Intel vs AMD on Toms HW

[klnyc]

You guys seen that video about the P4 vs Amd?.The P4 really survie on that test? and those AMD got fried with couple second.
Damn I have Athlon 1.4ghz I hope mine dont look that.

 

[ElFenix]

the p4 has really good thermal protection that clock throttles, certainly a useful feature when you have a $200 processor.

 

[Pabster]

klnyc wrote:

"You guys seen that video about the P4 vs Amd?."

Yes, you're only about the thousandth person to "point out" Dr. Pabst's "test" ...

"The P4 really survie on that test? and those AMD got fried with couple second."

Yes, Intel's thermal management is superior to AMD's. There's no question about it.

"Damn I have Athlon 1.4ghz I hope mine dont look that."

You wouldn't be commenting on Dr. Pabst's article if so, now, would you?

 

[Haribo]

The only thing that anoys me in the test is that the Athlon MP fries like a chicken.

That the first Athlon fries is no prob... it has no thermal protection and is not suppose to be run with no HSFAN.

But the Palomino is a BIG dissapointment... it has "thermal protection" and it doesn't last a second longer than the Tbird
I wasn't expecting P4 "performance" from it bit I sure as hell was expecting it to freeze like the P3.

AMD is no goood right now, they havent fixed the cracking core prob or the "frie in 2 sec" prob

 

[GTaudiophile]

That article/video has basically caused me to rethink my next CPU purchase. I was going to go with nForce/Palomino, but now I am leaning towards an Intel DDR-based chipset/Northwood. I need my computer to last me at least 3 years, so I need reliabilty! I am glad that Intel knows much I'm spending on their CPU and has protection of my investment in mind.

 

[klnyc]

Well, I guess I know where Im heading on my next cpu chip ..Something doesnt fried like a chicken.
Anyhow let me abuse this Athlon 1.4 for now

 

[John]

<< I need my computer to last me at least 3 years, so I need reliabilty! >>



We all want that. However, as long as your HSF doesn't fall off AMD should be a viable choice.

 

[WTBURNETTE]

Also, if you use something like MBM to monitor your temp and fan speed, you can have your system automatically shut down in the event your fan fails, so it's not really a factor...

 

[Haribo]

<< Also, if you use something like MBM to monitor your temp and fan speed, you can have your system automatically shut down in the event your fan fails, so it's not really a factor... >>



MBM doesn't always work... Ive seen many systems where nothing happens.. MBM just lets the temp rise.
They are ofcourse configured correctly and all necessary hardware components for it to work is present.

Software solutions is not to be trusted... that's why Intel has a hardware solution... and why AMD implemented one in the MP but as we now see it's no good.

 

[fkloster]

One more thing to remember. These chips were destroyed in less than a second in some cases. Not enough time for a thermal management system to deter its death.

 

[Athlon4all]

I haven't even looked at the video yet, and I am in total shock! A huge disappointment on AMD's part. This does not mean that you should buy a P4 over an Athlon just because of this, but this is an issume, and AMD really needs to work on it. I really think even more that the P4's approach to Thermal issues (throttling) is very wise. It is clearly the best implementation.

 

[shathal]

Interesting to observe how all of a sudden oppinions on thermal throttling (after Van Smith's rather bad & highly inaccurate report) have gone 180 degrees & it is accepted as a good idea now.

I'm still somewhat disappointed with AMD's solution - hope that this will change in the next processors they bring out. We're not talking here about FHS magically falling off your CPU - that doesn't happen very often.

However - a "scenario" like this is quite possible to occur in other flavours - badly applied thermal grease/solution - a FHS socket breaks off or bad contact between FHS & CPU / dead Fan, etc.

Thermal protection & detection should DEFINATELY both be on-die ... one shouldn't rely on a motherboard to be able to react to the immensy quick temperature changes of a CPU...

 

[rommel]

mmmm....i like fried chicken......lol

 

[Texmaster]

<< I haven't even looked at the video yet, and I am in total shock! A huge disappointment on AMD's part. This does not mean that you should buy a P4 over an Athlon just because of this, but this is an issume, and AMD really needs to work on it. I really think even more that the P4's approach to Thermal issues (throttling) is very wise. It is clearly the best implementation. >>



Actually for many of us who bought a P4 it was one of the biggest reasons. I dont want all my faith in a 20 dollar fan (without the heatsink). Its all about priorities.

I was actually shocked to see Tom of all people doing this. In fact I thought it was a little biased towards Intel!

What caught my eye was the HUGE fan that first Athlon had on it. Now we all know that isn't a standard AMD fan yet the Intel fans looked remarkably Intel Standard.

