I'd expect spelling mistakes on a tech site, but writing that a DX11 GPU will only support OGL 3.1 is seriously wrong. But OpenCL 10.1 ?? You expect Ivy Bridge in 10 years or what ?
Relax, it's an extra 0. I've fixed the typo. As for OpenGL support, the real issue there is going to be drivers, which is something Intel has never really focused on in their IGPs. Sandy Bridge does fine in DX games (relatively speaking), but I know of at least a couple OpenGL titles where performance is pathetic. I'm not entirely sure where Kristian got the OpenGL/OpenCL support information, though, so I've added "(?)" to the areas where we're not certain of specs.
OpenGL and OpenCL were reported by a German site who leaked quite a lot of Sandy Bridge data too which turned out to be right IIRC. Anand mentioned 16 EUs in IB already in the SB review and I also think he mentioned something about the OpenCL support. The same site is also reporting that the IB IGP will come in flavors of 6 and 16 EUs.
What the heck is your problem? It was a simple typo, if you had any clue what the correct versions were you could have politely replied that it was wrong and what the correct terminology would be. I heard MS is bringing back Clippy, but I hope he is not built to be as much as an @$$ as you.
typo ? typo would be 11 (a missing .) but 10.1 implies 1.01. Current OpenCL is 1.1, I guess Intel would know better than to design future products to implement old technologies.
Oh lord! They hit an extra key on the number pad! What's next? Outright lies? I, for one, will never voluntarily read this top-notch source of information again. In fact I think I see an angry mob forming! (That was sarcasm, you crazy person.)
Probably Intel not wanting people to whine and complain when they buy DDR3-1333 that's not running at the DDR3-1600 rated speed the memory controller is spec'ed at. DDR3-1600, as far as I recall, is still considered somewhat in the upper "power user/enthusiast" realm compared to mainstream DDR3-1066 and DDR3-1333 SDRAM.
The official value has to work in a 100F room with the system clogged in dirt. Our systems are much less hostile environments so we can push the parts harder.
Intel also doesn't want to toss the 1 or 2% of chips with the poorest IMCs when a few hundred extra MHz has almost no impact on normal use.
So that's somewhat humorous: laptop chips have DDR3-1600 support with SNB, but desktop parts like the 2600K don't. Of course, laptop RAM tends to be less power hungry in the first place which might be part of it. Anyway, the *desktop* Ivy Bridge will get official DDR3-1600 support, which is something desktop Sandy Bridge doesn't have.
I didn't realize it was different for mobile systems. It could also be that 99% of laptops only have room for 1 dimm/channel which makes the bus signaling cleaner.
Ah, that could be the reason. I know some laptops that will sell you DDR3-1600, but only if you don't use four DIMMs (look at the mobile workstations from Dell and HP). Still, it's odd considering enthusiasts have run far higher than DDR3-1600 speeds. Oh well.
It'll be interesting to see what Intel does with the added die space that 3D manufacturing gives them. Already, the largest singular part of SNB is the GPU, dwarfing the size of any individual CPU core. Hopefully we'll get something that's better than their previous offerings.
probably the GPU performance will enhance a lot, lets hope they get into the range of basic gaming gpu, but the main problem is that by the time that IVY arrives the AMD APU Fusion series will also have launched there second gen, and it is known already that this will be much more powerfull.
It's not just that, by the time IB has launched TSMC should have it's 28nm process working which means that AMD/nVidia's discrete cards will be twice as fast (twice the transistor density) leaving the IGPs back in the dust again.
SB's IGP wouldn't look as good as it did if TSMC's 32 nm process hadn't been scrapped giving intel a half node advantage. They'll have the same with IB, but only going from 12 to 16 EUs makes it likely they're using a big chunk of the gain to cut costs instead of boosting performance.
It doesn't matter how fast IGP's are compared to discrete graphics (it never has been). What matters is if the IGP allows for playable framerates at common desktop/notebook resolutions (we're talking 1280X1024 19" and similar widescreen resolutions). People with 1080p and higher resolutions won't be gaming on these, the other 95% of people that go to Best Buy and Dell to get a system will be.
And we need these "awesome" IGP's now. They maintain the bottom teir graphics that development houses still have to cater too (and consoles). Once those graphics capabilities become "good enough" for moderate resolutions the game makers can finally start cranking up the graphics quality and giving these new cards a workout.
I've never gamed on an IGP and never plan too, but I understand just how important they are to us gamers.
X67/68 do not exist. Z68 will come out soon, and like P67/H67 and a few others we mention, they will support Ivy Bridge with the appropriate firmware/BIOS updates. (That will be more a factor of the motherboard manufacturers taking time to update the firmware than anything.) X58 will be replaced by X79, which as we noted in an earlier article is a completely different socket.
looks like it's not going quite as fast as was expected; prior rumors had it stripped from the x79 chip too.
OTOH if filling in all the expansion slots in an ATX board would've been tricky with x79, it's going to be even worse for enthusiast z7x boards unless the number of lanes on the southbridge is increased, especially if DMI isn't given a boost to allow the SB lanes to run at 3.0 speeds as well. 2x gb nics, and 3rd party audio will take 3 lanes. Additional usb3 and sata 6gb controllers will take 1 each. That leaves at most 3 SB lanes to make 1x slots. The situation is worse if the board makers insist on adding legacy firewire or pata ports. Thunderbolt ports will IIRC eat a PCIe lane each as well. I suspect 4 to 8 lane pcie bridge chips (or possibly PCIe to PCI bridge chips) will end up being fairly common on these boards.
Do the current 6-series chipsets connect to the CPU with a 20Gbps DMI link, which isn't even enough for all the potential peripherals if they're all maxed out, but in practice that almost never happens.
