Ivy Bridge models and clocks leaked

[tweakboy]

its 77 watts TDP for the normal Ivy i5/i7 quads and your Q6600 has a TDP of 105 watts.

Aida64 latest version tells me different 168watts max power.

 

[Idontcare]

Aida64 latest version tells me different 168watts max power.

Aida64 is telling you wrong.

Its also telling you your individual cores are clocked at 4248 MHz (check your screenshot).

Never trust software measurements for power-usage, not the consumer grade ones we all have easy access to at least.

 

[Idontcare]

When intel introduced the 3D gate there web site showed the next step in Intel evolution quantuam well. So I suspect will see that @ 14nm maybe 9 nm

3D xtors are not a 1-node solution.

Intel would not have spent 10yrs developing them to the point of being production worthy simply for the sake of the 22nm node.

In this industry everything is developed as a minimum 2-node solution. If it can't be scaled and used for at least 2 nodes then it simply won't be pursued in the first place.

3D xtors will be employed at Intel for a minimum of 22nm and 14nm. Minimum.

Then we have the 11nm and 8nm nodes. Personally I will be surprised if Intel does not retain 3D xtors and continue scaling them at least down to the 8nm node (making it a 4-node solution, if not longer).

 

[IntelUser2000]

60% better will play a lot of 'console'-class games at lower resolutions (think laptops here). Sure, not as good as Llano for GPU, but much better power consumption and superior CPU performance. Discrete mobile GPUs will start to be really only for high-end laptops...

At the worst it's going to be pretty damn close:
http://www.anandtech.com/show/4476/amd-a83850-review/5
http://www.guru3d.com/article/amd-a8-3850-apu-review/12

Same in mobile:
http://www.anandtech.com/show/4444/amd-llano-notebook-review-a-series-fusion-apu-a8-3500m/11

Here's the graphics performance slide:


On the Performance Preset, the HD 4000 in i7 3770 is 3x better than the i7 2600.

Note that both of the CPUs are non-K's. The regular i7 2600 has HD 2000, but with 1350MHz clock. Here, it's showing the HD 2000 in i7 2600 performs 20% better than the HD 2000 in i5 2500: http://www.behardware.com/articles/815-9/intel-core-i7-and-core-i5-lga-1155-sandy-bridge.html

The HD 3000 in the 2600K performs 35-40% better than HD 2000 in 2600.

In summary:

Crysis Warhead

HD 2000 in regular Sandy Bridge desktop SKUs = 1x
HD 2000 in i7 2600 = 1.2x
HD 3000 in i5 2500K = 1.4-1.5x
HD 3000 in i7 2600K = 1.35-1.4 x (i7 2600's HD 2000) = 1.6-1.7x

-HD 4000 in i7 3770 in 3DMark Vantage Performance config = 2.9-3 x (i7 2600's HD 2000)

While the top end performance gain(3770 vs 2600K) is only 60%, average gain will be greater because 2500K is the more popular chip for HD 3000 usage. In that case it will end up to be 1.9-2x.

 

[PreferLinux]

Sandy Bridge's tying of everything to the bclk allows maybe 5% at best increase for overclocking beyond multiplier limitations. If rumors are true, Ivy Bridge is supposed to have something similar to Sandy Bridge E, meaning it might offer you coarse steps of bclk you can choose from, like back in the Pentium MMX days. Say 100/125/150MHz. Maybe + or - 5% from that.

And remember you can drop the multiplier as well, even on locked chips.

 

[toyota]

Aida64 latest version tells me different 168watts max power.
http://www.hammiestudios.com/168w.jpg

once again you make me wonder how you had the ability to even build that pc. the official TDP for your cpu is 105 watts. that takes all of 5 seconds to find using google. heck right there in your own screenshot is a link to the official info for your cpu. http://ark.intel.com/products/29765/Intel-Core2-Quad-Processor-Q6600-(8M-Cache-2_40-GHz-1066-MHz-FSB)

 

[piesquared]

At the worst it's going to be pretty damn close:
http://www.anandtech.com/show/4476/amd-a83850-review/5
http://www.guru3d.com/article/amd-a8-3850-apu-review/12

Same in mobile:
http://www.anandtech.com/show/4444/amd-llano-notebook-review-a-series-fusion-apu-a8-3500m/11

Here's the graphics performance slide:


On the Performance Preset, the HD 4000 in i7 3770 is 3x better than the i7 2600.

