Buy the Cybertron BX80623I72600K 16GB DDR3 2x
TigerDirect
$2,474.50
Rakuten
$327.99
A Matter of Fax
$369.69

Ivy Bridge Architecture Recap

At IDF Intel disclosed much of Ivy's CPU architecture, but below is a quick summary:

- 4-wide front end with µOp cache from Sandy Bridge
- OoO execution engine from Sandy Bridge
- Data structures previously statically shared between threads can now be dynamically shared (e.g. DSB queue), improves single threaded performance
- FP/integer divider delivers 2x throughput compared to Sandy Bridge
- MOV instructions no longer occupy an execution port, potential for improved ILP when MOVs are present
- Power gated DDR3 interface
- DDR3L support
- Max supported DDR3 frequency is now 2800MHz (up from 2133MHz), memory speed can be moved in 200MHz increments
- Lower system agent voltage options, lower voltages at intermediate turbo frequencies, power aware interrupt routing
- Power efficiency improvements related to 22nm
- Configurable TDP

I've highlighted the three big items from a CPU performance standpoint. Much of the gains you'll see will come from those areas coupled with more aggressive turbo frequencies.

On the GPU, the improvements are more significant. Some of the major changes are below:

- DirectX 11 Support
- More execution units (16 vs 12) for GT2 graphics (Intel HD 4000)
- 2x MADs per clock
- EUs can now co-issue more operations
- GPU specific on-die L3 cache
- Faster QuickSync performance
- Lower power consumption due to 22nm

Introduction The Lineup
POST A COMMENT

198 Comments

View All Comments

  • bigboxes - Thursday, March 08, 2012 - link

    Look at the chart. Nehalem (i920) was a tock. Reply
  • just4U - Thursday, March 08, 2012 - link

    I have to disagree with you on the i920 being such a huge leap. As someone who goes thru virtually every cpu line up for AMD/Intel I'd have to say the C2D (or quad) 6x series was the biggest leap forward in the past decade. Before that it was the A64 and X2 variants (altho.. we didn't get alot of use out of those secondary cores) Reply
  • IntelUser2000 - Thursday, March 08, 2012 - link

    LOL, this must be the most hilarious argument I've heard in a while.

    How do you relate 30+ % graphics gain as being ALL CPU? Don't be ridiculous, and that's an understatement.
    Reply
  • Silma - Thursday, March 08, 2012 - link

    Are low-res testing really relevant for graphics?

    Most players play at 1920x1080 or higher.
    1368x720 or 1680x1050 does not seem relevant to me at all for most people, especially those purchasing a computer with this processor.
    Reply
  • dagamer34 - Thursday, March 08, 2012 - link

    Most players who game at 1920x1080 also have graphics cards that cost more than $100. That's not what this was testing. Reply
  • geekfool - Thursday, March 08, 2012 - link

    I intend to get a 1900x1200 monitor to replace a 1024x768 but don't generally game with anything more sophisticated than OpenTTD...I figure a 7750 Radeon will be enough of a graphics card. Reply
  • Yojimbo - Thursday, March 08, 2012 - link

    uhh.. i don't think that's true. graphics intensive applications are not the only ones that benefit from fast CPUs. Reply
  • krumme - Thursday, March 08, 2012 - link

    Where does Charlie claim to be biased?

    But i agree, this sacred, aura, "this is not sanctioned by Intel" is a pain to read. It makes thesse articles a little bit difficult to start reading :)

    But how profitable, and how good a business do you have if you dont have "good conections"? - charlie uses his for underhand information, anand his to get info before the others. Its very obvious for us to interprete Anands article because we know the obvious, - it have to be profitable for both anand and Intel. But what about Charlie, what is the motives for the people leaking info to him? - its not quite so obvious and transparrent.
    Reply
  • awg0681 - Thursday, March 08, 2012 - link

    "Sure, he was comparing Intel graphics to Intel graphics, except he wasn't, because he himself threw Llano in there to compare."

    By the same token, if he had not included Llano results people would be wondering where they were and complaining that they weren't included. Puts Anand in a catch 22 when deciding whether or not to include Llano.

    There is validity to the complaint about the numbers being incorrect. Those should be looked at and corrected. Glossing over the results and no mention of Llano being more capable, again, this was mainly to compare Intel v Intel in a preview of their new chip and improvements they've made since last gen. Sure, he could've been more thorough with the AMD v Intel side, but that's not really what this article was about. We could also go to a steakhouse and complain there's not a large vegetarian meal selection too.
    Reply
  • Azeraph - Thursday, March 08, 2012 - link

    it doesn't really matter if the igp isn't that great most people don't buy them for their graphics power.I get the feeling that maybe intel is just putting them out there to keep it's base solid against AMD,Not that it needs it and i'm an amd fan. i found something the other day that will possibly change how tomorrows processors will use light instead of electricity.

    http://scitechdaily.com/penn-researchers-build-a-c...
    Reply

Log in

Don't have an account? Sign up now