CPU gone to AMD heaven

[Nasty]

I got my Thunderbird 1.4 3 weeks ago and the CPU just blew it's load, sitting in windows and it restarted for no reason, after this it threw my into the BIOS menu, restarted again and then a black screen.

My motherboard is ASUS A7V133-C and I have a 300 watt power supply. Idle temp was about 47 C and full load was about 55-58 C, there is no way it fried.

I put it down to faulty workmanship on the chip, never had this problem with INTEL, makes you wonder, maybe INTEL isn't that bad.




AMD Thunderbird 1.4
Coolermaster EP5-6I11
2 Case Fans 3000 rpm, 1 System Blower 15000 rpm
freq 10.5 - FSB 133
voltage 1.85

47C idle 55-58 full load
Motherboard A7V133-C
256 MB PC133 SDRAM
Nvidia TNT2 M64 32MB SDRAM
10 GB Quantum Fireball
CDRW drive

[donny_paycheck]

Yeah, if you don't mind paying 2x as much for a chip that's 3/4 as fast...

[FRANK]

Originally posted by Nasty
I got my Thunderbird 1.4 3 weeks ago and the CPU just blew it's load, sitting in windows and it restarted for no reason, after this it threw my into the BIOS menu, restarted again and then a black screen.

My motherboard is ASUS A7V133-C and I have a 300 watt power supply. Idle temp was about 47 C and full load was about 55-58 C, there is no way it fried.

I put it down to faulty workmanship on the chip, never had this problem with INTEL, makes you wonder, maybe INTEL isn't that bad.




AMD Thunderbird 1.4
Coolermaster EP5-6I11
2 Case Fans 3000 rpm, 1 System Blower 15000 rpm
freq 10.5 - FSB 133
voltage 1.85

47C idle 55-58 full load
Motherboard A7V133-C
256 MB PC133 SDRAM
Nvidia TNT2 M64 32MB SDRAM
10 GB Quantum Fireball
CDRW drive

If the CPU is only three weeks old, get an RMA and return it. Although these CPU's do run on the hot side, if you follow the generally accepted practices concerning overclocking, particularly the cooling, these are very reliable products. Most of the problems that people have with these CPU's stems from the HSF.

To attest to their durability, I've been very successful in my endeavors with the socket A CPU's, DURON 600's to better than 1100MHz and 1G T-BIRDS @ 1.42G running the FSB as high as 166MHz. The only problem I've encountered is one chipped core on a 1.2G T-BIRD that still overclocks to better than 1.4G.

Considering all of the changes that I've made to my systems and the number of times that I've removed and replaced the HSF, (probably more than any "normal" consumer would on ten computers in their lifetime) I don't find it an issue.

By the way, the ASUS Probe is known to report temperatures that are as much as 10C higher than actual. Both my A7V and my A7V133 regularly report the temperatures in the low 60C's and they both run as stable as Windows will let them.

I'll concede that Intel makes a high quality product, but this company is certainly not without it's issues,. If I remember correctly, the 820 chipset was one of their "black eyes" as well the recall of the entire production of the 1.13G PIII, not to mention the early Pentiums that couldn't count.

If you're not comfortable with the products that AMD has available, then buy which ever works for you. But the general consensus is the you will get the most bang for your buck, both in performance and cost with the products that are available from AMD.

My .02

[Myself and I]

For some reason something does not sound right in this picture
"I got my Thunderbird 1.4 3 weeks ago and the CPU just blew it's load"!!!

Where did your cpu blew it's load on?

[Christoph]

FRANK, your temps are in the low 60 °C range? Or did you mean that the reported temps varied from the actual temps?
If it's 60 °C, that's waaaaaay too hot for my tastes, but I suppose you senior members know what you're doing.

[FRANK]

Originally posted by IdeaMagnate
FRANK, your temps are in the low 60 °C range? Or did you mean that the reported temps varied from the actual temps?
If it's 60 °C, that's waaaaaay too hot for my tastes, but I suppose you senior members know what you're doing.



As I clearly stated in the above post;

>"the ASUS Probe is known to report temperatures that are as much as 10C higher than actual"<

It has been well documented in this as well as many other forums about this shortcoming on the ASUS boards, the A7V and A7V133 in particular. Having owned both of the above ASUS socket A motherboards, as well as several others from FIC and Epox, my experiences have confirmed this. I have swapped CPU's from one system to another with all things being equal and have seen a temperature difference of 10C+/-, with the ASUS boards consistently being on the high side



And lastly, I may be wrong but I think I detected a note of sarcasm in your reply.

>but I suppose you senior members know what you're doing<

I must say that I can't speak for the other seniors, but I have never steered anybody in any direction that would jeopardize their components beyond the assumed risk naturally involved in overclocking. Any suggestion that I have made to anybody, I have either done myself or have done the necessary research that many of them have neglected to do themselves. If you don't feel comfortable with "as reported" temperatures in the 60's, then by all means don't let them get that high. Either back down on your clock speed or get a more reliable temperature reporting solution.

