Mobile Ivy Bridge Specs Leaked

Eh, I see a 200 MHz base clock bump and a 300 MHz turbo bump, best-case for the same 17w TDP. That's a %10 clock speed increase across the board. You probably won't notice the difference in a blind test.

This is NOT what I was expecting from Intel's "amazing new 22nm process that is supposed to kill leakage and bring world peace," especially when you get down to the prized 17w chips that are cherry-picked. The May availability for for those babies is also troubling for Intel's new fleet of Ultraboooks.

The rumored %50-60 performance increase for the GPU does NOT account for all the power savings we were supposed to see. Unlike Llano, Intel uses significantly less than half the die area for the GPU, so even a %50 performance increase in the GPU area would not constitute more than a %20 overall power consumption increase (if you did this on 32nm). A good die shrink should reduce power consumption more than that.

We will have to wait and see if the claims by Intel of up to a %15 performance boost clock-for-clock are warranted, but until that is confirmed Ivy is looking pretty meh.
 
Last edited:
defaultluser said:
Eh, I see a 200 MHz base clock bump and a 300 MHz turbo bump, best-case for the same 17w TDP. That's a %10 clock speed increase across the board. You probably won't notice the difference in a blind test.

This is NOT what I was expecting from Intel's "amazing new 22nm process that is supposed to kill leakage and bring world peace," especially when you get down to the prized 17w chips that are cherry-picked. The May availability for for those babies is also troubling for Intel's new fleet of Ultraboooks.

The rumored %50-60 performance increase for the GPU does NOT account for all the power savings we were supposed to see. Unlike Llano, Intel uses significantly less than half the die area for the GPU, so even a %50 performance increase in the GPU area would not constitute more than a %20 overall power consumption increase (if you did this on 32nm). A good die shrink should reduce power consumption more than that.

We will have to wait and see if the claims by Intel of up to a %15 performance boost clock-for-clock are warranted, but until that is confirmed Ivy is looking pretty meh.
Click to expand...

It's not power consumption at 100% utilization and voltage that should be focused on, especially when you're looking at a mobile product. Same as the reason you don't measure mileage at a car's maximum speed; because it's important to have a real world measurement.

edit: from my understanding Ivy Bridge power savings are more profound at lower clocks. 25-35% less voltage (or so) is required for the base clock compared to 10-15% less at the maximum clock speed. I forget the formula to figure out power consumption at a certain frequency and voltage when given the standard power consumption, frequency, and voltage, but I can tell you that the relationship between voltage and power consumption (when the frequency stays the same) is definitely not linear. Since (in my experience and most others around me) laptop CPUs spend most of the time almost idle, I think it's safe to say mobile Ivy Bridge will be a big step forward for Intel regarding average power consumption making battery life very noticeably longer
 
Last edited:
mike_6289 said:
It's not power consumption at 100% utilization and voltage that should be focused on, especially when you're looking at a mobile product. Same as the reason you don't measure mileage at a car's maximum speed; because it's important to have a real world measurement.

edit: from my understanding Ivy Bridge power savings are more profound at lower clocks. 25-35% less voltage (or so) is required for the base clock compared to 10-15% less at the maximum clock speed. I forget the formula to figure out power consumption at a certain frequency and voltage when given the standard power consumption, frequency, and voltage, but I can tell you that the relationship between voltage and power consumption (when the frequency stays the same) is definitely not linear. Since (in my experience and most others around me) laptop CPUs spend most of the time almost idle, I think it's safe to say mobile Ivy Bridge will be a big step forward for Intel regarding average power consumption making battery life very noticeably longer
Click to expand...

Good point. Idle power is not tied to TDP.

Intel could surprise the entire laptop world if Ivy drops processor idle power by half or more.
 
Earlier this year Anandtech had stated we'll see a 35W quad core part - http://www.anandtech.com/show/4773/ivy-bridge-will-bring-a-35w-quadcore-i7

I don't see it on the leaked slides. There's no replacement for the 26xxQM on those slides, so maybe it's just not on those slides.

The iGPU gains are nice, but I suspect the only aspect that will matter for most people is the 4K video support. I guess a handful of gamers might find the improved iGPU very useful.

The improved power usage will be nice, but I want to see some real world results before jumping up and down. :) 30% reduction at idle is just theory at this point.

The release dates are stabbing my heart - April and May! I've been itching to buy a nice laptop, and I figured we'd see IB laptops in March. If the processors are out in April, we probably won't see a mass release of laptops till a May end or even June/July. I guess any fear Intel had from AMD is gone.
 
You must log in or register to reply here.
Back
Top