whao you actually got it to throttle? i've been keeping logs w/ throttlewatch and i have yet to ever see it throttle.
- May 16, 2002
- 27,286
- 16,123
- 136
Throttlewatch doesn't show it !!!! S&M (I think) wa what Duvie was running showing this. We got down to 2600 @ less than 60c !!!! From 3.4 or 3.5. Its all a blurr now, but bottom line is CRAP CPU !!!!! Watch for Duvies replies...
Actually we hit 2600mhz at 68-69c right before shutdown temp of 70c....At 60c we already had over a 100-150mhz clock throttling going on.....
Go check my other thread..I verified the cpu mhz was directly related to heat with the checking of it by removing and replacing the case side multiple times....It is undeniable...
Go check my other thread..I verified the cpu mhz was directly related to heat with the checking of it by removing and replacing the case side multiple times....It is undeniable...
hey mark, duvie,
i ran s&m 1.73, put CPU load on Long, 100%, for about 9 hrs today.
here are some results...
thrrottlewatch
RMclock
kind of kicking myself because i forgot to checkmark the box on speedfan for cpu temp and voltages
but anyways, this is supposed to be the 820D 2.8ghz @ 3.15ghz. what do you think?
AkumaX,
PM me any more pics if you have them....I would say under 50mhz is not much to be concerend since you are obviously having up and down spikes and I cant say is if I ever saw a time the core 1 was more then 3-5mhz above core 0 let alone the set clock speed.....
What are your vcore and temps like???
PM me any more pics if you have them....I would say under 50mhz is not much to be concerend since you are obviously having up and down spikes and I cant say is if I ever saw a time the core 1 was more then 3-5mhz above core 0 let alone the set clock speed.....
What are your vcore and temps like???
- May 16, 2002
- 27,286
- 16,123
- 136
Well, I was hoping to have my first encoding benchmark result this morning, but the 820 is not done yet. Here is the scenario:
1) 90 minute avi file from my camcorder used as input to windows media encoder on both boxes
2) Both boxes running win2k sp4, mcafee enterprice 8 and microsoft antospyware and 2xF@H units
3) both in the same room with 14c outside temp with the window open 18 inches from the 820 and an open case side ~20c in that area.
4) The 820D has one gig memory in dual channel mode while the 3800+ has 512 meg running single channel mode (more coming next week after this) This could have hurt the 3800, since right before I shut down WME, at was 540 meg used, and the 820D right now is at 637 meg (still encoding). That page swapping could have been bad.
The 3800+@2550 finished the job with the default settings (told it it was going to DVD) in 11 1/2 hours
The 820D is at 13 1/2 hours, and has ~20 left to go. I will update this thread when it finishes, should be in 3-4 hours. (I couldn't wait to post something)
Regardless of the real number there appears to be a REAL difference in speed. Stay tuned.
Edit: I included the temps, since the 820D was at 55C and shouldn't be throttling at that temp. The 3800+ case is closed, and the system is noticably quieter than the 820 system with the case open, but I need it open to not throttle.
1) 90 minute avi file from my camcorder used as input to windows media encoder on both boxes
2) Both boxes running win2k sp4, mcafee enterprice 8 and microsoft antospyware and 2xF@H units
3) both in the same room with 14c outside temp with the window open 18 inches from the 820 and an open case side ~20c in that area.
4) The 820D has one gig memory in dual channel mode while the 3800+ has 512 meg running single channel mode (more coming next week after this) This could have hurt the 3800, since right before I shut down WME, at was 540 meg used, and the 820D right now is at 637 meg (still encoding). That page swapping could have been bad.
The 3800+@2550 finished the job with the default settings (told it it was going to DVD) in 11 1/2 hours
The 820D is at 13 1/2 hours, and has ~20 left to go. I will update this thread when it finishes, should be in 3-4 hours. (I couldn't wait to post something)
Regardless of the real number there appears to be a REAL difference in speed. Stay tuned.
Edit: I included the temps, since the 820D was at 55C and shouldn't be throttling at that temp. The 3800+ case is closed, and the system is noticably quieter than the 820 system with the case open, but I need it open to not throttle.
- May 16, 2002
- 27,286
- 16,123
- 136
I will attach pictures of both summaries when the the 820D finally finishes, but yes, page swapping out over 512, and single channel, and closed case and smaller HSF... How many handicaps can I give to the 820D and it still is ~30% slower and much hotter.
Nice work so far Markfw900...
When you finish with the 820D & mobo, send it my way for a couple of weeks to see what my VapoChill can do with it....
When you finish with the 820D & mobo, send it my way for a couple of weeks to see what my VapoChill can do with it....
- May 16, 2002
- 27,286
- 16,123
- 136
Originally posted by: Technonut
Nice work so far Markfw900...
When you finish with the 820D & mobo, send it my way for a couple of weeks to see what my VapoChill can do with it....
Only one problem with that. The bottom of the line dual-core, the cheapest Intel has, the 820D, and the word vapochill just don't go together.
- May 16, 2002
- 27,286
- 16,123
- 136
OK, here is the results from the two encoding sessions:
The 3800 time 11 1/2 hours
The 820D, 17 hours !