8x PCIe 2.0 lanes = 40Gbps 2x SATA 6Gbps = 12Gbps 4x SATA 3Gbps = 12Gbps 14x USB 2.0 = 6.72Gbps Gigabit Ethernet = 1Gbps -------------------------------- Total Potential = 71.72Gbps!
I believe the 7-series is still using DMI 2.0, so it remains at 20Gbps, but now the total potential bandwidth for all ports is:
8x PCIe 2.0 lanes = 40Gbps 4x USB 3.0 = 19.2Gbps 2x SATA 6Gbps = 12Gbps 4x SATA 3Gbps = 12Gbps 10x USB 2.0 = 4.8Gbps Gigabit Ethernet = 1Gbps -------------------------------- Total Potential = 89Gbps
Realistically, the SATA ports are pretty much never going to all be 100% utilized at the same time, and the same goes for the PCIe ports and USB ports. It's possible Intel will tweak the DMI interface to boost bandwidth as well, which could easily accommodate the additional bandwidth heavy devices.
I was looking at the number of devices to connect, not total bandwidth (pcie to pci or pcie-pcie (more lanes) bridge chips don't help with that.
My observation is that after you connect all the devices to the PCIe bus that need to be connected, you don't have enough lanes left to fill out all 7 slots on an enthusiast level full atx board. Depending on the chipset you have anywhere between 1 and 3 slots off the CPU, and at most 3 1x slots off the southbridge. With thunderbolt potentially eating a 6th, and a few of the 6x series boards having enough USB3 controllers to have 8 or 10 USB3 ports, it's entirely possible to use up all of the southbridge's 8 lanes just on onboard controllers. The main advantage I see in potentially running the SB lanes at PCIe3 speeds isn't total bandwidth which isn't likely to be saturated; but allowing board makers to use 4xUSB,3 4x SATA6GB, or 2x GB nic controllers to get their port inflation numbers up while only consuming a single PCIe lane instead of 2.
That or they should bump the number of lanes on the southbridge from 8 to 12ish to allow connecting as many low average bandwidth devices as they could when the PCI controller was still available. IIRC the controller on Intel's current desktop boards supports 5 PCI devices.
If you remember, current P67 motherboards share PCIe lanes. If you drop a card in the bottom PCIe gfx 4x slot, it disables some on-board things (firewire, some USB headers, etc). In essence, the lanes are shared amongst the devices already.
I wasn't aware of anything like that, I assume the disabled headers are USB3 since there's no reason to disable any of the USB2 ports. The firewire port surprises me though, I'd always assumed it would have been stuck on the legacy PCI bus since there's plenty of capacity there and no need for the higher speeds of PCIe.
While you're probably right about it not seeing much traction in the enthusiast market, assuming it trickles down to the mobile chipset as well, support for a third monitor would be really nice for my work laptop.
This basically are the rumors we have heard from other places summed up here. Of coz being Anandtech this means the rumors are realistic.
The EU are properly a lot different to current SandyBridge, We expect to have double the Gfx Performance from Ivy Bridge. If any of the previous Intel slide are to be trusted. 22nm would also allow much higher Clock Speed for GPU part.
I hope Anandtech could find out whether Ivy Bridge have FMA.
$100 for a 20GB SLC SSD is really expensive, compare to the previous rumors of 40 - 50 USD.
Ivy Bridge IGP will only have 16 EUs which is only 33% more than what SB has. Like Anand mentioned in the SB review, it unfortunately looks unlikely that the IB IGP would be twice as fast, which is a shame.
In the original version, I said the "Larsen Creek" SSD will be priced at around 50$ like the rumors have suggested but apparently Jarred edited that. Of course it is safer to say 100$ and then be happily surprised when it turns out to be 50$ instead of getting those angry comments when it turned out to be 100$ instead of 50$.
My thought is that these "Next Gen EU " would be completely different to current EU, that is why it is not fair to compare EU numbers to performance.
And if this new EU runs 33% faster per clock, and runs at 33% higher clock while having 33% more EU then previous EU, should land us at 100% performance increase.
Of coz, Hardware dont matter much on GPU, not at all. It is software, drivers that makes ALL the difference. As we have seen with S3. Nvidia has double the Software Engineering compare to hardware working on it. And if Intel continue the way it is, their GPU hardware will never get the respect they deserve.
I think Intel GPUs get pretty much the respect they deserve. As long as they keep crippling the mainstream parts by putting in crap like HD2000 they don't deserve any respect. The only decent GPU they have is the HD3000 in the 'k' series which will never be used by the people that buy 'k' series parts (ok, 1% might).
Yeah, but all of the mobile chips pretty much get HD 3000, and they're the place you're more likely to use the IGP. Maybe Ivy Bridge will be better in this respect; guess we'll wait and see.
Kristian, thanks for an article full of details. However, I find this article harder to read than necessary. Because it is peppered with abbreviations that I'm not familiar with. They might be daily use for an Intel enthisiast but they have not yet reached the mainstream.
May I suggest AnandTech adds to their awesome Benches a Glossary and links the first use (in an article) of any name or abreviation to a short definition page with a list of serach results for the term (good Google juice if you knwo what I mean). That would make it so much easier to brush up on a term the reader is not familiar with.
Alternatively (or in addition) spell out the names of technologies with appreviations at theri first use, such as S... R... T ... (SRT) so that one can form a mental word for the appreviation and better understand what is said.
That said, keep up the good work. One wants to read these articles and understand in detail as opposed to skimming just for buzz words and chatter (slashdot I'm looking at you).
P.S.: The formatting of tables, etc. often does not scale with enlarging the font, but that is for another post.
