The mountain that AMD have to climb to compete at the top end of the CPU performance scale has just got a bit steeper, and a little higher as well.

Trinity has got to be good and the desktop "Piledriver" has got to be very good and very cheap to compete with these latest CPU's from Intel.

As far as the CPU usage and power draw is concerned with certain types of media/players. This is either a cock-up with the new Intel GPU, or its a software issue..... no doubt we will find this out soon.


Andy

PS: Good review.

Higher idle power consumption with the integrated graphics than sandy bridge. Other reviews would have missed that.

Meanwhile Intel would have you believe idle power is going down thanks to memory controller power gating. Maybe laptop CPUs will not measure so badly relative to sandy bridge mobiles.

Unless I've missed something, SPCR compared the idle power consumptions of the boards. And mobile boards are obviously better.
As you can see in the reviews, there is a large difference between boards. Unless Intel somehow botched the new GPUs (or there's a driver issue), none of these CPUs consumes much at idle.

Since SB and IB can work in the same motherboard, I'd be interested to see if the power consumption differences are still there on the same board? Maybe something with the Intel Z77 board is causing the higher power when the IGP is in use and you would see the same with a SB CPU. And maybe some of the more mainstream boards might give some better numbers.

I bet the Intel is pretty efficient. But the more features and I/O you've got, the higher the power consumption. So this Z77 board can't be expected to have an H61-like power consumption. Maybe all the new boards are more powerful than the old H61s. In that case, none are likely be as efficient as the best ones from last year.

Thanks for the review!

HT4U (in Googlish) shows idle power with iGPU similar between IVB and SNB.

So, could be drivers, could be mobo differences..

Competition is good I'm FX ready so let's hope piledriver sorts things out for AMD in both performance and power consumption and yes price too.
Regardless, I continue to support AMD as I have done for many years now.

I simply can't build an Intel PC because we need AMD to be there.
Overall despite lagging Intel I've not been unhappy with my build, not have the folks I've built pc's for been anything other than very pleased.
I want to see them get better though, no question on that.

"The die is arranged the same way but the manufacturing process has been shrunk from 32nm to 22nm utilizing Intel's much lauded Tri-gate transistor technology (it's explained nicely by AnandTech), the integrated graphics chips has been beefed up with more horsepower, the memory controller has been updated to officially support DDR3-1333..."

Is that supposed to be DDR3-1600?

I'm looking at some of Asus' new features in the upcoming motherboards, I'm really interested in the advanced fan control options, maybe SPCR can cover that extensively . I believe it effectively determines minimum and maximum fan speeds for each individual fan connected to every header and it uses that to actually determine it's "Silent" preset, etc. I believe you can also set more advance parameters. This sounds great to me.

While maybe not "extensive", SPCR always makes note of a motherboards fan control or lack there of. They've touched on the ASUS fan control in articles like Sandy Bridge, Part 2: Intel DH67BL & Asus P8H67-M EVO H67 Motherboards. I know I've seen people at the Tech Report pushing for more fan control options so there might be some articles there more specific about ASUS's implementation which appears to be the best.

hopefully the higher temp won't be too problematic

A very interesting discussion regarding why IB has such high temps (apparently, Intel did not solder the IB engineering samples to the heatspreader, but rather, used TIM. It is unclear if the retail IB cpus will be soldered like SB cpus):

http://www.overclockers.com/ivy-bridge-temperatures

IMO Intel limited the overclocking potential of the Ivy Bridge CPUs on purpose.
They not only used TIM instead of solder, but they used a low quality TIM. Check this article:
http://vr-zone.com/articles/ivy-bridge- ... 15844.html
After reading this I am considering delaying my i7-3770k purchase, maybe Intel will admit the 'mistake'
and use a better TIM in the next stepping of Ivy. But a lot of people should protest to make this happen.

I sent a query to my Intel contacts and got this reply:

Meh... so for now Intel sticks to 'it meets quality and reliability expectations'.
I understand that they didn't want to use solder, for some technical reasons.
But when not-overclocked load temperatures could be 8-11 degrees lower just by
using a quality TIM inside the package, which would cost Intel less than 1$ more per CPU,
I find their attitude annoying. I am sure Ivy will make a lot of profit for Intel anyway.

Mike - that seems to be the stock Intel PR response I've seen on other sites.



In bulk, a decent TIM would only add pennies more. So, yeah, Intel screwed the pooch on this. Saved a few pennies on their end, causing all of their customers to spend a lot more on the cpu cooling solution for decent temps - even at stock. If they use this same, rather crappy, solution on their laptop CPUs...there will be some noisy laptops this summer.

After surveying forum posts on Ivy Bridge overheating on various forums, it seems that even people using top heatsinks
and fast spinning ( read: loud ) fans are only getting marginal temperature improvements. The only major improvement is by IHS TIM replacement.
Since the IHS TIM replacement procedure is risky and I couldn't find a detailed description of it, for now I didn't buy an i7-3770k.
Just a small mistake can ruin a 350 euros processor, and I would feel depressed if this happenend to me.

Intel just launched the many-core Xeon Phi ( AKA Knights Corner, derived from Larabee ), built on the same 22nm process as Ivy Bridge.
I wonder if they used the same crappy TIM on that, or the die is naked and the heatsink makes direct contact.