oh did intel do something wrong with the new ivy bridge

[steve210]

the sad news that i dont like to hear is that the ivy bridge run hotter than the sandy bridge cpus is it because the cpus are smaller in size or is that why they run hotter did intel really mess up by not test them right or did we see intel rush these cpus out

[kalniel]

Hi. No, they didn't do something wrong, or rush them. The chips are faster than the previous generation, use less power, and will cost less to make.

[shaithis]

Whilst their core temperatures can run higher, the amount of heat they generate is less.

It's odd and will certainly hamper overclocking....and perhaps long-term it may effect failure rates.

But, 4.5-5GHz on air (for the i5) in all the reviews I've read so far. Whilst it doesn't appear that it's going to challenge for any new overclocking records, they are still going to be the best chip on the market for a plethora of tasks.

[Biscuit]

what more was expected of IB like, looks like it pretty much lived up to expectations to me?

[wasabi]

Will be interesting to see power draw versus performance on the new i3 xxxxT chips in ivy bridge form. Possibly even as an underclocking experiment.

[CAT-THE-FIFTH]

For overclocking your better off getting a Core i5 2500K though as you can use a cheaper cooler,and it seems the Core i5 2500K overclocks slightly better overall negating the IPC increase with IB.

Power consumption only really seems to much lower under sustained load for IB. Looking at a few reviews,at idle and under lower loads power consumption does not seem that much different to Sandy Bridge. It might be enough for me to get IB for my system as I use a SFF system.

The IGP is the major improvement but mostly down to performance in Quick Sync and for HTPC use although the 23.976 FPS bug still remains.

IGP performance in games is better,but the HD4000 model will only be found in three desktop models - the Core i7 3770K,Core i5 3570K and the Core i3 3225. The latter won't be out for a few months,all the other models are stuck with the HD2500 which is only 10% to 20% faster than the HD2000. The HD4000 for desktop seems to be only around HD6530D level performance(or less) and this is with the top end IB Core i7 3770K which costs over three times more than a Llano A6-3650.

On the laptop front the Anandtech review did seem to indicate that the mobile HD4000 could beat the mobile HD6620G. However,the Llano laptop tested used slower 1333MHZ SODIMMs running asynchronously(one 2GB and one 4GB) whereas the IB laptop tested use a pair of 4GB 1600MHZ SODIMMs. It seems if the Llano laptop had been running a pair of 1600MHZ SODIMMs the score would have been 10% to 15% higher.

Moreover,the IB laptop tested used a 45W part when compared to the 35W part used in the Llano laptop. Due to the agressive Turbo Boost the HD4000 has it means it can boost higher more often and this was reflected in higher power consumption.

However,there were two big wins for the mobile Core i7 and those were in Skyrim and Batman - these are quite CPU limited games. AMD does run the risk that their mobile CPUs might end up bottlenecking their own IGPs. Hopefully Trinity will have better CPU performance.

[shaithis]

With the 24p bug still there, the chip is near-useless for HTPC users.

I wonder if they could have fit another core or 2 into the silicon without the IGP there.......

[Terbinator]

The 24p issue has/is being fixed?
http://www.anandtech.com/show/5773/i...-perspective/5

[shaithis]

So a fix is finally starting to surface - nice.

[shamus21]

Not much head room when you come to overclock it. Then it will come down to price. I can see some big price cuts coming on the SB`s to shift what left of stock over the coming months.

[Gideon]

Now from the reviews i've been looking at it seemed all the original reviews had them running considerably hotter than the SB which were over the last few months. The benchmarks and reviews i've noticed over the last day or two since they officially released it seem to pretty much be the same temps as the SB and clocked to 4.9 vs the 4.8 for the 2500K on standard cooling. Power consumption was considerably less as well, so it would appear as expected it's if you have a 2500K there's no point in spending out but if your starting a new rig like myself it's the better option.

The early reviews were based on engineering samples which i never consider to be the complete finished article.