Going from Kentsfield to Haswell, can you see the difference?

Subject: Systems | August 9, 2013 - 05:17 PM |
Tagged: q6600, i7-4770k, upgrade

Haswell has not been recieved with screams of joy from enthusiasts as we appreciate the power savings but are honestly more interested in MOAR POWER!  As an upgrade from Ivy Bridge there are only a few benefits but what about updating from an old Q6600 on an X38 motherboard?  That is what Legit Reviews just did, moving to a Gigabyte Z87X-UD4H and i7-4770K and as an upgrade it is more than impressive; everything from gaming to the productivity of the system skyrocketed as you can see below.  If you are still running a Q6600 on your main rig it looks like saving your pennies would be a very good idea.

View Full Size

more impressive benchmarks if you follow the link

"We all love doing computer upgrades, but sometimes we have to ask ourselves if a full system upgrade is worth it or necessary. There are times that you may only gain a small percentage of a performance boost, where there are other times that you gain significant increases. Today I am going to compare my original Intel Kentsfield CPU and X38 motherboard to a shiny new Haswell CPU and Z87 motherboard."

Here are some more Systems articles from around the web:


August 9, 2013 | 05:25 PM - Posted by stompy (not verified)

From what I understand here in terms of CPU
One core processing very disappointing.

What has happened in recent years only reduced the nanometer and quite a bit performance ..
Added processors ... And have improved very little each core

August 9, 2013 | 05:48 PM - Posted by John_Gr (not verified)

I think going from 6600 to an FX 8350 will be also a big leap ahead with half the cost. OK we are not talking about AMD here, but I just think that anyone considering an upgrade from a 5 years old system is lucky enough to have both companies as an option. On the other hand someone with an Sandy/Ivy Bridge looks only what is the latest option from Intel.

August 9, 2013 | 07:01 PM - Posted by Jeremy Hellstrom

That would be another good comparison to do

August 9, 2013 | 10:25 PM - Posted by SetiroN

Those are two overclocking oriented builds, he should have tested overclocked speeds.
I'm pretty sure my 4GHz Q6600 wouldn't be humiliated like that.

August 10, 2013 | 02:35 AM - Posted by HeavyG (not verified)

Compared to an overclocked Haswell... I am pretty sure it would be humiliated just as much. Considering you are likely reaching 4.0GHz under some sort of custom water loop (only assuming as I have overclocked these CPUs before at 4.0GHz was impossibly unstable and hot at these clocks on air), the Q6600 would be at a 66% overclock, likely running at 1.5-1.52v, and not likely very stable either.

Overclocking the Haswell to 4.0GHz wouldn't be much of a reach and could accomplish on the most affordable air cooler, and it would still run circles around the Q6600.

But it's okay... the Q6600 is old, less efficient, and less powerful. It doesn't mean the Q6600 wasn't a beast when it was released.

August 10, 2013 | 07:44 PM - Posted by Terminashunator (not verified)

4Ghz on air was a lot more practical than you think.

August 11, 2013 | 03:01 PM - Posted by HeavyG (not verified)

It depends on the BIN, I am sure. I recall seeing some that would hit 4.0GHz at 1.48v, but the two I had would require more than 1.50v, which means you need some pretty serious cooling.

I had both an ASUS Silent Knight cooler and some high rated Thermalright (both out performed the Zalman 9500), and 4.0 GHz meant things were running pretty hot under full load.

Personally, I upgraded one of mine to the i5-750, and I felt like it was a pretty big upgrade. Everything ran... just a little better, even at stock clocks compared to my OC Q6600.

August 10, 2013 | 04:51 AM - Posted by Branthog

Isn't the Q6600 from 2007? More than six years later, shouldn't the 4770K be almost five times faster than the Q6600, going by Moore's law, instead of three times?

Where are we with that, these days? I've been falling off of keeping up with hardware news, but it seems to me we're seeing more like 20% (at best) processor performance improvements per 18 months -- not double...?

Anyway, I really miss the excitement that used to come with putting together a new build. A new high end rig every couple of years and you could see a massive difference. Not so much, anymore.

August 10, 2013 | 07:14 AM - Posted by Sublym3 (not verified)

"Moore's law is the observation that, over the history of computing hardware, the number of transistors on integrated circuits doubles approximately every two years."

Moore's law is about transistors not performance

August 11, 2013 | 10:16 PM - Posted by Branthog

True enough, but that is not how it has generally been applied and measured and, anyway, performance has generally followed suit.

It feels like we are at a plateau. Yes, we're focusing on energy efficiency and so on - but at what point does performance power return to the forefront?

August 10, 2013 | 11:03 AM - Posted by razor512

Of course it will be significantly faster, but those results are a little low, at stock, you can generally get over 3 points on cinebench 11.5

Anyway, an overclocked q6600 can get about 4 (higher if you have a good heatsink) points on cinebench and handle most demanding tasks pretty well.

Their benchmark should have included overclocked benchmarks to see how long an older CPU can remain relevant to modern tasks.

August 10, 2013 | 08:06 PM - Posted by Anonymous (not verified)

You know, I agree with the above observation. I usually get a mid to high end cpu in the hopes that once it starts to show its age it can be oc'ed a bit to get the most life out of it.

August 13, 2013 | 08:48 PM - Posted by Gabe Newell (not verified)

For the -Average- PC user, an overclocked Q6600 with an SSD and Mid to High range video card would handle 99%+ of their needs, including gaming.

I've got both a Q6600 and i7-4770K, both have GTX 770 (not sli) and they both run games at ultra with almost no noticeable difference I could feel.

Video encoding and Photoshop, there was a difference I could feel, but it wasn't huge like I was expecting. Kind of disappointed because mentally I was expecting a colossal increase in speed and processing but it just wasn't there.

January 13, 2015 | 10:48 AM - Posted by homee (not verified)

went from a q6850 to a haswell i3 4330 one example in gaming wnt from 60 fps average benchmark f1 2013 to 135 fps avg with the same graphics card but new mboard ram and cpu

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

By submitting this form, you accept the Mollom privacy policy.