Coolest AMD CPU

The forum for non-component-related silent pc discussions.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
DarkNMSS
Posts: 25
Joined: Mon Apr 28, 2003 12:12 pm

Coolest AMD CPU

Post by DarkNMSS » Fri May 23, 2003 12:29 am

I'm sorry if this question has been asked, but I couldn't find it. I am wonder which CPU is the coolest. From what I know, it should be the Thoroughbred B core 1700+ DLT3C processor. Is this correct or not? If someone has this chip, please tell me what type of temperatures you are getting and what heatsink set up you have. TIA.

dago
Patron of SPCR
Posts: 445
Joined: Wed Apr 23, 2003 8:50 am
Location: BE, CH
Contact:

Post by dago » Fri May 23, 2003 1:04 am

Yep, just got one, plugged onto a asus A7V333 motherboard, default voltage (1.5V), planning to undervolt when I've time to change a jumper.

Cooled with a Zalman 6000 (gallery pics), the fan cooling it being a 120mm fan over the CPU and a 92mm zalmann (@5 V). Airflow not very good because // to the zalman -> planning to mod a cheap psu to put a 120mm (like the fortron ones) and to remove the over-the-heatsink fan.

Now, to answer your question : normal use 40°C-45°C. Never overcomes 50°C even when folding or cpu burning.

DarkNMSS
Posts: 25
Joined: Mon Apr 28, 2003 12:12 pm

Post by DarkNMSS » Fri May 23, 2003 1:32 am

Thanks for the reply. 40-45 degrees is pretty good considering my Tbird 1.33GHz is running at 55 degrees full load. My AlCu Zalman heatsink/92mm fan combo is pretty loud right now and I was thinking about getting one of these new cores so I can turn down the cpu fan. When I turn down the fan to almost inaudible, the temperatures rise to 67 degrees which is a tad too high for my liking. It's not even summmer time yet. I always thought the temperaturess would be cooler than that though for a Tbred B core.
Last edited by DarkNMSS on Fri May 23, 2003 1:57 am, edited 1 time in total.

dago
Patron of SPCR
Posts: 445
Joined: Wed Apr 23, 2003 8:50 am
Location: BE, CH
Contact:

Post by dago » Fri May 23, 2003 1:43 am

Keep in mind that the airflow is very poor : it flows (on the pictures) from the right to the left, because the PSU doesn't suck any air (no inlet on that side).

In any case, I feel at ease with temps between 40 and 50, it's not high, and I don't see any reason why lower will be better (overclocking apart) as the other components aren't hot.

Also another notes : even if it's not yet summertime, we've got a wonderfull spring which leads to ambiant temps in my south-directed flat around 23. Now that it's more cloudy/rainy, temps dropped a few degrees for my CPU also.

Gandalf
Posts: 331
Joined: Tue Dec 24, 2002 9:04 am
Location: Belgium

Post by Gandalf » Fri May 23, 2003 6:14 am

Wait for the AMD64 .. they'll be (apparently) a lot cooler :D.
Another thing I'm waiting for is for DDR-II to become the standard ... less power :).

stsmytherie
Posts: 25
Joined: Tue Oct 15, 2002 8:38 am
Location: Seattle, USA
Contact:

Post by stsmytherie » Fri May 23, 2003 7:29 am

I posted information about this recently, and posted thermal data for all AMD procs.

Try this: http://forums.silentpcreview.com/viewtopic.php?t=4637

Athlon Powers
Posts: 376
Joined: Sun May 04, 2003 4:33 pm
Location: Athlonville, My Computer
Contact:

Re: Coolest AMD CPU

Post by Athlon Powers » Fri May 23, 2003 9:41 am

DarkNMSS wrote:I'm sorry if this question has been asked, but I couldn't find it. I am wonder which CPU is the coolest. From what I know, it should be the Thoroughbred B core 1700+ DLT3C processor. Is this correct or not? If someone has this chip, please tell me what type of temperatures you are getting and what heatsink set up you have. TIA.
That would be it right there, because of it's P4-like core voltage a lot of people can get almost a 100% overclock with a good board! :o:D

Gandalf
Posts: 331
Joined: Tue Dec 24, 2002 9:04 am
Location: Belgium

Post by Gandalf » Sun May 25, 2003 8:44 am

Blergh. It's disappointing that the Barton still disappates that much heat :(.