Thanks for the feedback. I'm still learning and sometimes it's just hard to use the full name instead of the abbreviation, mainly because in the forums I go everyone uses abbreviations.
While I don't want to blame Jarred for this, in the original version at least the SRT was first mentioned as Smart Response Technology (SRT) so people get the idea of what the SRT means. In the original version, there is also info on the Z68 and SRT in general but due to the NDA, we couldn't publish it so that might have added confusion since some parts of the other areas of the article referred to that.
If you really want to be helpful, send me an email (click my name above the article) and give me more examples of the cases where you wish that the abbreviation should be explained more clearly :-)
I also did a google search for "Intel SRT" (didn't know what it was) and got to Intel's site that talks about "System Recovery Technology" - same acronym two different things...
Just to clarify, we are under NDA on Smart Response Technology until Z68 launches, at which point Anand will have a full explanation as well as performance results. That also takes care of the Larsen Creek SSD, so I'm as interested as the next guy to see what it will do for performance. Basically, I had to cut a whole page from the review where Kristian discussed SRT and Z68 features, but we'll have the official launch soon enough to fill in the gaps. :-)
This roadmap screams to me that Intel does not support Thunderbolt for non Apple systems.
I'd have expected the next set of chipsets to support it for display and other purposes. Instead we see USB 3.0 accross teh board. Not that that is bad, but the lack of Thunderbolt in all of 6 new chipsets is really sendign a message (It's Apple only technology) despite all the statements otherwise.
Thunderbolt is dumb. Many systems get 1 port. That gets used when you plug in your display via DisplayPort, and there it goes completely unused! Even if you have a spare port and find something to plug into it, it won't be any faster than USB3 would be. So what's the point?
I think Apple wanted it since it was "cool, light!" for their marketing, and then engineering common-sense managed to win, since metal wires are cheaper. Intel likes it since they get to sell boatloads of Intel-only controllers. USB3 makes more sense in many ways.
Thunderbolt supports daisy-chaining so you can connect up to 7 devices into one port. I'm pretty sure most of the TB devices will support DC when they start rolling out. Plus, you can always get a hub if you have devices that don't support.
Moreover, there are currently two computers from Apple that have TB so you can't conclude that most systems in the future will be stuck with one port. The 27" already has two BTW.
Can someone answer these questions for me? would appreciate it.
Will X79 support Ivy Bridge CPU's? And if so... then will Ivy Bridge CPU's be available for X79 at around the same time as IB CPU's for the lower platforms or will the high end be left out again?
No. X79 will be for Sandy Bridge-E, which uses socket LGA2011. There's a schism between the ultra-high-end enthusiast platforms and the "mainstream enthusiast" platforms, which is a continuation of the X58 and P55 split. The problem is that where X58 was clearly superior to P55 platforms in most performance metrics, X58 vs. P67 suddenly didn't look so compelling. We'll eventually get the "fix" for that in Q3/Q4 when SNB-E launches, but just like X58 vs. P67 we'll have X79 vs. Z77 in 1H'2012 to make things difficult.
Thanks Jarred. So the X79 platform will have no CPU upgrade path at all? It will ONLY ever take SNB-E CPU's? Being as Ivy Bridge is just a "tick" and not a new architecture i would have thought X79 would get a IB CPU for LGA2011 that just needs a BIOS update.
I'd still be wary of the crap that Intel did to us 1366 users. Promising an 'affordable' 6 core upgrade, but never releasing a chip coming even close to $500. That's what happens when AMD can't compete.
Yeah, i7-970 was 880$ when it was launched in July last year (before that there was only 999$ 980X). It was dropped to 583$ when i7-990X was launched in February IIRC.
That would make sense. It looks like Intel has switched their product release cycles. Before, it was high-end chips first and then mainstream chips later on (at least this was the case with Nehalem, I'm too lazy today to check anything older that that :D). Now it is the vice versa. Mainstream chips in H1 and high-end/enterprise in H2.
This make sense though since enterprise level stuff is more complex and thus needs more time and money to be designed, tested etc.
It does make sense in some ways, the other half though is that there're lots of people who bought 2500/2600K systems that probably would've bought LGA2011 quads at the higher margins if the high end part launched first.
Lets say someone drops an Ivy Bridge CPU in an existing 6 series motherboard, could it run at DDR3-1600 or only at 1333? Technically, the memory controller is in the CPU die, so it sounds reasonable to say it would do 1600. On the other hand though, the motherboard would have to be able to handle the speed and would doubtlessly have some sort of UEFI settings related to it. Any thoughts here?
Personally I want to see confirmation that Ivy Bridge will be on X79 (in good time) otherwise there would be no hesitation. But I forsee a dead end for X79 like with X58 and I just don't like it. If IB won't be on X79 quickly enough and IB turns out to be quite a bit faster at a smaller process.. but that means waiting for IB. Bloody hell
Yup, that's what I gathered. Frankly I don't see any point in LGA 2011 if, as I expect, IB is much faster than SB at the same clock speeds. You then get to wait for IB while the supposedly more 'mainstream' platform is faster than the high-end platform.
I would say the primary market for LGA 2011 parts is the enterprise market. The mainstream products (LGA 1155) lack many necessary features (PCIe lanes, ECC, DP setups...) that are required by servers and higher-end workstations. While there is Xeon E3-1200 lineup, it does not offer more than four cores and there is no option for DP setup either.
For enterprises, the raw CPU performance is not always the number one concern. Fast I/O options are usually crucial and the 16 PCIe 2.0 lanes provided by mainstream SB may not be enough for those needs. In most cases, the performance per watt is also essential, especially in servers which are online 24/7. That is why Xeon lineups often offer low-power options.
