AMD CPUs in ASRock motherboards

Just wanted to say thanks for all the replies - I've downloaded
Speedfan and will give it a try. I will also check the airflow inside
the case and the interplay between the PSU fan and the CPU fan - that
may be part of the problem.

Just one more thing - I'm now confused about the gungy stuff you spread
on the CPU before putting the fan on - thermal wotsit doodah compound.
Some have told me that I've probably used too little and am now getting
poor heat flow up through the heatsink, but in this thread people say
it should be a very thin layer, and even to scrape it off prior to
applying more. Can somebody please talk an idiot through the
application of this stuff? I've been building PCs for ten years now and
I've never built one with an overheating problem. I just unpack the CPU
and fan, squirt on the gunge and put it all together. It usually works
fine. If I ever change the CPU or the fan, I apply another squirt of
stuff just to be on the safe side.

CC
 >> Stay informed about: AMD CPUs in ASRock motherboards 

"George Macdonald" <fammacd=!SPAM^nothanks@tellurian.com> wrote in
message news:8f9ol1ho68se2onpk0klabosq5hfh2sedd@4ax.com...
>
> The noisiest thing in my system is the fan on the 9600XT video card
which
> ocasionally makes quite a buzz on startup but settles down later as
it
> warms up - it's been like that since almost new so I've just left
it.

My video card makes absolutely no noise. None whatever. That's
because my mobo is an Asus K8S-MX with integrated video, so I have no
video card with noisy fan.
 >> Stay informed about: AMD CPUs in ASRock motherboards 

Felger Carbon <fmsfnf DeleteThis @jfoops.net> wrote:
> My video card makes absolutely no noise. None whatever.
> That's because my mobo is an Asus K8S-MX with integrated
> video, so I have no video card with noisy fan.


My deepest condolences. You have shared memory.
1200x1024x32bpp at 72 Hz refresh eats 354 MB/s of
your RAM bandwidth. Plus latency ever x? transfers.

-- Robert
 >> Stay informed about: AMD CPUs in ASRock motherboards 

On Mon, 24 Oct 2005 15:46:51 GMT, "Felger Carbon" <fmsfnf.RemoveThis@jfoops.net>
wrote:

>"George Macdonald" <fammacd=!SPAM^nothanks@tellurian.com> wrote in
>message news:8f9ol1ho68se2onpk0klabosq5hfh2sedd@4ax.com...
>>
>> The noisiest thing in my system is the fan on the 9600XT video card
>which
>> ocasionally makes quite a buzz on startup but settles down later as
>it
>> warms up - it's been like that since almost new so I've just left
>it.
>
>My video card makes absolutely no noise. None whatever. That's
>because my mobo is an Asus K8S-MX with integrated video, so I have no
>video card with noisy fan.

Hmm, funny thing: I opened my case to check about something else and found
the video card fan sitting stationary - won't even respond to a poke or
prod... it's dead.<sigh> I'd just been thinking how the system seemed to
be nice & quiet the past couple of days - good thing I hardly ever play
games.

--
Rgds, George Macdonald
 >> Stay informed about: AMD CPUs in ASRock motherboards 

You only need a thin layer of goop. Use a drop about the size of a
matchhead,spread it with a credit card(or other old plastic card.). Thermal
goop is only supposed to fill in tiny pits in cpu and heatsink.
You might want to clean the old stuff of with nail polish remover. It
evaporates and doesn't leave any residue. Don't use 70% alcohol.

"magnate" <chrisc.RemoveThis@dbass.demon.co.uk> wrote in message
news:1130165382.789415.240370@g49g2000cwa.googlegroups.com...
> Just wanted to say thanks for all the replies - I've downloaded
> Speedfan and will give it a try. I will also check the airflow inside
> the case and the interplay between the PSU fan and the CPU fan - that
> may be part of the problem.
>
> Just one more thing - I'm now confused about the gungy stuff you spread
> on the CPU before putting the fan on - thermal wotsit doodah compound.
> Some have told me that I've probably used too little and am now getting
> poor heat flow up through the heatsink, but in this thread people say
> it should be a very thin layer, and even to scrape it off prior to
> applying more. Can somebody please talk an idiot through the
> application of this stuff? I've been building PCs for ten years now and
> I've never built one with an overheating problem. I just unpack the CPU
> and fan, squirt on the gunge and put it all together. It usually works
> fine. If I ever change the CPU or the fan, I apply another squirt of
> stuff just to be on the safe side.
>
> CC
>
 >> Stay informed about: AMD CPUs in ASRock motherboards 

"Robert Redelmeier" <redelm DeleteThis @ev1.net.invalid> wrote in message
news:Ca97f.6709$tV6.469@newssvr27.news.prodigy.net...
> Felger Carbon <fmsfnf DeleteThis @jfoops.net> wrote:
> > My video card makes absolutely no noise. None whatever.
> > That's because my mobo is an Asus K8S-MX with integrated
> > video, so I have no video card with noisy fan.
>
> My deepest condolences. You have shared memory.
> 1200x1024x32bpp at 72 Hz refresh eats 354 MB/s of
> your RAM bandwidth. Plus latency ever x? transfers.

By golly you're right, Robert! That what, 3% overhead will prevent me
from sending and receiving email, browsing, or reading this newsgroup!
Uhm... so how have I been doing these things just fine all this time?


Not everybody plays games or does nuclear explosion simulations.
 >> Stay informed about: AMD CPUs in ASRock motherboards 

Felger Carbon <fmsfnf RemoveThis @jfoops.net> wrote:
> "Robert Redelmeier" <redelm RemoveThis @ev1.net.invalid> wrote in message
>> My deepest condolences. You have shared memory.
>> 1200x1024x32bpp at 72 Hz refresh eats 354 MB/s of
>> your RAM bandwidth. Plus latency ever x? transfers.
>
> By golly you're right, Robert! That what, 3% overhead will

I believe DDR has a max burst bandwidth of 3200 MB/s, so
it's at least 11%. But the bursts have very limited length,
a cacheline or 4 before there's a whole new latency cycle. (I
think VRAM can burst longer). As a result AFAIK average
bandwdith is around 1 GB/s. Your shared mem eats 35%.

Not something I'd want to hobble a great CPU with. IMHO, you'd
be _far_ better off disabling on board graphics and getting
an old Matrox PCI card, especially if you're only doing 2D.

-- Robert
 >> Stay informed about: AMD CPUs in ASRock motherboards 

On Mon, 24 Oct 2005 20:47:47 +0000, Robert Redelmeier wrote:

> Felger Carbon <fmsfnf DeleteThis @jfoops.net> wrote:
>> "Robert Redelmeier" <redelm DeleteThis @ev1.net.invalid> wrote in message
>>> My deepest condolences. You have shared memory.
>>> 1200x1024x32bpp at 72 Hz refresh eats 354 MB/s of
>>> your RAM bandwidth. Plus latency ever x? transfers.
>>
>> By golly you're right, Robert! That what, 3% overhead will
>
> I believe DDR has a max burst bandwidth of 3200 MB/s, so
> it's at least 11%. But the bursts have very limited length,
> a cacheline or 4 before there's a whole new latency cycle. (I
> think VRAM can burst longer). As a result AFAIK average
> bandwdith is around 1 GB/s. Your shared mem eats 35%.

Why a "whole new" latency cycle? Isn't the page/row still active?

> Not something I'd want to hobble a great CPU with. IMHO, you'd
> be _far_ better off disabling on board graphics and getting
> an old Matrox PCI card, especially if you're only doing 2D.

Felg, I gots some *old* Matrox PCI cards if you want 'em[*]. I use a
"new" AGP Matrox 2D card.

[*] One has been doing dual-display duty for my laptop, which will be
replaced in a week or so, or so I'm told.

--
Keith
 >> Stay informed about: AMD CPUs in ASRock motherboards 

keith <krw.TakeThisOut@att.bizzzz> wrote:
> On Mon, 24 Oct 2005 20:47:47 +0000, Robert Redelmeier wrote:
>> I believe DDR has a max burst bandwidth of 3200 MB/s, so
>> it's at least 11%. But the bursts have very limited length,
>> a cacheline or 4 before there's a whole new latency cycle. (I
>> think VRAM can burst longer). As a result AFAIK average
>> bandwdith is around 1 GB/s. Your shared mem eats 35%.
>
> Why a "whole new" latency cycle? Isn't the page/row still active?

I'm no DRAM expert but the page probably, the row maybe.
AFAIK there are definite limits to how many rows can be burst
out. A double cache line 256 byte burst is 32 rows in row.
I thought this was the supposed advantage of RDRAM, _much_
longer bursts at the cost of latency.

-- Robert
 >> Stay informed about: AMD CPUs in ASRock motherboards 

On Tue, 25 Oct 2005 02:51:07 +0000, Robert Redelmeier wrote:

> keith <krw.TakeThisOut@att.bizzzz> wrote:
>> On Mon, 24 Oct 2005 20:47:47 +0000, Robert Redelmeier wrote:
>>> I believe DDR has a max burst bandwidth of 3200 MB/s, so
>>> it's at least 11%. But the bursts have very limited length,
>>> a cacheline or 4 before there's a whole new latency cycle. (I
>>> think VRAM can burst longer). As a result AFAIK average
>>> bandwdith is around 1 GB/s. Your shared mem eats 35%.
>>
>> Why a "whole new" latency cycle? Isn't the page/row still active?
>
> I'm no DRAM expert but the page probably, the row maybe.
> AFAIK there are definite limits to how many rows can be burst
> out. A double cache line 256 byte burst is 32 rows in row.
> I thought this was the supposed advantage of RDRAM, _much_
> longer bursts at the cost of latency.

I'm no expert either, but I don't see why one would close a page
unnecessarily. Burst <> row or page. The *one* advantage of DRDRAM was
*supposed* to be that multiple (as in many) pages could be kept open. I
don't think SDRAM forces one to have a page open only when it's bing
accessed. Even page-mode DRAM (what, 15 years ago?) didn't require pages
to be closed after each access.