Also, the AMD on the last test might have been bumped because whoever was taking off the heatsink had a hard time. Notice the intel fan clips were already loose. I just found that odd.

I also think a more realistic test would be to simply cut the fan power off. I would have really liked to have seen those results.

 

[Garet Jax]

<< have gone 180 degrees >>



Every time I see or hear any one use this expression it reminds when Jason Kidd joined the failing Phoenix Suns and said...

"I'm looking forward to turning this franchise around 360 degrees."

 

[AGodspeed]

I need my computer to last me at least 3 years, so I need reliabilty!

Bah, when has your HSF fallen off! I bet never. All Tom's article did for me was assure me that the Palomino's thermal diode is not sufficient for the experiment Tom picked. I've known for a while now that the Palomino would fry just as quickly as the T-bird would in that situation.

The only thing the Palomino's thermal diode does is reduce heat by 25%, which is a good improvement over the T-birds. Now AMD has to work on making a processor that won't fry in 2 seconds with no HSF. Heck, my 1.4 GHz T-bird has been on for 5 weeks straight now, and it's fine. I don't know why you guys think this is a "reliability" problem.

 

[Kingofcomputer]

I don't care,
I only buy retail cpu,
if it's fried, just do a rma.

 

[bigbootydaddy]

i love seeing the amd zealots here...or not here.

not having to worry about heat...is one less worry

booty

 

[Mikewarrior2]

Agodspeed,

where did you see 25% heat reduction? Often times, it is a reduction of 12-18%, depending on speed aplomino. A large key to this is because both palomino's and t-birds are teh same core voltage....



Mike

 

[fkloster]

<< The only thing the Palomino's thermal diode does is reduce heat by 25%, which is a good improvement over the T-birds >>



25% of 600 degrees is 150 degree reduction.. not enough to save a 450 degree part. By the time the Palamino's diode kicked in anyways the CPU was already dead no?

 

[AGodspeed]

25% of 600 degrees is 150 degree reduction.. not enough to save a 450 degree part. By the time the Palamino's diode kicked in anyways the CPU was already dead no?

lol. What? Who said anything about 600 degrees. We're not talking about a Palomino without an HSF. We're talking about the heat/power reduction of a Palomino over a T-bird under the most common/normal conditions (like conditions with an HSF, duh!).

where did you see 25% heat reduction? Often times, it is a reduction of 12-18%, depending on speed aplomino. A large key to this is because both palomino's and t-birds are teh same core voltage....

Mike, where did you read "often times it is a reduction of 12-18%." I have heard figures of between 20-25% less power requirement and heat production. If you can prove your point, I might actually attempt to garner the strength to find and list the 1000 sites that support mine. Or I'll just list the same site that experimented with this whole falling off HSF thing! lol. Dr. Tom Link

 

[fkloster]

<< We're talking about the heat/power reduction of a Palomino over a T-bird under the most common/normal conditions (like conditions with an HSF, duh!). >>



Ummm, wah? I thought the whole purpose of the on board diode on the Palomino was to kick in power reduction AS A RESULT of loss of fan or heatsink. Is there any other reason, besides loss of HS +/or fan, that would cause a chip to become so dangerously hot as to dammage the chip?

 

[AGodspeed]

I thought the whole purpose of the on board diode on the Palomino was to kick in power reduction AS A RESULT of loss of fan or heatsink.

Actually, that's a good question, because I thought the same thing too at one time. Here's what Dr. Pabst had to say on the matter:

the motherboard can read the actual core temperature and apply appropriate measures to safe the life of the CPU (or avoid a notebook meltdown). What I am missing however is the automatic shut down of the Athlon 4 at a certain temperature, without any external electronics.

Here's the link. (It may take a while to load).

So I guess this thermal diode feature doesn't really work? Or does it simply reduce power? I'm not really sure. All I know is that the Palomino will have somewhere between 20-25% less heat production and power requirements than current T-birds.

 

[jcovercash]

yea I saw the video, I dont really know about it, cuz I dont think it means if the HSF is taken off that the chip will survive, but that it will survive if the fan dies. i think

 

[meson2000]

Actually, I thought the themal diode was just for the AMD Power now technology used in notebooks.
So when the AC is unpluged the CPU throttles down to suck less power from the battery. Thunderbirds
and Durons (with the exception of the new 1ghz) don't have PowerNow. It is saddening that the diode
doesn't protect the CPU from burning, but I don't plan on moving away from AMD because of it. I haven't
had a problem with heat at all on my 1.33 and I am only using the stock coolermaster heatsink (not some
monster heavy heatsink). I am not worried at all about it falling off.