Now I see the bit rate on audio was different, even though I used the exact same procedure. I just used the wizard, said convert a file, for use in a dvd. Why would this occur ? How would it affect the benchmark.
Next, performance difference. I don't know which one to use as the denominator, but its either 33% slower or 49% slower than the 3800. You decide.
The temp on the 820 was 51-53c most of the time, but I did see 55c once. One other odd thing, the 3800 was using a pretty consistent 86-88% of the 2 available cores, but the 820 was using 59-92% of the cores, it was very errectic. 59,85,66,92,77,64,etc.... Also, the temp on the 3800 was consistant at 46c, never really changed and with the case cover ON, where the 820 was case cover off, and a breeze from the window coming right in the side cover (see pics at the top of this post, but I removed the side cover.
The 3800 time 11 1/2 hours
The 820D, 17 hours !
Now I see the bit rate on audio was different, even though I used the exact same procedure. I just used the wizard, said convert a file, for use in a dvd. Why would this occur ? How would it affect the benchmark.
Next, performance difference. I don't know which one to use as the denominator, but its either 33% slower or 49% slower than the 3800. You decide.
The temp on the 820 was 51-53c most of the time, but I did see 55c once. One other odd thing, the 3800 was using a pretty consistent 86-88% of the 2 available cores, but the 820 was using 59-92% of the cores, it was very errectic. 59,85,66,92,77,64,etc.... Also, the temp on the 3800 was consistant at 46c, never really changed and with the case cover ON, where the 820 was case cover off, and a breeze from the window coming right in the side cover (see pics at the top of this post, but I removed the side cover.
I am sure now after seeing it personally, anything that will hold load 100% load and with case side on will reach 55c and therefore begin to throttle....
The erratic cpu usage was seen in the infamous THG review....I am wondering if it has to do with the shared cache and maybe it has to flush out the cache causing moments of less usage....I dont know for sure but the cardiac pattern was something I pointed out during that THG review as very peculiar...
I say you drop it back to something that was stable...At 3.5ghz is froze up with 2 instances of prime after like 5 minutes...It is not stable at those speeds...
You should really test it at 3.43ghz and the 1..45v or 3360mhz and 1.4v....
The erratic cpu usage was seen in the infamous THG review....I am wondering if it has to do with the shared cache and maybe it has to flush out the cache causing moments of less usage....I dont know for sure but the cardiac pattern was something I pointed out during that THG review as very peculiar...
I say you drop it back to something that was stable...At 3.5ghz is froze up with 2 instances of prime after like 5 minutes...It is not stable at those speeds...
You should really test it at 3.43ghz and the 1..45v or 3360mhz and 1.4v....
- May 16, 2002
- 27,286
- 16,123
- 136
I am leaving the side cover off. and it is stable now at 3.5. I tried prime for an hour, and about 36 hours with 2x F@H + WME taking about 75%, for 100% total usage, no problems. I am now using S&M just to monitor any throttling. Maybe I will try WME again for a while just to verify it wasn;t throttling under that circumstance.
What about the 192 vs the 384 audio rate ??? Any insight as to why that happened when I set it up the same way ? Does it affect the benchmark ?
What about the 192 vs the 384 audio rate ??? Any insight as to why that happened when I set it up the same way ? Does it affect the benchmark ?
It didn't freeze because of heat...so that doesn't make sense...It froze becuase of power...it needed more vcore...
- May 16, 2002
- 27,286
- 16,123
- 136
I forgot to mention that I pumped the vcore to 1.475, and in windows under full load it used to be 1.36 while set to 1.45. Now at 1.475, it stays at 1.4v. Sorry for the omission. And I am using S&M now purely as a monitoring tool. Maybe I run the test again with S&M to see if it starts throttling ???
Edit: one more interesting note now that I have throttlewatch running. When first firing up an application, it actually goes up to 3700 or more, and settles down to the 3500 speed after 10-20 seconds. Very strange CPU....
Edit: one more interesting note now that I have throttlewatch running. When first firing up an application, it actually goes up to 3700 or more, and settles down to the 3500 speed after 10-20 seconds. Very strange CPU....
start the app and then click back and forth between the the core 0 and core 1 tab...notice the real time mhz readings....
that's it? hahah.. i can't do that all day.. i wish it would log like speedfan/rmclock/throttlewatch
did you see in your thread how high the temps went and the fluctuations of the vcore+cpu mhz was for my 820D while running S&M on long/100%?
Mark,
that was quite illuminating and it has definitely helped me stay away from the P-D's. I currently own a 3.6 P4 (stock, no OC'ing). Would you or anyone care to comment if non-multithreaded apps will feel as fast on the X2 as it currently feels on my 3.6?
Would love to hear some responses!
Also, I'd like to add, I am NOT a gamer.
that was quite illuminating and it has definitely helped me stay away from the P-D's. I currently own a 3.6 P4 (stock, no OC'ing). Would you or anyone care to comment if non-multithreaded apps will feel as fast on the X2 as it currently feels on my 3.6?
Would love to hear some responses!
Also, I'd like to add, I am NOT a gamer.
You do realize that if you got it off Ebay, there's a very good chance that it was a lemon cpu and somebody just wanted to offload it, right...
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 23K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter