• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Will Ivy Bridge Be Worth It? (Buyers Guide)

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Sentential

Contributing Member
Joined
Sep 16, 2003
Location
Knoxville, TN
Ive been mulling around if I should get a 2600K now or wait for Ivy Bridge. One thing I noticed is how similar this time feels to when Kentsfield was moved to Penryn; so I decided to use a little avaition fuzzy math to try and rationalize whether it was going to be worth buying or not. I think the data I produced is really quite interesting and I decided to share. What I used were a couple reviews but they werent as detailed as I wanted however; using Anandtech review numbers for Q6600 and Q9650 I was able to notice a lot of interesting info...

Clarkfield - 2.4 <> 3.5 (32% overclock of base freq)
Penryn - 3.0 <> 4.0 (25% overclock of base freq)

Process difference 3.5 <> 4.0 (13% increase of max oc)
= 13% increase for 30% shrink (going from 65nm to 45nm)

Sandybridge - 3.4 <> 4.8 (30% overclock of base freq)
Ivybridge - ? <> ?

That said these numbers look erriely similiar so one could intreperate these as 1:1 with a slight rounding error

If a die shrink from 65nm to 45nm = 30% = 13% max freq increase
Die shrink from 32nm to 22nm = 32% = 13/30 = .43 X 32 = 13.86% max freq increase

Expected Ivy Bridge per Anandtech = 4800 X .1386 = 665.28 + 4800 = 5468.28

Now to find base clock:

2400/3000 = 20% increase
3400 X .20 = 4080 (round off to 4000)

Expected E3700K = 4.0 base

Turbo?

2500k = 3.3 -> 3.7 = 11%
2600k = 3.4 -> 3.8 = 11%

So therefor = 4000 x .11 = 440 + 4000 = 4440 rounded = 4400

Expected E3700K = 4.0 Base 4.4ghz Turbo



----------


Extrapolating based on personal results

Clarkfield Q6600 - 2.4 <> 3.2 (25%)
Penryn Q9400 - 2.6 <> 3.5 (24%)

Average 24.5% vs 32% / 25%

Process difference based on personal results = 9% vs 13%

I tend to not have very good overclocking chips as I purchase them at release and heat has always been a factor


So lets average them together


------------


Expected Ivybridge clocks for 3700K based on realistic data (mine) with max overclock @ 4.5ghz

Low = 4905mhz @ 9%
Median = 4995mhz @ 11%
High = 5085mhz @ 13%


Expected Ivybridge clocks for 3700K based on optimistic data (Anandtech) with max overclock @ 4.8ghz

Low = 5232mhz @ 9%
Median = 5328mhz @ 11%
High = 5468mhz @ 13%

---------------------

TLDR:

Expected New Average Overclock for IB: 4885 + 5328 /2
= 5.106ghz
Expected Stock Clocks For i7 3700K
= 4.0 / 4.4

So is Ivy Bridge worth it? If you are at 5ghz no, 4.5? Most likely yes.
 
Last edited:
Honestly for use in a daily driver including gaming going from a SB rig that does 4.5 ghz to a IB 5 ghz (that is if the Anand projections are correct) will not yield any notible difference...in benching yes.
What I expect is that your rig will run cooler and draw less power with IB giving you higher performance clock for clock compared to SB.
 
Interesting post - the 4 GHz/4.4 GHz does sound about right I think, for what will be a pretty high end chip.
 
Honestly for use in a daily driver including gaming going from a SB rig that does 4.5 ghz to a IB 5 ghz (that is if the Anand projections are correct) will not yield any notible difference...in benching yes.
What I expect is that your rig will run cooler and draw less power with IB giving you higher performance clock for clock compared to SB.

My bad for not explaining; I wasnt using their projections I was projecting based on the review of the Q6600 compared to the Q9650 and extrapolating their values to sandy bridge.

Basically what I wanted to know what to get an idea of what the clocks would end up being. From what I found 4.9 is going to be the average give or take for what I normally get. Unfortunatly this CPU does not appear to be able to deliver 5.0 to the degree I hoped it would which may make people think twice about it. I for one will buy it at release especially given that if my numbers are right the base turbo for the 3700K will be identical to my current overclock.
 
My bad for not explaining; I wasnt using their projections I was projecting based on the review of the Q6600 compared to the Q9650 and extrapolating their values to sandy bridge.

Basically what I wanted to know what to get an idea of what the clocks would end up being. From what I found 4.9 is going to be the average give or take for what I normally get. Unfortunatly this CPU does not appear to be able to deliver 5.0 to the degree I hoped it would which may make people think twice about it. I for one will buy it at release especially given that if my numbers are right the base turbo for the 3700K will be identical to my current overclock.

Wow, awesome work Sentential! That's a really interesting way of projecting OCs and such. :thup: I will most likely stick with my 2500k for a good while after IB comes out, as I mostly game, and wouldn't see a huge difference in gaming with IB.
 
Back