DarkNMSS
Posts: 25
Joined: Mon Apr 28, 2003 12:12 pm

Post by DarkNMSS » Sun May 25, 2003 10:37 am

I can't seem to find much information on the Athlon64. How do you know it will run cooler? Is the core voltage going to be lower than 1.5V or something? I don't really mind too much for speed as my current 1.33GHz is fast enough for my needs (for now). How much cooler would a P4 be compared to the AMD? Is a P4 2.4 800MHz CPU going to be hotter than the P4 2.4 533MHz? And are these two P4s hotter or cooler than the AMD?

DryFire
Posts: 1076
Joined: Sun May 25, 2003 8:29 am
Location: USA

Post by DryFire » Sun May 25, 2003 10:57 am

the athlon 64 will use SOI technology. It should make the chips run much cooler.

From what i have read the opterons run ~44 dg C on a 1 u rackmount server. Which are very hard to cool.

Not to mention the A64 should be slightly cooler then it's opteron counter part.

Mark Larson
Posts: 105
Joined: Mon Feb 10, 2003 4:04 pm
Location: MN

Post by Mark Larson » Sun May 25, 2003 11:14 am

DarkNMSS wrote:I can't seem to find much information on the Athlon64. How do you know it will run cooler? Is the core voltage going to be lower than 1.5V or something? I don't really mind too much for speed as my current 1.33GHz is fast enough for my needs (for now). How much cooler would a P4 be compared to the AMD? Is a P4 2.4 800MHz CPU going to be hotter than the P4 2.4 533MHz? And are these two P4s hotter or cooler than the AMD?
Both the P4s will be hotter.

DarkNMSS
Posts: 25
Joined: Mon Apr 28, 2003 12:12 pm

Post by DarkNMSS » Sun May 25, 2003 11:16 am

I guess it would be wise to wait until September to purchase a new CPU since that's when these babies are going to be in production. Probably won't be seeing it till November though. :lol:

Thanks for the input on the subject guys. It's invaluable information for me.
Last edited by DarkNMSS on Sun May 25, 2003 11:19 am, edited 1 time in total.

KenAF
Posts: 114
Joined: Sat May 17, 2003 4:32 pm

Post by KenAF » Sun May 25, 2003 11:18 am

Don't forget, the Athlon64 is based on the current Athlon core, but is able to take better advantage of it thanks to the on-die memory controller (which normally consumes power and dissipates heat in the northbridge--and will now do the same on the cpu); the same core at higher ipc means more work, which means more heat. These will serve to offset much of the remaining benefits of SOI.

Given an identical processor design, SOI allows use of higher frequencies than possible otherwise OR it lets you run the chip cooler. It does not do both. Moreover, it takes lots of work to optimize for and take advantage of SOI. In all likelihood, forthcoming Athlon64 chips at 2+GHz will be just as hot as current chips. The Opteron at 1.8GHz already dissipates comparable heat to a P4 at 2.6GHz--again, because it does more work per cycle than Athlon chips, and it has the heat from the integrated memory controller and HT links to deal with.

Gandalf
Posts: 331
Joined: Tue Dec 24, 2002 9:04 am
Location: Belgium

Post by Gandalf » Sun May 25, 2003 11:23 am

Bullshit.
The 64's will more than likely start at 1.8gig (3000+ apparently .. :/) and be put in *ceramic* casings (like early cheap ass durons) to lower the costs *and* because they won't get as hot. (If they were as hot as the current ones they'd be in the organic casing).

A 3000+ 64 will be cooler than a 3000+ XP, simply put.

KenAF
Posts: 114
Joined: Sat May 17, 2003 4:32 pm

Post by KenAF » Sun May 25, 2003 12:36 pm

Gandalf,

My employer (Broadcom) designs ICs, and we have looked at SOI. We own two Opteron systems and they dissipate comparable heat to our Xeon 2.6 systems, thought not quite up to the 80.6 watts specified on AMD's web site. For reference, Barton's max is 76.8 watts @ 2.2GHz, while the P4 has a TDP of 61.5 watts @ 2.53GHz.

I never said a 3000+ 64 would put out more heat than a 3000+ XP. I said (implied) they would be comparable at identical clock speeds, given the other differences in the two processors (higher IPC + memory controller + HT link), and the lack of SOI process maturity. As the SOI process matures in the next 12-18 months, there may be some additional benefit--which will probably be used to increase clock speeds rather than reduce power requirements. On occasion, we discuss this on the RWT forum.
Last edited by KenAF on Sun May 25, 2003 3:01 pm, edited 2 times in total.

DryFire
Posts: 1076
Joined: Sun May 25, 2003 8:29 am
Location: USA

Post by DryFire » Sun May 25, 2003 2:45 pm

ken is quite right.

Post Reply