From consumer's standpoint, SB-E may look like it makes no sense but in the end, it is not really aimed at them. There will be few CPUs for enthusiasts but other than that, it's solely Xeons.
I really wanna slap someone over at Intel in their Marketing Dept.
Why isnt this new chipset simply the X69? did all the engineer's giggle too much or do they somehow think the SB-E NB/PCH is magically next generation because its taking so long to release.
If you're not including USB3 or LightPeak, it doesn't deserve to be a new generation.
Its bad enough there will be 4 different sockets for Servers next year, and I'm sure at least a dozen procs per socket (except maybe on the ultra high end, half dozen there more likely)
Anyone know if the Ivy Bridge IGP is supposed to support 120Hz desktop? I don't expect true stereoscopic 3D on it, but it'd be nice for a 3D laptop to not switch from 120 to 60 when running on IGP. 120Hz would also be helpful for playing 24FPS blu-ray.
"A final interesting point for many users is that Ivy Bridge is pin compatible with Sandy Bridge, and it will work on current LGA1155 motherboards with the appropriate chipset and a firmware and BIOS update (H61, H67, P67, and Z68 are capable of support IB). Intel will also launch new 7-series chipsets, which we’ll get into below."
Yeah, I believe it when I see it. They said the same thing with the 915/925 chipsets, but when it came down to it, they changed the VRM requirements and controls and a new motherboard was required for Core 2 chips anyway. I bet they do the same thing.
On the other side of the fence, I have a Nvidia Nforce 590SLi board (socket AM2) that supports even the newest 6 core chips with only a bios upgrade, if I want to replace it. I recently upgraded it from the Athlon 64 6400+ chip (3.2GHz with no overclocking potential and 125W power consumption) to a Athlon II X2 260 (3.2GHz, overclockable to 3.7GHz, and only 65W power usage) for $70. It doesn't sound like an upgrade, but the power savings alone will pay for it in less than 6 months, considering its my file server.
Intel does have a big history of making people upgrade an entire platform to upgrade. I don't expect them to change that.
You guys are under NDA about Z68 until next week, yet there's already a bunch of information available on computer parts websites for Z68 motherboard pre-orders XD
appart from what i have seen looks it nice but why does nt intel take out the integrate hd graphics in 3 type of processors and make them hexacore with hyperthreating. i mean sb is extremely powerfull and ivy bridge wil be more powerfull and more efficient. and for me it does not make that much sense such powerull procesoors and igpu?
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.
Facebook
Twitter
RSS
74 Comments
Back to Article
haplo602 - Friday, May 6, 2011 - link
I'd expect spelling mistakes on a tech site, but writing that a DX11 GPU will only support OGL 3.1 is seriously wrong. But OpenCL 10.1 ?? You expect Ivy Bridge in 10 years or what ?JarredWalton - Friday, May 6, 2011 - link
Relax, it's an extra 0. I've fixed the typo. As for OpenGL support, the real issue there is going to be drivers, which is something Intel has never really focused on in their IGPs. Sandy Bridge does fine in DX games (relatively speaking), but I know of at least a couple OpenGL titles where performance is pathetic. I'm not entirely sure where Kristian got the OpenGL/OpenCL support information, though, so I've added "(?)" to the areas where we're not certain of specs.Kristian Vättö - Friday, May 6, 2011 - link
OpenGL and OpenCL were reported by a German site who leaked quite a lot of Sandy Bridge data too which turned out to be right IIRC. Anand mentioned 16 EUs in IB already in the SB review and I also think he mentioned something about the OpenCL support. The same site is also reporting that the IB IGP will come in flavors of 6 and 16 EUs.futrtrubl - Saturday, May 7, 2011 - link
Now it just needs to be fixed in the text too. 2nd paragraph after the table.Shane1 - Friday, May 6, 2011 - link
What the heck is your problem? It was a simple typo, if you had any clue what the correct versions were you could have politely replied that it was wrong and what the correct terminology would be. I heard MS is bringing back Clippy, but I hope he is not built to be as much as an @$$ as you.haplo602 - Friday, May 6, 2011 - link
typo ? typo would be 11 (a missing .) but 10.1 implies 1.01. Current OpenCL is 1.1, I guess Intel would know better than to design future products to implement old technologies.NARC4457 - Friday, May 6, 2011 - link
It was a simple mistake, you don't have to be a douche about it.andymcca - Friday, May 6, 2011 - link
Oh lord! They hit an extra key on the number pad! What's next? Outright lies? I, for one, will never voluntarily read this top-notch source of information again. In fact I think I see an angry mob forming!(That was sarcasm, you crazy person.)
DanNeely - Friday, May 6, 2011 - link
The mob is forming, but we're massing in front of haplo602's house, not Kristian's.Megatomic - Friday, May 6, 2011 - link
Haha, this is like the nitpicking they do to the writers over at DailyTech.DanNeely - Friday, May 6, 2011 - link
The difference is that some of the writers at DT regularly make major gaffe's not just silly typos.Kristian Vättö - Friday, May 6, 2011 - link
It is long journey to come all the way to Finland so I doubt anyone would voluntarily travel here anyway :Dasuglax - Friday, May 6, 2011 - link
I know some Sandy Bridge processors already support up to DDR3-1600. The i7-2820QM in my laptop does.JarredWalton - Friday, May 6, 2011 - link
Unofficially, yes, but official support is supposed to be DDR3-1333.asuglax - Friday, May 6, 2011 - link
Intel has it listed as such here: http://ark.intel.com/Product.aspx?id=52227&pro...AstroGuardian - Friday, May 6, 2011 - link
What are all these official-unofficial BS? Either it supports 1600 or not? Why complicating simple things?Stahn Aileron - Friday, May 6, 2011 - link
Probably Intel not wanting people to whine and complain when they buy DDR3-1333 that's not running at the DDR3-1600 rated speed the memory controller is spec'ed at. DDR3-1600, as far as I recall, is still considered somewhat in the upper "power user/enthusiast" realm compared to mainstream DDR3-1066 and DDR3-1333 SDRAM.TL;DR = lowest common denominator?
