amd should correct its exposed core design

[capybara]

always using amds, with their delicate exposed core right on top of the chip,
i never gave it a second thought. until an intel user pointed out to me, this is
a design flaw. and i must agree. why cant amd protect its cpu core
just like intel does? heat dissapation issues?

 

[CTho9305]

Costs? Lack of a problem? OEMs put heatsinks on properly, and only a few other people break cores, so I really don't think it is a particularly big deal. I woudl think OEMs make up the bulk of most hardware manufacturers's sales anyway.

 

[supiammike]

if it was a big deal wouldnt amd fix it? and wouldnt we see more poeple that have problems? I have put together 3 computers and all three amd chips work perfectly for me

 

[Peter]

If it were a design flaw, isn't it strange then that Intel still makes plenty of exposed-core parts too - including CPUs. Having a heat spreader on is a bit safer for the DIY market where clumsy users might be on the case; but not having one is substantially better in other markets - the heat transfer is better, and in some cases you need the few millimeters won in height, e.g. in notebooks. In the largest chunk of the market, which is premounted complete systems where skilled or at least practiced people assemble the systems, it just doesn't matter.

But if it makes you happier, Opteron does have a heat spreader on.

 

[Rainsford]

Durability is not the first thing on the list of things I look for in a CPU...they've got better things to do.

 

[AEB]

i look at overall design because the future is important, AMD's do run hotter than P4's im not saying the design is why but details are important to a lot of people, it shows to me that thought was put into the design. Thats why i like german cars

 

[Peter]

Athlons running hotter than P4? That's a myth, dude. P4 power consumption is no lower than Athlon's. Not at all.

 

[GoHAnSoN]

the core is pretty much protected anyway. the open side of the core is quite thick protective layer

 

[mechBgon]

If there's a design flaw that needs correcting, it's that the SocketA heatsink is a directional design but can be installed backwards. It would've been simplicity itself to make the heatsink's clip with two lugholes on one end and three on the other, or any other method of preventing a reversal. I wish I had a nickel for every time I've had to ask whether the heatsink's stepped end is matched to the raised, solid-plastic end of the CPU socket

Looks like Opteron and Athlon 64 use completely symmetrical heatsinks

 

[Warin]

In 100+ installations of Athlon and Duron processors, I have never once cracked a core. Even when the processor was not properly seated (I relied on the shop I bought the mobo from to install the processor properly. DUH! ) I havent broken one.

The integrated heat spreader is a great idea to make it totally foolproof, as well as helping with heat dissipation. I think thats why AMD has an IHS on the Opteron and Athlon 64.

 

[Mingon]

The integrated heat spreader is a great idea to make it totally foolproof, as well as helping with heat dissipation. I think thats why AMD has an IHS on the Opteron and Athlon 64.

The IHS does not really help with heat dissipation - it is just another thing in the way - but it does help with heatsink surface alignment and protection

 

[Goatsan]

well they ARE getting rid of it, have you seen their new operton cpu's, they have a heat spreader on them

 

[Jeff7181]

Heatspreaders should be on Retail processors only... so when billy bob goes to Best Buy and buys one of them new fangled processor thing-a-majiggys he doesn't break it.

 

[Jeff7]

Originally posted by: Jeff7181
Heatspreaders should be on Retail processors only... so when billy bob goes to Best Buy and buys one of them new fangled processor thing-a-majiggys he doesn't break it.

Sounds like a good idea. Of course, there's still that piece of plastic sitting over the thermal bubble gum stuff that kills procs when people don't know what they're doing AND don't read the directions..
And they'll also wonder why their new socket A processor won't fit in the socket 478 on the motherboard. They both say "Designed for Windows XP!" so what gives!!!!!

 

[Peter]

Well ...

Experience has shown that the extra time and effort spent to make something idiot-proof is entirely wasted.

That's because while you're improving your product, the idiots will have improved at least as much.

Just peek into any large DIY retailer's returned-casualties bin. Don't ever believe you've seen everything.

 

[kpb]

Peter is definitely right on that one. Having done tech support for a variety of companies has definitely taught me that there is always a bigger idiot out there. No matter how idiot proof you make it they will always find some way to screw it up. As far as the exposed core the doesn't bother as much as the amazing amount of pressure that you need to apply to clip current fans on amd chips. I haven't done any p4 systems personally but I have to beleive there's gotta be a better way to handle attaching the heatsink-fan than they currently use. It just seems like the system wasn't designed with any where near the power disipation that they currently have in mind. Can't really change either tho with out causing problems and require new mb or heat sink designs.

 

[AEB]

Originally posted by: Peter
Athlons running hotter than P4? That's a myth, dude. P4 power consumption is no lower than Athlon's. Not at all.

i dont base it on power consuption i base it that intels can run on stock air and AMD;s have a hard time doing so

 

[Peter]

Sorry, if you don't base a thermal statement on thermal properties, then you need a math adjustment. Urgently.