As far as my track record with the socket A platform, it goes something like this;

I started with one of the first socket A motherboards the FIC AZ11 with the notorious KT133 chipset, along with a DURON 600 AKBA. Very limited overclocking potential with this board. I followed this with an ASUS A7V although still saddled with the KT133 chipset, it had multipliers and I was able to overclock that DURON to 1.1G (110*10). I picked up a 650 T-BIRD for the AZ11 at that time. Next in line was the A7V133 with the much improved KT133A chipset and another AKBA DURON 600. This combination is good for 1.13G (8.5*133), I did manage to get the FSB as high as 140MHz (140*10) with an AXIA 1G T-BIRD. About this time DDR RAM made it's debut, again I jumped in early, this time with the FIC AD11 and an AXIA 1.2G T-BIRD. This setup is good for 1.47G. While fiddling with the FIC board I noticed that the temperatures were consistently lower on this board than either of the ASUS boards regardless of the combination of CPU and state of overclocking. And finally, the advent of the new KT266A chipset and the much acclaimed Epox 8KHA+ with it's support for very high FSB speeds, right now I'm running the 1G T-BIRD@1.41 (166*8.5) again with temperatures as reported, much lower than any ASUS board that I've worked with.

I've always used quality heat sinks with Global Win being my hsf of choice. I consider a successful overclocking attempt being able to run apps as well as the generally accepted burn-in programs such as SIS Soft Sandra etc. I run SETI 24/7 on all of my systems yielding a 100% CPU load. These systems are very stable, as much as Windows will let them be, I use Win 98SE, ME, 2000 pro as well as LINUX. The results that I've been able to get indicate to me that there are no heat related issues on my systems, other than inaccurate reporting by certain on board programs.

If I'm mistaken about the tone of your reply, I apologize for my terse response.

[Christoph]

I didn't mean to seem sarcastic.
If I see some obvious newbie rejecting good advice, I'll indulge in a bit of sarcasm. I know however that you seniors are seniors because of your knowledge and willingness to help others, and I give your opinions the weight they merit.
If Asusprobe gives you inaccurate temps, why not use something like MBM 5? As for me, I really wouldn't like seeing 60°C, even if I knew it was inaccurate.

[FRANK]

Originally posted by IdeaMagnate
I didn't mean to seem sarcastic.
If I see some obvious newbie rejecting good advice, I'll indulge in a bit of sarcasm. I know however that you seniors are seniors because of your knowledge and willingness to help others, and I give your opinions the weight they merit.
If Asusprobe gives you inaccurate temps, why not use something like MBM 5? As for me, I really wouldn't like seeing 60°C, even if I knew it was inaccurate.

I apologize for the attitude of my above response. No offence taken I hope.

As far as the MBM5 as a substitute for ASUS Probe, keep in mind that regardless of which program that you use to monitor the sensors, they are still monitoring the on-board sensors. My experience has been that MBM5 gives the same readings as ASUS Probe, in addition MBM5 has an option in their setup to adjust the readings of the various sensors. So if you are not sure where you're at to begin with, then, in my opinion, MBM5 gives no advantage .

The only reliable way to attain an accurate temperature reading is with and additionally installed thermistor and that too could be regarded as subjective depending on it's placement. I will concede that this method even though still possibly inaccurate to some degree would be better than the situation on the ASUS boards as it is now. I've found that in my case, I am satisfied to monitor the reported temperatures which are generally stable, and watch for any unexplained increases, than it is to be critically accurate to within 1/2 of a degree or so. I've found that if I use a quality hsf and follow good cooling practices, the temperatures as reported will not be my limiting factor.

MY .02

[kardinal]

ill prolly get flamed for posting this but, i have an asus a7a266, and i emailed asus about my temp problems and here is what they said.

The thermal resistor used for monitoring, is placed at different locations on both motherboards. It is closer to the
CPU on A7A266, thus registering a higher temperature. The temperature is actually more accurate with the
A7A266.

that may go against the grain but it makes sense to me...

i know i know the newbie knows nothing

/me runs and hides in the corner

:D:D:D;)


*edited in*
they also said this about my 1.4ghz tbird...

The temp issue is not caused by our motherboard, this should be the AMD processor, Please refer to the
following link with regard to the AMD thermal information. Under page 4 of the document it says that the
maximum die temperature is 95 degree celcius for 1100MHz and higher.

[Christoph]

Glad that's resolved.

[MrX]

Look....Ive personally only built computers for 7 years but Ive also overclocked a good amount of those systems. I know, I know, the guys system was running a little hot....but at no time did it go over 58C! Many other posts and friends experiences prove that this may degrade the chips life but not down to 3weeks! The main problem I ascribe to this was the use of an OEM HSF shipped with his chip, a hot running Asus and a bad quality EYHJA. I observed him in everything he did and at what appeared to be a random time whilst in software the effects he mentioned happened. Apon closer examination the chip, HSF and MOBO are unmarked in mint condition. At AMD they say 95C cas temp is the max. If this case had gone even half of that his chip wouldnt even be able to boot.