Wow, what’s with all the CULV hate? I’ve never seen a quietly introduced, low volume, mobile Intel CPU get so much negative press before.

CULV stands for Consumer Ultra Low Voltage. It’s a badge affixed to certain mobile Core 2 Duos that run at lower clocks and lower voltages than standard mobile Core 2 Duos. Just as some CPUs can overclock higher than others, some CPUs can run at lower voltages than others. It all has to do with the bell curve for CPU yield; while the majority will run at a normal frequency/voltage range, some will turn out to be exceptional parts.

The CULV parts also use the smaller packaging Intel first debuted in the MacBook Air. The bottom line is that these chips will enable smaller, thinner laptops but without dropping down to Atom-level performance. Intel has traditionally always done an ultra low voltage mobile CPU, this time it's just getting marketed a bit stronger - partially because of the smaller packaging, and additionally because of the stronger market for ultra portable notebooks now.

Processor Clock Speed L2 Cache TDP
Intel Core 2 Duo SU7300 1.30GHz 3MB 10W

 

The max TDP for the CULV line tops out at 10W, with the lowest end single core chips dissipating a maximum of 5.5W. That’s not much more than an Atom processor.

In contrast, standard voltage mobile Core 2s have TDPs ranging from 17 - 35W.

It’s not all about dropping the voltage though, clock speed unfortunately suffers. Most CULV chips run in the 1.2 - 1.4GHz range. While that’ll still be much faster than an Atom, it’s roughly half the frequency of a standard mobile Core 2. You can argue that clock speed doesn’t matter, but you’ll notice the difference between a 1.2GHz Core 2 and one running at 2.4GHz.

And that’s ultimately why OEMs view CULV as a failure in North America: performance. Or at least that’s what they’ve been going around telling everyone.

To a certain extent even Apple has recognized the poor performance of ultra low voltage chips. When it introduced the MacBook Air, instead of going with a 1.2GHz ultra low voltage Core 2 Apple used a low voltage 1.6/1.8GHz Core 2 Duo. The difference in clock speed is one of the things I pointed out as a reason why the MacBook Air didn’t suck.

ASUS must’ve come to the same conclusion, because although its new UL line use a 1.3GHz CULV Core 2 Duo the chip ships overclocked. Yep.

The ASUS UL50Vt: Overclocked CULV
POST A COMMENT

51 Comments

View All Comments

  • gstrickler - Monday, September 14, 2009 - link

    Why not just use the SU9600 CPU rather than overclock the U7300? Cost is the only reason I can think of. SU9600 has a 10W TDP, runs at the same voltage, and is sped'd @ 1.86GHz. The SU9600 might actually use less power than overclocking since the memory bus would not be overclocked (800MHz vs 1066MHz), and the faster CPU/cache would negate most differences in memory bandwidth. Reply
  • snakeoil - Monday, September 14, 2009 - link

    We are still waiting for the lynnfield results without the overclock (turbo overclock).
    the lynfield review is not valid because it corresponds to an overclocked processor.If we don't get those results we don't know what to do with lynfield.
    Reply
  • fitten - Tuesday, September 15, 2009 - link

    It corresponds to a processor as it operates in normal mode as it is intended to operate and, indeed, will operate on every machine shipped. By shutting off features of the CPU, you're not benchmarking how the actual processor would be working, you're only chasing some academic discussion about the processor. NO ONE will run the processor in that mode (it'd be stupid to shut off features because they "aren't fair") so any results obtained are useless except for tangential academic discussion.

    Why not shut off the L2/L3 caches as well or at least make them run at only the size of the competition? That's "fair", right? You might as well make sure that all tests are done with the processors running at the same clock speed... all in the name of "fairness", you know...
    Reply
  • Lonyo - Monday, September 14, 2009 - link

    Go troll elsewhere please.
    After reading your comments for other articles, it's clear that the only thing you care about is asking Anand to benchmark a processor with a built in feature disabled.

    Maybe next time you should ask to get this system benchmarked with the overclocking disabled? And the underclocking too.
    Or maybe you could ask for a review of an HD4870X2 with Crossfire turned off, because Crossfire gives it an advantage over, say, a single GTX280 which doesn't have to run SLI.

    Keep your irrelevant and stupid posts elsewhere, if possible on another website.


    One request for the review of this is to maybe look at doing a browser test, especially with the built in overclocking feature enabled.
    Following on from the other browser test, it would be interesting to see how a higher variability in CPU clockspeeds would change battery life when a variety of browsers are used.

    Def. looks like the 13" model could be a nice step up from high-end netbooks if the price is right.
    Who would pay $600 for an 11" netbook with 8 hour battery life and poor (relative) performance if you can get a slightly larger 13" laptop with "real" processor, similar battery life, and a minimal increase in size and weight?
    Reply
  • MamiyaOtaru - Monday, September 14, 2009 - link

    Turbo comes with the processor and requires no user intervention. It's a feature. You are asking for a review of an intentionally gimped processor where a standard feature is turned off. There is no reason for such a review. Reply
  • jonup - Monday, September 14, 2009 - link

    Very relavant place for your post. Reply
  • DigitalFreak - Monday, September 14, 2009 - link

    "If we don't get those results we don't know what to do with lynfield. "

    Die?
    Reply
  • yacoub - Monday, September 14, 2009 - link

    No, but we are waiting for the P55 motherboard round-ups that were supposed to start late last week. :( Reply
  • PotatoMAN - Monday, September 14, 2009 - link

    This reminds me of the days of turbo buttons and such when I first started playing with computers (around 80486 days and such). This seems neat, although I would be curious to see if doing simple web browsing makes the system scale back CPU cycles at all? I don't doubt the 7 hours of battery life, but I am a little confused by the intent to go to "turbo" mode then just surf the internet (why would you need to go to turbo mode for this, unless flash playback is bad in regular mode?). Why not go to "turbo" mode and max out the CPU with some Prime95 tests or something? Just a thought... thanks for all the good work and techy journalism AT! Reply
  • XBoxLPU - Monday, September 14, 2009 - link

    Why not even look at the SU4100 and the tasks usually done by these typically Netbook segment products. The SU4100 is a dual core CPU that generally improves multi tasking greatly, even for single threaded applications. I would wager that a SU4100 would eat any Atom CPU alive Reply

Log in

Don't have an account? Sign up now