And what's "stock air" anyway? Use an appropriate cooling solution, mount it correctly, and you're in. And it's not like P4 cooling solutions are any smaller or less complicated than those for Athlons - this is because the thermal output of these CPUs is about the same, with AMD currently even a bit lower.

 

[CTho9305]

Originally posted by: AEB

Originally posted by: Peter
Athlons running hotter than P4? That's a myth, dude. P4 power consumption is no lower than Athlon's. Not at all.

i dont base it on power consuption i base it that intels can run on stock air and AMD;s have a hard time doing so

Chips don't have a "hard time" running with given cooling. They are either 100% stable, or defective. If you are having problems, maybe your case is defective and not providing the manufacturer-recommended airflow. Don't bash a product as a whole because you are using it wrong (or haven't ever used it and just like to spread FUD - fear, uncertainty, and doubt).

If you want to claim that there is something wrong with AMD's retail cooling setup, please provide facts.

 

[capybara]

Originally posted by: kpb
....... As far as the exposed core the doesn't bother as much as the amazing amount of pressure that you need to apply to clip current fans on amd chips. I haven't done any p4 systems personally but I have to beleive there's gotta be a better way to handle attaching the heatsink-fan than they currently use. It just seems like the system wasn't designed with any where near the power disipation that they currently have in mind. Can't really change either tho with out causing problems and require new mb or heat sink designs.

well, Xactly right! i was once putting an "amazing amount of pressure" on the clip attaching a hsf to an athlon, when i felt i grinding feeling between the hsf and the exposed core.
and then an intel user mentioned, "intels dont have exposed cores" and i thought,
wow, why hasnt amd thought of that! duh!

 

[Peter]

Actually, the current top-of-the-line Athlons have LOWER thermals than the highest speed grade of the original socket-A Athlon.

If you want to see questionable attachment design, look at a P4 board after the heatsink has been mounted. From the side. Look at how much the board bends.

 

[Smilin]

Originally posted by: Peter
Actually, the current top-of-the-line Athlons have LOWER thermals than the highest speed grade of the original socket-A Athlon.

If you want to see questionable attachment design, look at a P4 board after the heatsink has been mounted. From the side. Look at how much the board bends.

That's a no-Sh1tter there.

 

[thraxes]

Originally posted by: capybara

Originally posted by: kpb
....... As far as the exposed core the doesn't bother as much as the amazing amount of pressure that you need to apply to clip current fans on amd chips. I haven't done any p4 systems personally but I have to beleive there's gotta be a better way to handle attaching the heatsink-fan than they currently use. It just seems like the system wasn't designed with any where near the power disipation that they currently have in mind. Can't really change either tho with out causing problems and require new mb or heat sink designs.

well, Xactly right! i was once putting an "amazing amount of pressure" on the clip attaching a hsf to an athlon, when i felt i grinding feeling between the hsf and the exposed core.
and then an intel user mentioned, "intels dont have exposed cores" and i thought,
wow, why hasnt amd thought of that! duh!

The actual design of the socket and its cooling capabilities still dates back to the original socket A athlon. At that time the packaging technology used by AMD and Intel with their P3 were almost the same. Both had an exposed DIE and relied on tiny plastic clips on the side of the socket for attaching the HSF. Intel has since brought out a new Generation of CPUs called P4 with an infrastructure (HSF mounting, power supply) designed fresh from the ground up, whereas AMD has simply optimized their older design because they lack the ressources that Intel has. Now with their own next Gen CPU AMD has als redesigned the infrastructure around the CPU. To do so before would make the Athlon chips incompatible with certain mainboards, heatsinks etc... It could be done but it would just be a confusing mess to the consumer.

 

[Webthug]

Originally posted by: CTho9305

Originally posted by: AEB

Originally posted by: Peter
Athlons running hotter than P4? That's a myth, dude. P4 power consumption is no lower than Athlon's. Not at all.

i dont base it on power consuption i base it that intels can run on stock air and AMD;s have a hard time doing so

Chips don't have a "hard time" running with given cooling. They are either 100% stable, or defective. If you are having problems, maybe your case is defective and not providing the manufacturer-recommended airflow. Don't bash a product as a whole because you are using it wrong (or haven't ever used it and just like to spread FUD - fear, uncertainty, and doubt).

If you want to claim that there is something wrong with AMD's retail cooling setup, please provide facts.

having built many Athlon and P4 systems the only improvement i could dee would to be to place the metal strip that attaches the HSF to the Mobo outside the heatsink. This is primarily because there is a patch under the metal strip that recieves less cooling, and where the strip touches the heatsink there is only limited cooling through the metal strip. This design also comprimises the surface area of the heatsink, making it less effective. With the P3's having a similar HSF design to th Athlon, Intel has solved this problem with the all outsize setup, which i think that AMD needs to follow intel's lead. Whilst the current cooling solutions are adiquare, this design improvement would improve their efficiency not only for the Opteron but alos fot the current athlon XP's

 

[Peter]

Nicely spotted, however the center of the heatsink doesn't receive much airflow anyway since it's in the dead center of the spinning fan above it. So the clamp running through the middle of the heatsink actually doesn't make any mentionable difference at all.