Note that both of the CPUs are non-K's. The regular i7 2600 has HD 2000, but with 1350MHz clock. Here, it's showing the HD 2000 in i7 2600 performs 20% better than the HD 2000 in i5 2500: http://www.behardware.com/articles/815-9/intel-core-i7-and-core-i5-lga-1155-sandy-bridge.html

The HD 3000 in the 2600K performs 35-40% better than HD 2000 in 2600.

In summary:

Crysis Warhead

HD 2000 in regular Sandy Bridge desktop SKUs = 1x
HD 2000 in i7 2600 = 1.2x
HD 3000 in i5 2500K = 1.4-1.5x
HD 3000 in i7 2600K = 1.35-1.4 x (i7 2600's HD 2000) = 1.6-1.7x

-HD 4000 in i7 3770 in 3DMark Vantage Performance config = 2.9-3 x (i7 2600's HD 2000)

While the top end performance gain(3770 vs 2600K) is only 60%, average gain will be greater because 2500K is the more popular chip for HD 3000 usage. In that case it will end up to be 1.9-2x.

Unfortunately, one cannot talk about graphics performance without talking about image quality. Assuming IntEl can get their drivers working, and provide image quality on the level of AMD and NV, given their history in creating graphics solutions, it's highly unlikely they'll be anywhere near competitive in an apples to apples comparison.

 

[RussianSensation]

Intel Reveals Official Next-Gen "Ivy Bridge" Chip Performance Numbers to Partners.

According to the documents, Intel Core i7-3770 (4 cores with HyperThreading, 3.40GHz, 8MB cache) will deliver the following advantages compared its predecessor Core i7-2600 (4 cores with HyperThreading, 3.40GHz, 8MB cache):

+7% higher overall SYSmark 2012 score;
+14% higher overall HDXPRT 2011 score;
+15% higher Cinebench 11.5 score;
+13% better ProShow Gold 4.5 result;
+25% faster performance in Excel 2010;
+56% faster performance in ArcSoft Media Expresso;
+192% higher overall 3DMark Vantage score;
+17% faster performance in 3DMark Vantage CPU benchmark;
+199% faster performance in 3DMark Vantage GPU benchmark;

Ivy Bridge will generally inherit Sandy Bridge micro-architecture and will sport a rather significant number of improvements.

1) Firstly, it will have certain improvements that will boost its performance in general applications by around 20% compared to Core i "Sandy Bridge" chips (e.g., enhanced AVX acceleration).

2) Secondly, the forthcoming chip will have a new graphics core with DirectX 11 and OpenCL 1.1 support, 30% higher performance compared to the predecessor as well as new video processor and display controllers.

3) Thirdly, Ivy Bridge will feature PCI Express 3.0 x16 interconnection as well as PCIe 2.0 x4 controller.

4) In fourth, the processor will support a number of power management innovations.

The most interesting part for me is overclocking!!

 

[toyota]

maybe this was already brought up but something on that chart in the OP is odd. its has the 3770k with 3.5 base and 3.9 turbo but the 3770 with 3.4 base and 3.9 turbo. that makes no sense to have the k version of the same chip with .1 higher base clock. the 3750k and 3750 both have the same base on oc clocks so it makes me think the 3770 is really supposed to be 3.5 for a base.

 

[Nemesis 1]

3D xtors are not a 1-node solution.

Intel would not have spent 10yrs developing them to the point of being production worthy simply for the sake of the 22nm node.

In this industry everything is developed as a minimum 2-node solution. If it can't be scaled and used for at least 2 nodes then it simply won't be pursued in the first place.

3D xtors will be employed at Intel for a minimum of 22nm and 14nm. Minimum.

Then we have the 11nm and 8nm nodes. Personally I will be surprised if Intel does not retain 3D xtors and continue scaling them at least down to the 8nm node (making it a 4-node solution, if not longer).

If you say so it must be true.

Intel's partnership with UK R&D company Qinetiq has borne further fruit: the pair this week said they had made a quantum-well transistor with a gate length of 85nm.

Quantum-well transistors are also known as high electron-mobility transistors, and they excite companies like Intel because of their ability to operate simultaneously at low voltages and very high clock speeds. They also dissipate far less heat than today's transistors.

The Intel/Qintiq quantum-well transistor is based on a material called indium antimonide (InSb), which Qinetiq has been researching for some time through is R&D partnership with the UK's Ministry of Defence. According to Qinetiq, transistors made from InSb would consume a tenth of the energy gobbled up by today's state-of-the-art transistors yet deliver the same performance. Or they could be used to triple performance for the same power consumption.

