x201 - Intel® Processor Identification problems & Low Benchmarks

Having benchmarked my new x201 with wPrime ( http://www.wprime.net/Download/ ), i noticed a big difference not only with the benchmarks in the NotebookReview.com Χ201 review, but with my Sony VGN-SR29VN (p8600) too.

x201 Config: i5-540M (2.53GHz, 3MB Cache), 4 GB RAM, Win7 Pro 64 bit, HDD 320GB 7200 rpm
This is a new machine, just out of the box, nothing new installed, anti-virus was not installed/running.

wPrime 2.03 results: 47.219 secs (default settings, ie 4 threads, 32M)
wPrime 2.00 results: 41.98 secs (default settings, ie 4 threads, 32M)
x201 windows power mode set to "maximum performance" for both tests (which were performed many times, the above are mean values. I also tried other power modes, and checked that indeed the "Maximum/Minimum CPU processor state" values were at 100%).


But according to the NotebookReview.com review, x201 gets 17.93 secs (wPrime version is not mentioned, but i suppose it is the older 2.00 version)!!! In contrast, my Sony (p8600) gets about 35 secs in wPrime 2.00.

Windows7 identify the processor as [email protected]

I then downloaded the latest Intel® Processor Identification Utility - Windows
Version: Tue, 09 Mar 2010
Support for the Intel Processor Identification Utility

What i'm getting is (maximum performance, 100% processor state as max&min)
Intel(R) Core (TM) i5 540M CPU @2.53 GHz
EXPECTED speed: 2.53 GHz
REPORTED speed: 1.19 GHz !!!

Bus speed: 133 MHz
L3Cache MemoryQ 3 Mb

I tried various power modes (maximum, high, power saver, battery, etc), but the REPORTED speed stays the same, ie: 1.19 GHz (even on Maximum Battery with processor state at 5% max/min)

Am i missing something here, or is this a faulty processor/unit

(along with the slight aesthetic blemish the chassis has? http://forum.notebookreview.com/lenovo-ibm/485153-faulty-x201-keyboard-fan.html )

 

do you have the battery installed when you run the test?

 

No, the battery is NOT installed at all.
I've got the x201 plugged in (and checked the windows7 MAX/MIN CPU processor states when plugged-in are indeed set at 100% for "maximum performance").

 

unless you have the 90 watt adapter, removing the battery would cause CPU throttling. Repeat the test with the battery installed, and you would get a different result.

 

Hmm... just checked my adapter, is the 65W 20V one (didnt know there were different ones)
Wasnt expecting this !!! @Lenovo

You are absolutely right, just installed the battery and got 18.436 secs, which is the expected value.

And the Intel identification utility now shows
REPORTED speed: 2.79 !!!
EXPECTED speed: 2.53
I suppose there is no reason to panic about this difference!

I cant thank you enough

 

the speed difference (2.53 and 2.79) is due to turbo boost kicking in, which is intel's own integrated overclocking utility.

 

all of the X-series come with a 65W adapter.   there's absolutely no reason for any of them to need a 90W adapter as they are designed to be used with a battery installed (as you've discovered ).

 

Can you guys explain why the battery being installed makes a difference at all?

 

There were a few 170W power adapters for like 15$ in the outlet... would have been a good idea to pick one up, but I didn't

(it would charge your battery in about a 1/3 of the time... I guess that's not a huge deal)

 

apart from w70x thinkpad, no Thinkpad laptop have a 170 w power adapter (but they are not compatible with the normal thinkpad).

Battery charge up is not just controlled by the size of the adapter per se, as the on board controller will determine the optimum charge up rate. There is upper limit on how fast a large lithium ion cell takes to charge up, in order to prevent irreversible battery damage and run away chain reaction.

So having a 170 w (even if works) won't mean your battery will take only 1/3 of the time.

 

the short answer is that ThinkPads were designed to be run with a battery.   the battery serves as a buffer to accommodate power spikes in the system.   all power states in the BIOS are set with an installed battery in mind.

since benchmark apps typically max out a system, you'll see lower numbers without the battery installed since that buffer is missing.   the supplied a/c adapter by itself typically cannot provide enough power.   this is why a larger one usually can make up for the lack of battery installed.

 

Does this mean that, for example, a 90W adapter can endanger the system, compared to the default 65W one ?

 

A 90w adapter is safe to use. It is my understanding that the adapters that are even higher wattage (for like the W70x) are not for use with other systems.

 

the w510 adapter can be used, but not sure whether it is going to speed up the charging a whole lot.

 

Power Supply 90W and 65W have the same voltage output, but different output current. And this current they are not issued for a maximum time. The problem comes when the power supply can not provide the desired current, or a very powerful unit is used with a low load, while he was working in the area of low efficiency.

Speed of charging with power supply at much higher power depends on the control system. In theory it should not consume more stored in the settings.