DanNeely - Friday, May 6, 2011 - link
The official value has to work in a 100F room with the system clogged in dirt. Our systems are much less hostile environments so we can push the parts harder.Intel also doesn't want to toss the 1 or 2% of chips with the poorest IMCs when a few hundred extra MHz has almost no impact on normal use.
JarredWalton - Friday, May 6, 2011 - link
So that's somewhat humorous: laptop chips have DDR3-1600 support with SNB, but desktop parts like the 2600K don't. Of course, laptop RAM tends to be less power hungry in the first place which might be part of it. Anyway, the *desktop* Ivy Bridge will get official DDR3-1600 support, which is something desktop Sandy Bridge doesn't have.DanNeely - Friday, May 6, 2011 - link
I didn't realize it was different for mobile systems. It could also be that 99% of laptops only have room for 1 dimm/channel which makes the bus signaling cleaner.JarredWalton - Friday, May 6, 2011 - link
Ah, that could be the reason. I know some laptops that will sell you DDR3-1600, but only if you don't use four DIMMs (look at the mobile workstations from Dell and HP). Still, it's odd considering enthusiasts have run far higher than DDR3-1600 speeds. Oh well.dagamer34 - Friday, May 6, 2011 - link
It'll be interesting to see what Intel does with the added die space that 3D manufacturing gives them. Already, the largest singular part of SNB is the GPU, dwarfing the size of any individual CPU core. Hopefully we'll get something that's better than their previous offerings.duploxxx - Friday, May 6, 2011 - link
probably the GPU performance will enhance a lot, lets hope they get into the range of basic gaming gpu, but the main problem is that by the time that IVY arrives the AMD APU Fusion series will also have launched there second gen, and it is known already that this will be much more powerfull.DanNeely - Friday, May 6, 2011 - link
It's not just that, by the time IB has launched TSMC should have it's 28nm process working which means that AMD/nVidia's discrete cards will be twice as fast (twice the transistor density) leaving the IGPs back in the dust again.SB's IGP wouldn't look as good as it did if TSMC's 32 nm process hadn't been scrapped giving intel a half node advantage. They'll have the same with IB, but only going from 12 to 16 EUs makes it likely they're using a big chunk of the gain to cut costs instead of boosting performance.
7Enigma - Tuesday, May 10, 2011 - link
It doesn't matter how fast IGP's are compared to discrete graphics (it never has been). What matters is if the IGP allows for playable framerates at common desktop/notebook resolutions (we're talking 1280X1024 19" and similar widescreen resolutions). People with 1080p and higher resolutions won't be gaming on these, the other 95% of people that go to Best Buy and Dell to get a system will be.And we need these "awesome" IGP's now. They maintain the bottom teir graphics that development houses still have to cater too (and consoles). Once those graphics capabilities become "good enough" for moderate resolutions the game makers can finally start cranking up the graphics quality and giving these new cards a workout.
I've never gamed on an IGP and never plan too, but I understand just how important they are to us gamers.
brokenbeaker - Saturday, May 7, 2011 - link
3D doesn't give them extra spece, it's the process shrink that does that. 3D gives them higher "on" current and less leakage current.Drazick - Friday, May 6, 2011 - link
When Intel get back to the SSD Controllers business?Will we see Intel Based Controller SSD this year?
Thanks.
imaheadcase - Friday, May 6, 2011 - link
That is the one everyone wants, less power/cheaper..Marlin1975 - Friday, May 6, 2011 - link
Let me guess, those that bought X67/8 chipsets will have to buy another board to use IB?Not that intel would ever do that.
JarredWalton - Friday, May 6, 2011 - link
X67/68 do not exist. Z68 will come out soon, and like P67/H67 and a few others we mention, they will support Ivy Bridge with the appropriate firmware/BIOS updates. (That will be more a factor of the motherboard manufacturers taking time to update the firmware than anything.) X58 will be replaced by X79, which as we noted in an earlier article is a completely different socket.DanNeely - Friday, May 6, 2011 - link
looks like it's not going quite as fast as was expected; prior rumors had it stripped from the x79 chip too.OTOH if filling in all the expansion slots in an ATX board would've been tricky with x79, it's going to be even worse for enthusiast z7x boards unless the number of lanes on the southbridge is increased, especially if DMI isn't given a boost to allow the SB lanes to run at 3.0 speeds as well. 2x gb nics, and 3rd party audio will take 3 lanes. Additional usb3 and sata 6gb controllers will take 1 each. That leaves at most 3 SB lanes to make 1x slots. The situation is worse if the board makers insist on adding legacy firewire or pata ports. Thunderbolt ports will IIRC eat a PCIe lane each as well. I suspect 4 to 8 lane pcie bridge chips (or possibly PCIe to PCI bridge chips) will end up being fairly common on these boards.
JarredWalton - Friday, May 6, 2011 - link
Do the current 6-series chipsets connect to the CPU with a 20Gbps DMI link, which isn't even enough for all the potential peripherals if they're all maxed out, but in practice that almost never happens.8x PCIe 2.0 lanes = 40Gbps
2x SATA 6Gbps = 12Gbps
4x SATA 3Gbps = 12Gbps
14x USB 2.0 = 6.72Gbps
Gigabit Ethernet = 1Gbps
--------------------------------
Total Potential = 71.72Gbps!
