream and nehelem seemingly only useing DDR3 it'll advance much faster.
Your right though I have a birthday comeing up in a few month's as is.
Trust me I know the difference I know shintai is saying "even in his jerk way of saying it", that you can't run PCI-E through to the PCH because DMI doesn't allow it since it is limited to 2GB. I asked him a question, it got ignored. So, to see if I'm on an ignore list, just post something blatantly wrong and wait for his correction. Not a single reply from about 7 other posts as well. I was just checking to see if I was completely ignored or an invisible poster. Sorry folks, my bad.
You guys can guess that I'm following this thread because even at 56, I have permawood waiting for Nehalem....
BUT what I see is a lot of guys who are much smarter than I am and know much more about the internal design of the cpus essentially trying to "one up" each other with that knowledge and where beliefs collide slamming one another even if it's done in a "cutesy" way.
Is it necessary? Can't people keep an open mind and say to themselves that maybe they don't know it all and the other guy might, just might understand one area better than they do?
It really comes across at times like two 15 year olders dropping trou and saying "mine is bigger than yours"
Ego is a thing best left in another room when accessing the forum and discussing.
You learn more, you make friends vs enemies and your blood pressure stays lower.
I sit here, I read and try to comprehend a technology that wasn't available back when I was in school and probably had a few million more brain cells, at least the ones that help with learning, but I honestly wonder and am frustrated at times with the apparent need to be "right"
Voice your opinion, back it up with facts if you have them or if not, just say that this is how I "think" it works and then read what the other guys says.
WE all learn this way and at the end of the day aren't cursing on another.
No, I missed that one.:rofl:
Definately not me. I've learned from many people here and hopefully passed on some info to help others.
Maybe it's because I never saw an issue when presented with info that voided my position and saying to the guy: "Thanks, I just learned something"
Usually when I respond if I'm positive of something I wil say it plain but if I just beleive it to be so I will say "I think"..
I mean it, this place is so much better when the jabs, fighting and ego's are left home. Life is short, enjoy it and leave the fighting to the trolls.
They are worthless anyway.
does this mean we won't be able to use faster memory with the 2.66GHz chip? Those are pretty much DDR2 speedsQuote:
Originally Posted by nordichardware
thats true, i just hoped that it'd be able to use DDR3-1600+ since we have all these high speed modules now
Just to finish the SLI topic (sorry very little sleep lately)
If Quadro works on ordinary X16 slots without that 200 chip thingy (technical term :) then that chip thingy is truly just useless garbage?
It's garbage in the sense that Nvidia most likely could make it work on any platform. The only thing that I believe would be a valid explanation is that they have some silly extra something going on under the hood that's facilitating cards talking to each other directly rather then all the way back to the CPU and down, but all that should be happening over the SLI connector.
Really there's probably 2 reasons. First, because they can. They have patent on it and can make mobo guys/Intel do whatever they want to "enable" it. In this case, buying a $30 chip. Second, if they were to enable SLI on Intel, they'd have to validate it's functionality on every board that comes with an Intel chipset.
I don't really care from a personal point of view, I'm not interested in multi-GPU and especially not Nvidia's multi-GPU, but it is quite stupid.
That's what annoys me. That intel has to put an nVidia chip on the mobo that will "enable" it to do something it already technically can.:down: It increases the prices of the mobo and it's just another chip that must be cooled, we all know the thermal properties of nVidia's chips:rolleyes:.
But to keep things on-topic, i'm really looking forward to the release of Bloomfield and Tylersburg. So much so that i'm delaying building my new rig until the launch, then a few weeks for prices to settle. Such a shame, i've got all these shiny new watercooling parts lying around with nothing to use them on. Only a few more months...
I'm just rather seriously considering dual 4870x2's instead, depending on how well quad CF works out. Wanted to be sure I was seeing the lay of the land correctly (that NV is inappropriately charging us for something). I'll wait and see, but if the drivers are solid and there is less to get actively cooled on the mobo, it looks likely that dual 280's is right out on my X58.
Have to see.
Then we can just hope that every board sold isn't SLI Ready. We can't influence the whole computer market but this one we can. Each SLI ready board sold validates nVidia crappy behavior or tells them what they're doing is OK. I wish folks buying X58 would at least make an effort to NOT buy any X58 with these chips.
OR maybe the NF200 on the X58 is nVidia version of the Dongle Cable from the original CrossFire. Maybe it's a handicap? If I buy X-58 I will looking for something without it.
I doubt the first boards will have NF200. You've seen the current uniprocessor Nehalem boards, they're all based on the reference design dug up by HKEPC apart from Intel's own - they're not going to go ahead and rip that apart just for some dumb SLI tax chips. If a company does that I'd say they're destined to get buggy boards and a crappy reputation.. It's not a board I'd want, change up the design two months before release? No thank you.
I think the first ones will be the basic Tylersburg 36S reference board they all have currently, no silly green chips at first.
If the chip is needed, they could just as easily add it to the GFX card. Just like ATi used to have the CF version of card, they could have the SLI version.
The first ones wont be worth buying anyways though. its always a month or 2 later where the companies come out with their "enthusiast" version, usually incorporating any hardware tweaks/fixes from things they learned from the release of the first, also to make more money on people who would buy the first, then the second so they still have the best.
Sure, but most of the time Intel based boards are sold to users for Beta testing like the nVidia versions:rolleyes: The last four of my Intel based boards were bought within two to four weeks after they went on sale. The last being a GA-P43. It's only problem was with two Muskin 2GB sticks of RAM. Flashed BIOS and wa-la fixed. Same goes for the GA-P35 and Foxconn P35 that also needed a BIOS flash. Didn't need to wait for other revisions for stability. Overclocking have been known to improve with later revisions though.
I don't think so! Putting the chip on the card wouldn't work because all of the data still wouldn't have a Direct link to the other or Second card. I'm not saying I know for sure the chip is like a Dongle but it is a very good possibility it is. Since the PCH doesn't factor in on this chip at all. The chip could help the link to the X-58 Chipset, it gets in the way of the Direct Link to Lynnfield.
Since I trashed nVidia, lets look at from a different side. What if the Chip proves to be the best option? What if its 10 to 15% faster running with the chip than without it?
I don't want SLI, with or without it. :p:
Besides, how would that be possible, and how would you explain the scaling betwen CF and SLI in games? Shouldn't there be a disparity if their magical 'SLI processor' actually made a difference? I mean, from what I can see, if anything, CF scales better.
Look, I didn't say it was magical and clearly said MAYBE IT IS NEEDED.:up:
CF scales better because ATI started with the Dongle and corrected it from the ground up. Maybe nVidia didn't. But even in AMD's case, checkout those 4870 X2 vs 2 4870s in CF?
http://www.anandtech.com/video/showdoc.aspx?i=3354&p=1
Now please notice that excellent Bridge chip used on X2? It doesn't matter if this was added to two cards or not. The X2 card works every time, the SLI and CF doesn't. I'd take X2 any time. Sorry but 2 X2's are kind of Crazy in most cases.
If I wanted X58, and I do want it, I'm happier with one 4870 X2 than two of either nVidia or AMD cards. So I don't want or need a NF200 on the board.
The X2 doesnt work each time. It got the same driver limitations and its just a matter of time before they are equal with profiles and such.
And the "SLi Processor" is basicly a PLX chip of nVidia. PCIe switch, nothing more, nothing less.