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


View All Comments

  • rpsgc - Wednesday, March 7, 2012 - link

    All that revenue, all that profit and yet, they STILL can't bet AMD in integrated graphics.

    I think that qualifies as a fail.

    Thanks for (kind of) proving his point?
  • dagamer34 - Thursday, March 8, 2012 - link

    They don't really care to. The point of a business is to make money, not have the best products. The latter only gets solved when AMD gets serious in competing with Intel on power/performance again. Reply
  • Operandi - Tuesday, March 6, 2012 - link

    The internet called,"stop wasting my bits". Reply
  • StevoLincolnite - Tuesday, March 6, 2012 - link

    You know what? All you do is bash AMD.
    If you think AMD sucks THAT much and it's engineers and everything else is incredibly bad...
    Then I have a challenge.

    Go build your own Processor or GTFO with the bashing.
  • bennyg - Wednesday, March 7, 2012 - link

    Do not feed the troll. Reply
  • StevoLincolnite - Tuesday, March 6, 2012 - link

    Except... Intels IGP drivers on Windows are bad already. They are allot worst on the Mac.
    Historically Intel has never supported it's IGP's to *any* great length and even had to throw up a compatibility list for it's IGP's so you know what games they could potentially run.

    Here is a good example:

    Heck I recall it taking Intel a good 12 months just to enable TnL and Shader Model 3 on the x3100 chips.

    Historically the support has just not been there.
  • earthrace57 - Tuesday, March 6, 2012 - link

    AMD's CPUs are going to die...sucks to be an AMD fanboy. However, whatever they are doing with their dedicated GPUs, they are doing something right...if they can manage to pull their act together on the driver side, I think AMD would live as a GPU company... Reply
  • earthrace57 - Tuesday, March 6, 2012 - link

    I'm sorry, but Llano APUs will stay on top for quite a while; Intel is still at heart a CPU, Llano is part GPU...if AMD can get drivers the quality of nVidias, they will most likely do extremely well on that front. Reply
  • zshift - Tuesday, March 6, 2012 - link

    I really enjoyed the added compilation benchmark. This site has the most comprehensive collection of benchmarks that I've seen, it's a one-stop shop for most of my reviews. Keep up the great work! Reply
  • Jamahl - Tuesday, March 6, 2012 - link

    Would be great to see power benchmarks of the IGP, especially vs Llano and the HD 3000. Let's see if the graphics improvements have come at the price of yet more power consumption or if intel has managed to keep that down. Reply

Log in

Don't have an account? Sign up now