I just came from market. The h55tc intel mobo that comes with i3 530 costs about the same as the DP55WB intel they paired with i5 750. Around $100 each. Not saying I am gonna buy the mobo they paired, but seeing that I thought the prices of mobo for i3 and i5 are around same?
Depends on how much you shop around. There are H55 mobos around for as little as $55 online, but some of those are refurbs. The cheapest non-refurb I can find is at Amazon for $78 + free shipping:
http://www.amazon.com/Biostar-Intel-.../dp/B003A4H4P2
$100 for a P55 board is not a bad deal, and if it does what you need it to do, then you aren't getting much from going H55.
Isn't there any i5 based on 32nm? I thought the new i3 , i5 and i7 processors were all based on 32nm. T
This is where Intel's processor naming scheme gets confusing. There are 32nm i5s, but they are dual-core Clarksdales that cost too much for what they do. The quad-core i5-750 is a 45nm Lynnfield. But you seem to have figured that part out.
The 860 and 920 are costing me exactly same. Although 860 is not readily available. What about 860 mobos?
The i7-860 is an LGA1156 processor (Lynnfield) that would use the same motherboard as an i5-750 or even an i3-530. You can think of it as a 750 + Hyperthreading.
the core performs calculations in steps, called a "pipline". The more steps you have, the longer your pipeline. HT moves two instructions along these at the same time and can result in some hardware that is not being utilized due to the nature of the current instructions to get utilized more and get better performance.
results? depending on the exact mathematical formula being computed, enabling HT can result from anywhere between a 5% DECREASE in speed to an 80% INCREASE in speed. (with a small increase in power consumption)... resulting in either a decrease or an increase in power efficiency depending on the type of calculations you are performing.
If I remember correctly, HT gives a significant speed and power efficiency boost to video editing. but you will have to look into detailed reviews that test the increase in performance with it on and off for various known tasks. (or you can perform said tests yourself, since you can turn off HT on any HT enabled processor).
Yes, that is a good description of what HT does. I was simply alluding to the fact that a HT-created logical processor handling a thread on a Clarksdale would probably use less power doing so than an HT-created logical processor handling the same thread on a Bloomfield.
http://techreport.com/articles.x/18448/4
How does dual core i3 530 consumes more power then quad core i5 570 at idle? Granted its clocked a bit higher but still. Shouldn't 2 more cores require alot more power then few mhz? Does it has something to do with motherboard?
If the test was fair, techreport should have been running the i5-750 on the same board as the i3-530. That being said, it could have something to do with how aggressive power-saving features are on the i5-750 vs. the i3-530. The 750 may also benefit more from power-gating but I'm not sure about that. Considering the difference is 4W, it doesn't seem like a big deal.
Considering this i3 is out of the picture.
920 is also out of the picture. Both idle and peak are above my liking.
Yeah, I figured you'd want a quad. Hell initially I figured you'd need a quad with HT, but now that you have described your usage patterns a bit more, maybe not so much. In that case an undervolted i5-750 should fit the bill. I only threw Clarksdale out there because they can be really efficient chips (just a bit underpowered).
I am almost considering i5 750. Great idle consumption with digestable peak. Although a little lower would have been great. I wonder if their peak wattage was measured with turbo boost running?
Probably, though you can always disable turbo. And undervolt.
870 idle looks great too. Peak is just a little above my liking. I can't find 870 in my local market. Have they just been released?
The i5-750, i7-860, and i7-870 were all released at about the same time.
There is no mention of 860 in the charts. Wonder why?
For whatever reason, some reviewers did not review this chip in their first batch of Lynnfields, and still others never reviewed it at all. Anandtech had to do an i7-860 followup article to supplement their initial Lynnfield release article:
http://www.anandtech.com/show/2839