I believe the 7-series is still using DMI 2.0, so it remains at 20Gbps, but now the total potential bandwidth for all ports is:
8x PCIe 2.0 lanes = 40Gbps
4x USB 3.0 = 19.2Gbps
2x SATA 6Gbps = 12Gbps
4x SATA 3Gbps = 12Gbps
10x USB 2.0 = 4.8Gbps
Gigabit Ethernet = 1Gbps
--------------------------------
Total Potential = 89Gbps
Realistically, the SATA ports are pretty much never going to all be 100% utilized at the same time, and the same goes for the PCIe ports and USB ports. It's possible Intel will tweak the DMI interface to boost bandwidth as well, which could easily accommodate the additional bandwidth heavy devices.
DanNeely - Friday, May 6, 2011 - link
I was looking at the number of devices to connect, not total bandwidth (pcie to pci or pcie-pcie (more lanes) bridge chips don't help with that.My observation is that after you connect all the devices to the PCIe bus that need to be connected, you don't have enough lanes left to fill out all 7 slots on an enthusiast level full atx board. Depending on the chipset you have anywhere between 1 and 3 slots off the CPU, and at most 3 1x slots off the southbridge. With thunderbolt potentially eating a 6th, and a few of the 6x series boards having enough USB3 controllers to have 8 or 10 USB3 ports, it's entirely possible to use up all of the southbridge's 8 lanes just on onboard controllers. The main advantage I see in potentially running the SB lanes at PCIe3 speeds isn't total bandwidth which isn't likely to be saturated; but allowing board makers to use 4xUSB,3 4x SATA6GB, or 2x GB nic controllers to get their port inflation numbers up while only consuming a single PCIe lane instead of 2.
DanNeely - Friday, May 6, 2011 - link
That or they should bump the number of lanes on the southbridge from 8 to 12ish to allow connecting as many low average bandwidth devices as they could when the PCI controller was still available. IIRC the controller on Intel's current desktop boards supports 5 PCI devices.Ammaross - Friday, May 6, 2011 - link
If you remember, current P67 motherboards share PCIe lanes. If you drop a card in the bottom PCIe gfx 4x slot, it disables some on-board things (firewire, some USB headers, etc). In essence, the lanes are shared amongst the devices already.DanNeely - Friday, May 6, 2011 - link
I wasn't aware of anything like that, I assume the disabled headers are USB3 since there's no reason to disable any of the USB2 ports. The firewire port surprises me though, I'd always assumed it would have been stuck on the legacy PCI bus since there's plenty of capacity there and no need for the higher speeds of PCIe.DanNeely - Friday, May 6, 2011 - link
While you're probably right about it not seeing much traction in the enthusiast market, assuming it trickles down to the mobile chipset as well, support for a third monitor would be really nice for my work laptop.iwod - Friday, May 6, 2011 - link
This basically are the rumors we have heard from other places summed up here. Of coz being Anandtech this means the rumors are realistic.The EU are properly a lot different to current SandyBridge, We expect to have double the Gfx Performance from Ivy Bridge. If any of the previous Intel slide are to be trusted. 22nm would also allow much higher Clock Speed for GPU part.
I hope Anandtech could find out whether Ivy Bridge have FMA.
$100 for a 20GB SLC SSD is really expensive, compare to the previous rumors of 40 - 50 USD.
Kristian Vättö - Friday, May 6, 2011 - link
Ivy Bridge IGP will only have 16 EUs which is only 33% more than what SB has. Like Anand mentioned in the SB review, it unfortunately looks unlikely that the IB IGP would be twice as fast, which is a shame.In the original version, I said the "Larsen Creek" SSD will be priced at around 50$ like the rumors have suggested but apparently Jarred edited that. Of course it is safer to say 100$ and then be happily surprised when it turns out to be 50$ instead of getting those angry comments when it turned out to be 100$ instead of 50$.
iwod - Friday, May 6, 2011 - link
My thought is that these "Next Gen EU " would be completely different to current EU, that is why it is not fair to compare EU numbers to performance.And if this new EU runs 33% faster per clock, and runs at 33% higher clock while having 33% more EU then previous EU, should land us at 100% performance increase.
Of coz, Hardware dont matter much on GPU, not at all. It is software, drivers that makes ALL the difference. As we have seen with S3. Nvidia has double the Software Engineering compare to hardware working on it. And if Intel continue the way it is, their GPU hardware will never get the respect they deserve.
I am looking forward to the Larsen Creek SSD.
fic2 - Friday, May 6, 2011 - link
I think Intel GPUs get pretty much the respect they deserve. As long as they keep crippling the mainstream parts by putting in crap like HD2000 they don't deserve any respect. The only decent GPU they have is the HD3000 in the 'k' series which will never be used by the people that buy 'k' series parts (ok, 1% might).JarredWalton - Friday, May 6, 2011 - link
Yeah, but all of the mobile chips pretty much get HD 3000, and they're the place you're more likely to use the IGP. Maybe Ivy Bridge will be better in this respect; guess we'll wait and see.ssj4Gogeta - Sunday, May 8, 2011 - link
Actually in that case it will be 1.33*1.33*1.33=2.35 times faster.7Enigma - Tuesday, May 10, 2011 - link
Oh if only math worked that way in real life. :)Conficio - Friday, May 6, 2011 - link
Kristian,thanks for an article full of details. However, I find this article harder to read than necessary. Because it is peppered with abbreviations that I'm not familiar with. They might be daily use for an Intel enthisiast but they have not yet reached the mainstream.