An 85nm gate-length may not sound impressive at a time when Intel and others are about to debut 65nm semiconductors before moving to 45nm in 2008, but it's a big improvement on the 200nm quantum-well transistor Intel and Qinetiq made last year. These are early days for the technique - Intel isn't expecting to be able to make use of it until around 2015.

The newest test transistor operates in 'enhancement mode', rather than the 'depletion mode' of the 200nm version produced last year. That means it more closely mimics the operational behaviour of the transistors found in modern CPUs.

 

[Nemesis 1]

Unfortunately, one cannot talk about graphics performance without talking about image quality. Assuming IntEl can get their drivers working, and provide image quality on the level of AMD and NV, given their history in creating graphics solutions, it's highly unlikely they'll be anywhere near competitive in an apples to apples comparison.

Man I admit you fight the good fight. But your on the loosing side. Just like your were with the BD hype . But this battle is a foolish fight , Doesn't matter to me if the 4000 is 3x better than last generation . Its still not an enthusiast gamer IGP . It will however give llano all it wants . On the graphics thing Please tell me whats wrong with what I see on my screen when using intel graphics . This driver thing is really old . It was a battle fought between ATI/NV and now you guys are taking it to the next level . AMD/NV fanbois ganging up on intel graphics . Just go by an AMD BD and live in peace with it . You will lose this battle and the war.

 

[grimpr]

Gotta give credit to Nemesis 1, he's a hellhound on tracking up very interesting information and a man with long memory, which means a good history afficionado. :thumbsup:

 

[Nemesis 1]

Intel Reveals Official Next-Gen "Ivy Bridge" Chip Performance Numbers to Partners.

According to the documents, Intel Core i7-3770 (4 cores with HyperThreading, 3.40GHz, 8MB cache) will deliver the following advantages compared its predecessor Core i7-2600 (4 cores with HyperThreading, 3.40GHz, 8MB cache):

+7% higher overall SYSmark 2012 score;
+14% higher overall HDXPRT 2011 score;
+15% higher Cinebench 11.5 score;
+13% better ProShow Gold 4.5 result;
+25% faster performance in Excel 2010;
+56% faster performance in ArcSoft Media Expresso;
+192% higher overall 3DMark Vantage score;
+17% faster performance in 3DMark Vantage CPU benchmark;
+199% faster performance in 3DMark Vantage GPU benchmark;

Ivy Bridge will generally inherit Sandy Bridge micro-architecture and will sport a rather significant number of improvements.

1) Firstly, it will have certain improvements that will boost its performance in general applications by around 20% compared to Core i "Sandy Bridge" chips (e.g., enhanced AVX acceleration).

2) Secondly, the forthcoming chip will have a new graphics core with DirectX 11 and OpenCL 1.1 support, 30% higher performance compared to the predecessor as well as new video processor and display controllers.

3) Thirdly, Ivy Bridge will feature PCI Express 3.0 x16 interconnection as well as PCIe 2.0 x4 controller.

4) In fourth, the processor will support a number of power management innovations.

The most interesting part for me is overclocking!!

So your saying the 20% performance increase I gave intel on SB is a fact . So why all the hate when I said that . Why can't you people except things for what they are? Its easy to make % what you want them to be . Include worse case exclude best case . I perferr excluding worse case and best case . Denile is a river sometimes a river of blood.

 

[Nemesis 1]

Gotta give credit to Nemesis 1, he's a hellhound on tracking up very interesting information and a man with long memory, which means a good history afficionado. :thumbsup:

You think thats a good thing? I wish I had really bad memory. You haven't a clue what its like to recall every rotten thing you have done in life its horriable to carry that weight .
Than being obsessed with history and religion and finding the truth and everyone including your own family thinking your insane . Than to have a near death like experiance without being near death . If I had a choice to have been born . I would have choosen not. To have the Love I have for my grandchildren and have my beliefs. Its not good its a freaken nightmare.

 

[podspi]

If you say so it must be true.

Intel's partnership with UK R&D company Qinetiq has borne further fruit: the pair this week said they had made a quantum-well transistor with a gate length of 85nm.

Quantum-well transistors are also known as high electron-mobility transistors, and they excite companies like Intel because of their ability to operate simultaneously at low voltages and very high clock speeds. They also dissipate far less heat than today's transistors.

The Intel/Qintiq quantum-well transistor is based on a material called indium antimonide (InSb), which Qinetiq has been researching for some time through is R&D partnership with the UK's Ministry of Defence. According to Qinetiq, transistors made from InSb would consume a tenth of the energy gobbled up by today's state-of-the-art transistors yet deliver the same performance. Or they could be used to triple performance for the same power consumption.

