![]() |
![]() |
#1 |
6D616 Posts |
![]()
Hello Everybody,
Hopefully somebody can help point me in the right direction here. I have just build a PC (WIN7 64bit) and are getting lots of BSOD, I have ruled out any driver issues and hence are now looking at hardware. I have ran memtest for 12 hours and got no error. I then ran Prime95 in the small FFTs option for a number of hours and got no errors, the power rails looked OK and the max CPU temp was only 57C. I then ran Prime95 in the Blend option and within 5 mins got this error on each of my 4 cores: FATAL ERROR:Rounding was 0.5, expected less than 0.4 As the blend option uses more memory tests, does this point to a motherboard issue or a CPU , or something else ???? I am running AMD Phenom II X4 965 with a Gigabyte 870A-UD3 motherboard, 2 x 2GB 1333 CL9 DDR3, 850W PSU. I am not overclocking at all, I just want a stable PC ! Have I got faulty hardware, is there any other tests I can do to pin point what I need to replace ? Thanks. |
![]() |
![]() |
#2 |
Dec 2010
10102 Posts |
![]()
Curious to see if you get help, all I got was we "can't help until you run memtest", which I did without errors. After posting the results, no reaction at all from this community!
Good Luck Rick S. |
![]() |
![]() |
![]() |
#3 |
Jul 2009
Germany
547 Posts |
![]()
It could possibly be due to the settings for your RAM (latencies, etc).
Place RAM settings in the BIOS to Auto and repeat the stress test. For my Ga790GPA-UD3h motherboard and OCZ RAM was this approach successfully |
![]() |
![]() |
![]() |
#4 |
"Lucan"
Dec 2006
England
2×3×13×83 Posts |
![]() |
![]() |
![]() |
![]() |
#5 |
Sep 2002
Oeiras, Portugal
1,451 Posts |
![]()
This type of problems is very likely to be originated by some memory issue, particularly in this case, as you are not OCing the system. I am also assuming that the machine goes awry under Prime95 but it is stable otherwise, is that true?
First thing to check: is the memory in the QVL for your MB? Although it is not a MUST, it is at least a SHOULD, specially in cases where the CPU/mobo/memory subsystem is put to stress regularly. Running the memory with the default (SPD) settings is also good advice when troubleshooting BSODs. But I guess you are already doing it. If everything (including voltages) is at factory default and you still got the errors, it is time to try to replace the mem sticks. Assuming you have two, try to remove one at a time and see if that helps: you may spot a faulty stick this way. If the errors persist with just one stick in place and that happens no matter which one is in there, you may consider to RMA them. If you can get somebody to lend you a couple of known good (and compatible!...) mem sticks for testing that would be great too. Last fiddled with by lycorn on 2010-12-06 at 20:14 |
![]() |
![]() |
![]() |
#6 |
Dec 2010
128 Posts |
![]() |
![]() |
![]() |
![]() |
#7 |
May 2010
32·7 Posts |
![]()
How good is your 850W power supply? Some of the generic versions have some stability and noise issues with their 12V buses that can trigger memory errors.
Can you check your Northbridge temperatures? Sometimes overheating there will cause problems. Some people have been surprised when using liquid cooling on the cpu because there's no fan nearby that will improve air circulation over the northbridge and RAM modules. As an experiment, I would try UNDERclocking your memory down to the 1066 MHz level and see if that affects the errors. I too found that memtest would report all happiness for hours while Prime95 would die in minutes with an error. I suspect that the memory strain of Prime95 is higher than imposed by memtest. |
![]() |
![]() |
![]() |
#8 | |
"Richard B. Woods"
Aug 2002
Wisconsin USA
1E0C16 Posts |
![]() Quote:
If you're at least as interested in fixing your system as in attracting posts to the thread you started, you could try looking in the dozens of other Information & Answers threads with "fatal", "error", and/or "rounding" in their titles. (The search function is not entirely useless.) You Advice given to one person is often equally applicable to other persons. Even we veterans sometimes tire of keying the same basic advice over and over and over to each newbie who's not interested in using the search function and/or thinks he's the very first person to have ever had his particular set of error messages and/or thinks that no advice ever given to anyone else can possibly be helpful to himself without having seen it specifically repeated in the thread he started. Really! Try looking in threads started by other people with titles similar to the one you started!! :-) Last fiddled with by cheesehead on 2010-12-07 at 00:13 |
|
![]() |
![]() |
![]() |
#9 | |
Dec 2010
2×5 Posts |
![]() Quote:
|
|
![]() |
![]() |
![]() |
#10 |
"Richard B. Woods"
Aug 2002
Wisconsin USA
22·3·641 Posts |
![]() |
![]() |
![]() |
![]() |
#11 |
6809 > 6502
"""""""""""""""""""
Aug 2003
101×103 Posts
19·499 Posts |
![]()
I can speak for myself, I haven't had to do a lot with hardware problems. When I had a machine that failed with Prime95, memtest did it for me. I was trying to get you start until someone else came along that is more familiar with your system and issues.
|
![]() |
![]() |
![]() |
Thread Tools | |
![]() |
||||
Thread | Thread Starter | Forum | Replies | Last Post |
FATAL ERROR: Rounding was 0.5, expected less than 0.4 | glal12 | Information & Answers | 2 | 2016-09-21 22:59 |
FATAL ERROR: Rounding was 0.5 les then expected 0.4, | sekramer10 | Hardware | 6 | 2015-11-16 13:31 |
Prime 95: FATAL ERROR: Rounding was 0.5, expected | etuckram | Hardware | 1 | 2011-05-27 00:11 |
FATAL ERROR: Rounding was 0.49xxxxxx, expected less 4 | tester | Hardware | 22 | 2003-02-25 22:25 |
FATAL ERROR: Rounding was 0.49xxxxxx, expected less 4 | tester | Software | 1 | 2003-01-17 04:45 |