May I suggest AnandTech adds to their awesome Benches a Glossary and links the first use (in an article) of any name or abreviation to a short definition page with a list of serach results for the term (good Google juice if you knwo what I mean). That would make it so much easier to brush up on a term the reader is not familiar with.
Alternatively (or in addition) spell out the names of technologies with appreviations at theri first use, such as S... R... T ... (SRT) so that one can form a mental word for the appreviation and better understand what is said.
That said, keep up the good work. One wants to read these articles and understand in detail as opposed to skimming just for buzz words and chatter (slashdot I'm looking at you).
P.S.: The formatting of tables, etc. often does not scale with enlarging the font, but that is for another post.
Kristian Vättö - Friday, May 6, 2011 - link
Thanks for the feedback. I'm still learning and sometimes it's just hard to use the full name instead of the abbreviation, mainly because in the forums I go everyone uses abbreviations.While I don't want to blame Jarred for this, in the original version at least the SRT was first mentioned as Smart Response Technology (SRT) so people get the idea of what the SRT means. In the original version, there is also info on the Z68 and SRT in general but due to the NDA, we couldn't publish it so that might have added confusion since some parts of the other areas of the article referred to that.
If you really want to be helpful, send me an email (click my name above the article) and give me more examples of the cases where you wish that the abbreviation should be explained more clearly :-)
HilbertSpace - Friday, May 6, 2011 - link
I also did a google search for "Intel SRT" (didn't know what it was) and got to Intel's site that talks about "System Recovery Technology" - same acronym two different things...JarredWalton - Friday, May 6, 2011 - link
Just to clarify, we are under NDA on Smart Response Technology until Z68 launches, at which point Anand will have a full explanation as well as performance results. That also takes care of the Larsen Creek SSD, so I'm as interested as the next guy to see what it will do for performance. Basically, I had to cut a whole page from the review where Kristian discussed SRT and Z68 features, but we'll have the official launch soon enough to fill in the gaps. :-)Conficio - Friday, May 6, 2011 - link
This roadmap screams to me that Intel does not support Thunderbolt for non Apple systems.I'd have expected the next set of chipsets to support it for display and other purposes. Instead we see USB 3.0 accross teh board. Not that that is bad, but the lack of Thunderbolt in all of 6 new chipsets is really sendign a message (It's Apple only technology) despite all the statements otherwise.
DanNeely - Friday, May 6, 2011 - link
Thunderboilt is just a combination displayport/PCIe lane in a single plug. I don't see any reason why it would need special chipset support.AnnonymousCoward - Friday, May 6, 2011 - link
Thunderbolt is dumb. Many systems get 1 port. That gets used when you plug in your display via DisplayPort, and there it goes completely unused! Even if you have a spare port and find something to plug into it, it won't be any faster than USB3 would be. So what's the point?I think Apple wanted it since it was "cool, light!" for their marketing, and then engineering common-sense managed to win, since metal wires are cheaper. Intel likes it since they get to sell boatloads of Intel-only controllers. USB3 makes more sense in many ways.
Kristian Vättö - Saturday, May 7, 2011 - link
Thunderbolt supports daisy-chaining so you can connect up to 7 devices into one port. I'm pretty sure most of the TB devices will support DC when they start rolling out. Plus, you can always get a hub if you have devices that don't support.Moreover, there are currently two computers from Apple that have TB so you can't conclude that most systems in the future will be stuck with one port. The 27" already has two BTW.
B3an - Friday, May 6, 2011 - link
Can someone answer these questions for me? would appreciate it.Will X79 support Ivy Bridge CPU's?
And if so... then will Ivy Bridge CPU's be available for X79 at around the same time as IB CPU's for the lower platforms or will the high end be left out again?
JarredWalton - Friday, May 6, 2011 - link
No. X79 will be for Sandy Bridge-E, which uses socket LGA2011. There's a schism between the ultra-high-end enthusiast platforms and the "mainstream enthusiast" platforms, which is a continuation of the X58 and P55 split. The problem is that where X58 was clearly superior to P55 platforms in most performance metrics, X58 vs. P67 suddenly didn't look so compelling. We'll eventually get the "fix" for that in Q3/Q4 when SNB-E launches, but just like X58 vs. P67 we'll have X79 vs. Z77 in 1H'2012 to make things difficult.B3an - Friday, May 6, 2011 - link
Thanks Jarred.So the X79 platform will have no CPU upgrade path at all? It will ONLY ever take SNB-E CPU's? Being as Ivy Bridge is just a "tick" and not a new architecture i would have thought X79 would get a IB CPU for LGA2011 that just needs a BIOS update.
DanNeely - Friday, May 6, 2011 - link
IIRC some of the leaked intel roadmaps have IB based LGA2011 chips coming out in late 2012, about a year after the SB ones.Pneumothorax - Friday, May 6, 2011 - link
I'd still be wary of the crap that Intel did to us 1366 users. Promising an 'affordable' 6 core upgrade, but never releasing a chip coming even close to $500. That's what happens when AMD can't compete.Kristian Vättö - Friday, May 6, 2011 - link
Well, there is i7-970 at 583$ but yeah, I have to agree with you. Sandy Bridge runs circles around Gulftown and is much cheaper.DanNeely - Friday, May 6, 2011 - link
Isn't that a fairly recent price drop? IIRC before SB launched the only hexes they offered were in the $900ish range.Kristian Vättö - Friday, May 6, 2011 - link
Yeah, i7-970 was 880$ when it was launched in July last year (before that there was only 999$ 980X). It was dropped to 583$ when i7-990X was launched in February IIRC.Kristian Vättö - Friday, May 6, 2011 - link
That would make sense. It looks like Intel has switched their product release cycles. Before, it was high-end chips first and then mainstream chips later on (at least this was the case with Nehalem, I'm too lazy today to check anything older that that :D). Now it is the vice versa. Mainstream chips in H1 and high-end/enterprise in H2.This make sense though since enterprise level stuff is more complex and thus needs more time and money to be designed, tested etc.
