1558 checks and test to do when it arrives

Need help from the 1558 and possibly 1557 owners to check when my 1558 comes in a couple of days. I am getting this from the outlet so I am a little more concerned about the possible issues. I am looking for guidance to look for typical issues people have seen so I can check them out as soon I get it.

Also looking for CPU and heat test software I can use to make sure it does not have heat issues.
Right now I have the Hmonitor and the Intel burnin test. Any advice on this front is appreciated.

My specs are
Studio 15 - 1558 Laptop
Certified Refurbished
64BIT Operating System CD
Processor: Intel Core i5-520M Processor (2.4GHz, 4 Threads,3M cache, turbo Boost up to 2.93GHz)
4 GB DDR3 SDRAM 1333MHz (2 DIMMs)
Back-lit Keyboard
500 GB SATA Hard Drive (7200 RPM)
512MB ATI Mobility Radeon HD 4570
Intel Centrino Advanced-N 6200 with Netgear WNR2000 Wireless N Router
125V Power Cord
Sound Blaster X-Fi MB Panzer software version 1.2
9 Cell Primary Battery
Integrated 2.0 Mega Pixel Web Camera
Dell Wireless 365 Bluetooth Module
8X DVD +/- RW w/dbl layer write capability
Microsoft Works 9.0
Dell WebCam Central
90W AC Adapter
Genuine Windows 7 Home Premium
15.6 inch Full High Definition (1080p) High Brightness LED Display with TrueLife
roxio Software CD
Black Chainlink

 

Prime95 and Furmark are two good tools to create some heat in the CPU and GPU.

Use ThrottleStop to keep an eye on the multiplier and clock modulation settings. An i5-520M has a default multiplier of 18. When fully loaded, it should not drop below that value. When it does, that's a sign of throttling.

At full load, clock modulation and chipset clock modulation should be steady at 100.0%. There is no reason to slow a laptop down with these but Dell continues to use these "features" on many of their laptops.

http://www.fileden.com/files/2008/3/3/1794507/ThrottleStop.zip

Send me a log file if you're not sure about the results.

IBT creates lots of heat but it does not create a consistent load so this might interfere with your results and make it harder to determine what is throttling and what are normal pauses between calculations when IBT is running. Prime95 Small FFTs creates less heat but is very consistent.

 

I find it hard to believe that Dell uses this. These "hacks" are useless on newer chips like the Core 2 Duo, i7-M, etc. as they have built in down clocking and power phase lane shutdowns built into the chip and chipsets.

 

Thanks, is Throttlestop for monitoring only or actual stress test.
Do you have experience with Burnin tester and Intel burn in test 2.4?
I downloaded them and planning to use them when the laptop comes in.

 

Unfortunately, Dell does use these "hacks" on a wide variety of modern laptops. The clock modulation feature is built into all Intel CPUs since the Pentium 4 era. It's original purpose was to control temperatures during emergency situations but unfortunately Dell and other manufacturers are using this to slow modern laptops down to a crawl during normal use, especially while trying to play a game. This combined with multiplier throttling can reduce a Core i laptop down to as little as 10% of its rated performance.

It's a lot cheaper to throttle a laptop at the bios level compared to the cost of engineering a quality laptop with adequate cooling and the additional cost to include a power adapter that can fully power both the CPU and GPU at the same time.

iafzal3: I prefer LinX to IBT but they both use the Intel Linpack libraries and either of them can give a CPU a good workout.

As each problem is solved, the load will drop significantly for a few seconds as the next problem is being set up. If you are sitting at your computer and are watching this then the multiplier suddenly changing will make some sense to you but the log file will be more difficult to track because of this. That's why I prefer Prime95 Small FFTs. The load from that program is extremely consistent. If you run Prime95 and Furmark at the same time, you will be able to give your laptop a thorough work out. There are no games or other real world applications that can equal this combined load so for testing, I consider this to be good enough.

ThrottleStop is for monitoring and will give you a very accurate look at what the CPU multiplier is really up to for each core / thread. On Core i based processors where turbo boost can be cycling on and off a hundred times a second, most software like CPU-Z, HWMonitor, etc. won't give you an accurate look at what's going on and neither of these correctly report the clock modulation that is being used by Asus, Acer and especially Dell.

 

Personally for CPU stress testing I like Everest by Lavalys. It'll take the CPU right up to its max in the stability test, see here.

 

OK I was able to get hold of Everest and run the test. Below are the result ran the test for 10mins and the CPUs went all the way 68C max but hovered between 60-68C range. Not sure if this is calibrated correctly as after the test was done they went to about 40C. When I ran the test with another software at idle the CPUs would be around 22C.
I have i5-520 with 5470 card. I noticed a multiplier of 20@ 2.66GHz stable running of the CPU. It is supposed to go to 2.9GHZ @ full load.
Any feedback on the tests? can I do anythign else with the Everest to check out the GPU and CPU and take the CPU to 2.93Ghz. I only noticed it to go to 2.66GHZ (20x) and in another test go to 28.60Ghz in another test.

 

That's not how these CPUs are designed to work. The default multiplier is 18 for a Core i5-520M. When a single core is active and the second core is asleep, turbo boost will allow the multiplier to briefly go as high as 22. When the second core wakes up, the maximum for both cores is a multiplier of 20.

When both cores are fully loaded, it is impossible to see more than the 20X multiplier. When running a single threaded benchmark like Super PI mod, the multiplier should rapidly cycle between 20 and 22 but the majority of software does not report this correctly.

You can add Everest to that list. Most software does not follow Intel's method to accurately monitor these CPUs. ThrottleStop does and it will show you EXACTLY what the multipliers for each thread are up to.

 

Thanks for the info
I used BurninTest to stress the CPU to max and then used Intel gadget to see the frequency of the CPU. I assumed it was stressing both cores but I can be mistaken. I could see the CPU fluctuate between 2.66Mhz(20x) and 2.79Mhz(21), never 2.93Mhz. I never saw it go to 22x but I do remember it jump to 22x multiplier while I was doing random stuff. Must be a single core running and that is why it went to 22x.

I will try Throttlestop to make another check but it seems my CPU is good no throttleing issue as far as I can see except at full load for 5 mins temps are almost at 70C.