Maybe I'll button-hole Dale if I see him around. He designed the stuff.

--
Keith
 >> Stay informed about: AMD CPUs in ASRock motherboards 

On Mon, 24 Oct 2005 20:47:47 GMT, Robert Redelmeier
<redelm RemoveThis @ev1.net.invalid> wrote:

>Felger Carbon <fmsfnf RemoveThis @jfoops.net> wrote:
>> "Robert Redelmeier" <redelm RemoveThis @ev1.net.invalid> wrote in message
>>> My deepest condolences. You have shared memory.
>>> 1200x1024x32bpp at 72 Hz refresh eats 354 MB/s of
>>> your RAM bandwidth. Plus latency ever x? transfers.
>>
>> By golly you're right, Robert! That what, 3% overhead will
>
>I believe DDR has a max burst bandwidth of 3200 MB/s, so
>it's at least 11%. But the bursts have very limited length,
>a cacheline or 4 before there's a whole new latency cycle. (I
>think VRAM can burst longer). As a result AFAIK average
>bandwdith is around 1 GB/s. Your shared mem eats 35%.

I dunno about that mbrd but the ATI chipset, has a feature they call
HyperMemory -- kinda like the TurboCache of low-end nVidia 6200 cards --
where you can have up to 128MB of memory hanging directly off the
integrated chipset.

--
Rgds, George Macdonald
 >> Stay informed about: AMD CPUs in ASRock motherboards 

On Mon, 24 Oct 2005 20:47:47 GMT, Robert Redelmeier
<redelm DeleteThis @ev1.net.invalid> wrote:

>Felger Carbon <fmsfnf DeleteThis @jfoops.net> wrote:
>> "Robert Redelmeier" <redelm DeleteThis @ev1.net.invalid> wrote in message
>>> My deepest condolences. You have shared memory.
>>> 1200x1024x32bpp at 72 Hz refresh eats 354 MB/s of
>>> your RAM bandwidth. Plus latency ever x? transfers.
>>
>> By golly you're right, Robert! That what, 3% overhead will
>
>I believe DDR has a max burst bandwidth of 3200 MB/s, so
>it's at least 11%. But the bursts have very limited length,
>a cacheline or 4 before there's a whole new latency cycle. (I
>think VRAM can burst longer). As a result AFAIK average
>bandwdith is around 1 GB/s. Your shared mem eats 35%.

Felger actually high-balled the amount of overall performance it costs
you though, usually you're looking at less than a 1% loss in
performance for most office tasks. Honest.

There's a good reason why something like 60% of all PCs use integrated
graphics these days, it just doesn't make a difference for what the
vast majority of us do.

-------------
Tony Hill
hilla <underscore> 20 <at> yahoo <dot> ca
 >> Stay informed about: AMD CPUs in ASRock motherboards 

On Tue, 25 Oct 2005 22:02:18 -0400, keith <krw.RemoveThis@att.bizzzz> wrote:

>On Tue, 25 Oct 2005 21:44:44 +0000, Robert Redelmeier wrote:
>
>> keith <krw.RemoveThis@att.bizzzz> wrote:
>>> I'm no expert either, but I don't see why one would close a
>>> page unnecessarily. Burst <> row or page. The *one* advantage
>>> of DRDRAM was *supposed* to be that multiple (as in many) pages
>>> could be kept open. I don't think SDRAM forces one to have a
>>> page open only when it's bing accessed. Even page-mode DRAM
>>> (what, 15 years ago?) didn't require pages to be closed after
>>> each access.
>>
>> No, but there are delays (smaller on same row) and burst
>> length seems to be limited to 8:
>>
>Ok, but the graphics card need only open a new page when it's crosses the
>line. I believe SDRAM can keep a couple of pages open too, but I'll
>have to try to find D.

All current DDR-DRAM chips have four banks which can all be kept open
simultaneously. Memory Controllers vary on how open pages on the banks are
handled: Intel has traditionally (since the 440-BX) allowed all banks on
all ranks (up to 32) to be kept open, with an idle cycle counter for
precharge; VIA used to precharge when jumping rank to rank. Dunno exactly
what AMD's does - I can't find a limit though there is a mention of a
maximum in the docs. It ain't easy to figure out with all the different
access modes... now in latest revisions we have "bank swizzling" (each
device bank bit is XORed from 3 different address bits) on top of Chip
Select Interleaving; they do have an idle cycle limit counter and for those
things to work they have to keep banks open across rank accesses, up to
some limit?

BTW Robert seemed to suggest a cache line of 128-Bytes but on both Intel &
AMD systems the maximum burst is 64-Bytes, i.e. 4-beat burst on dual
channel systems. The P4 L2 cache line *is* 128Bytes but each line is two
sectors of 64Bytes.

My personal feeling here is that the integrated video is not that big a
deal, though it would seem wasteful to have the video refresh cycles
hitting main memory, especially when a couple of 256Mb chips hanging
directly off the chipset chip could be put to good use there.

--
Rgds, George Macdonald
 >> Stay informed about: AMD CPUs in ASRock motherboards