An 85nm gate-length may not sound impressive at a time when Intel and others are about to debut 65nm semiconductors before moving to 45nm in 2008, but it's a big improvement on the 200nm quantum-well transistor Intel and Qinetiq made last year. These are early days for the technique - Intel isn't expecting to be able to make use of it until around 2015.

The newest test transistor operates in 'enhancement mode', rather than the 'depletion mode' of the 200nm version produced last year. That means it more closely mimics the operational behaviour of the transistors found in modern CPUs.

Is this a quote from somewhere? Because 65nm has been out for some time now :biggrin:

It is going to be really interesting finally getting some of these 22nm CPUs in hand...

 

[exar333]

Intel Reveals Official Next-Gen "Ivy Bridge" Chip Performance Numbers to Partners.

According to the documents, Intel Core i7-3770 (4 cores with HyperThreading, 3.40GHz, 8MB cache) will deliver the following advantages compared its predecessor Core i7-2600 (4 cores with HyperThreading, 3.40GHz, 8MB cache):

+7% higher overall SYSmark 2012 score;
+14% higher overall HDXPRT 2011 score;
+15% higher Cinebench 11.5 score;
+13% better ProShow Gold 4.5 result;
+25% faster performance in Excel 2010;
+56% faster performance in ArcSoft Media Expresso;
+192% higher overall 3DMark Vantage score;
+17% faster performance in 3DMark Vantage CPU benchmark;
+199% faster performance in 3DMark Vantage GPU benchmark;

Ivy Bridge will generally inherit Sandy Bridge micro-architecture and will sport a rather significant number of improvements.

1) Firstly, it will have certain improvements that will boost its performance in general applications by around 20% compared to Core i "Sandy Bridge" chips (e.g., enhanced AVX acceleration).

2) Secondly, the forthcoming chip will have a new graphics core with DirectX 11 and OpenCL 1.1 support, 30% higher performance compared to the predecessor as well as new video processor and display controllers.

3) Thirdly, Ivy Bridge will feature PCI Express 3.0 x16 interconnection as well as PCIe 2.0 x4 controller.

4) In fourth, the processor will support a number of power management innovations.

The most interesting part for me is overclocking!!

Impressive if true. Thats some amazing performance gains for a 'die-shrink' plus using considerably less power.

This REALLY makes me want a quad-core laptop. And amazingly enough, it would be plenty capable without a discrete GPU. This should really sell the 'Ultrabook' idea very well.

 

[Nemesis 1]

Is this a quote from somewhere? Because 65nm has been out for some time now :biggrin:

It is going to be really interesting finally getting some of these 22nm CPUs in hand...

Yes it is a quote from somewhere but like you I don't know where I got it from . I pulled it from our research machine.

 

[bryanW1995]

Sandy Bridge's tying of everything to the bclk allows maybe 5% at best increase for overclocking beyond multiplier limitations. If rumors are true, Ivy Bridge is supposed to have something similar to Sandy Bridge E, meaning it might offer you coarse steps of bclk you can choose from, like back in the Pentium MMX days. Say 100/125/150MHz. Maybe + or - 5% from that.

That would be awesome, but with no competition I'm not going to hold my breath.

 

[RussianSensation]

So your saying the 20% performance increase I gave intel on SB is a fact . So why all the hate when I said that . Why can't you people except things for what they are?

No, the 4 points listed out by me are directly from the article, not written by me.

 

[Idontcare]

Yes it is a quote from somewhere but like you I don't know where I got it from . I pulled it from our research machine.

My research machine is indicating to me that Google says you most likely got it from TheRegister

 

[dma0991]

Intel Delays 22nm Ivy Bridge Processor to April/May 2012

 

[exar333]

The biggest question on most minds here is 'How high will these IB CPUs clock'.

Can't wait for some initial leaks...

 

[MrTransistorm]

The second question would be, "How durable are they?"

The 32nm process has been around long enough that we know what range of voltages are safe before degradation sets in. What about 22nm? I'm definitely going to wait and see how well they handle OC'ing long term before I consider any upgrades.

 

[BadThad]

I agree with OneOfTheseDays. I am currently using an OCd Q6600. When I go to replace it with an IB model, I doubt I will overclock it at all. Time and stability are too important to me as I use it for a work computer. Funny thing is that I don't really need to upgrade as the Q6600 (at 3.34 GHZ) is enough processor for me...thanks to the Intel SSD

ME TOO! The Q6600 meets my needs and my old Asus mobo refuses to die. 4GB ram and an SSD keeps me happy on this Q6600.