DanNeely - Friday, May 6, 2011 - link
It does make sense in some ways, the other half though is that there're lots of people who bought 2500/2600K systems that probably would've bought LGA2011 quads at the higher margins if the high end part launched first.Mr Perfect - Friday, May 6, 2011 - link
Lets say someone drops an Ivy Bridge CPU in an existing 6 series motherboard, could it run at DDR3-1600 or only at 1333? Technically, the memory controller is in the CPU die, so it sounds reasonable to say it would do 1600. On the other hand though, the motherboard would have to be able to handle the speed and would doubtlessly have some sort of UEFI settings related to it. Any thoughts here?don_k - Friday, May 6, 2011 - link
Personally I want to see confirmation that Ivy Bridge will be on X79 (in good time) otherwise there would be no hesitation. But I forsee a dead end for X79 like with X58 and I just don't like it. If IB won't be on X79 quickly enough and IB turns out to be quite a bit faster at a smaller process.. but that means waiting for IB. Bloody hellKristian Vättö - Saturday, May 7, 2011 - link
Don't expect LGA 2011 Ivy Bridges before H2 2012. SB-E is Q4'11 so Intel won't release IB-E in Q1'12.don_k - Saturday, May 7, 2011 - link
Yup, that's what I gathered. Frankly I don't see any point in LGA 2011 if, as I expect, IB is much faster than SB at the same clock speeds. You then get to wait for IB while the supposedly more 'mainstream' platform is faster than the high-end platform.Kristian Vättö - Saturday, May 7, 2011 - link
I would say the primary market for LGA 2011 parts is the enterprise market. The mainstream products (LGA 1155) lack many necessary features (PCIe lanes, ECC, DP setups...) that are required by servers and higher-end workstations. While there is Xeon E3-1200 lineup, it does not offer more than four cores and there is no option for DP setup either.For enterprises, the raw CPU performance is not always the number one concern. Fast I/O options are usually crucial and the 16 PCIe 2.0 lanes provided by mainstream SB may not be enough for those needs. In most cases, the performance per watt is also essential, especially in servers which are online 24/7. That is why Xeon lineups often offer low-power options.
From consumer's standpoint, SB-E may look like it makes no sense but in the end, it is not really aimed at them. There will be few CPUs for enthusiasts but other than that, it's solely Xeons.
Casper42 - Friday, May 6, 2011 - link
I really wanna slap someone over at Intel in their Marketing Dept.Why isnt this new chipset simply the X69?
did all the engineer's giggle too much or do they somehow think the SB-E NB/PCH is magically next generation because its taking so long to release.
If you're not including USB3 or LightPeak, it doesn't deserve to be a new generation.
Its bad enough there will be 4 different sockets for Servers next year, and I'm sure at least a dozen procs per socket (except maybe on the ultra high end, half dozen there more likely)
Kristian Vättö - Sunday, May 8, 2011 - link
According to VR-Zone, 20GB "Larsen Creek" SSD will be 119$. 40GB X25-M/310 series is cheaper than that!DanaG - Sunday, May 8, 2011 - link
Anyone know if the Ivy Bridge IGP is supposed to support 120Hz desktop? I don't expect true stereoscopic 3D on it, but it'd be nice for a 3D laptop to not switch from 120 to 60 when running on IGP.120Hz would also be helpful for playing 24FPS blu-ray.
dgingeri - Monday, May 9, 2011 - link
"A final interesting point for many users is that Ivy Bridge is pin compatible with Sandy Bridge, and it will work on current LGA1155 motherboards with the appropriate chipset and a firmware and BIOS update (H61, H67, P67, and Z68 are capable of support IB). Intel will also launch new 7-series chipsets, which we’ll get into below."Yeah, I believe it when I see it. They said the same thing with the 915/925 chipsets, but when it came down to it, they changed the VRM requirements and controls and a new motherboard was required for Core 2 chips anyway. I bet they do the same thing.
On the other side of the fence, I have a Nvidia Nforce 590SLi board (socket AM2) that supports even the newest 6 core chips with only a bios upgrade, if I want to replace it. I recently upgraded it from the Athlon 64 6400+ chip (3.2GHz with no overclocking potential and 125W power consumption) to a Athlon II X2 260 (3.2GHz, overclockable to 3.7GHz, and only 65W power usage) for $70. It doesn't sound like an upgrade, but the power savings alone will pay for it in less than 6 months, considering its my file server.
Intel does have a big history of making people upgrade an entire platform to upgrade. I don't expect them to change that.
Kakkoii - Thursday, June 9, 2011 - link
You guys are under NDA about Z68 until next week, yet there's already a bunch of information available on computer parts websites for Z68 motherboard pre-orders XDhttp://www.ncix.com/products/?sku=61757&vpn=GA...
Kinda funny..
Kakkoii - Thursday, June 9, 2011 - link
I probably should have looked at AnandTech's Motherboard section before posting this comment. lol.karl96 - Wednesday, March 28, 2012 - link
appart from what i have seen looks it nice but why does nt intel take out the integrate hd graphics in 3 type of processors and make them hexacore with hyperthreating. i mean sb is extremely powerfull and ivy bridge wil be more powerfull and more efficient. and for me it does not make that much sense such powerull procesoors and igpu?