E4300 and 650i overclocking

'~misfit~' wrote:
| Ahh, OK. having done some research I now know that diamonds are *very*
| thermally conductive. I only hope that your paste has a high percentage of
| diamonds to carrier. Maybe you could heat it and hope that the diamonds
| settle? Probably not at that size.
|
| Anyway, good luck. :-)
Glad to hear from you. Having tried butter as a thermal grease, I guess I'm
willing to try anything B^) Diamond is extraordinarily conductive for heat
(also, unfortunately, diamond is also very conductive electrically.) If I
remember correctly, there are two axes for thermal conduction, the less
conductive axis is about twice as conductive as copper and the more
conductive axis is about seven times as conductive as copper. With random
orientation that is still going to be at least twice as conductive as silver
(or so I suppose.) Diamonds for machining have really dropped in price; at
McMaster-Carr (which is NOT a low cost tool and material supply house), a
one carat diamond, mounted for use in shaping surface grinder wheels, is
less than $65 US. The fine stuff for lapping should be even cheaper, so
five grams should have quite a bit of diamond powder. I'd guess that the
major expense is separating the particles by size. The finest grade is
likely TOO fine. But for $7 US, I'm willing to try it. Graphite didn't
work out well for me, but maybe diamond will show promise. Then there's the
possibility that the oil based carrier will be to thin. I've just looked at
the Intel boxed retail E4300, and it comes with thermal compound already
applied to the heatsink. Evidently these Core 2 Duos have MUCH flatter and
smoother surfaces than past CPUs. The amount of compound is so small that
not only will the coverage be translucent, it will be darn near invisible.

I do plan to test the 'diamond grease' in a jig before I consider using it
on a CPU.

The motherboard is so densely packed that anything short of chilling the
entire case volume is going to be tricky, especially since there are FOUR
heatsinks in addition to the CPU heatsink; two chipset heatsinks connected
by a heat pipe and using a fan and two DC-DC voltage converter-regulator
strip heatsinks without fans. And then there are the memory heatsink. All
will need adequate airflow.

So far I have all the pieces except for the two memory modules. I order
everything else from ZipFlyZoom.com (at the same time as I order the memory
modules from newEgg (ZipFlyZoom was backordered.) Both orders placed about
2 PM Wednesday. The ZipFlyZoom order arrived about 2 PM Friday via 2nd day
FedEx - 13# for ~ $10 US. The newEgg order is scheduled to arrive Monday
via 2 day UPS - 6 ounces for ~ $4 US.

I agreed with Ed Medlin about the power requirements and picked up an Antec
550 Neo HE ( three 12 VDC rails, 18 Amps each) at a CompUSA going out of
business sale.

If this turns into a blog, just shoot me ... please.

Phil Weldon

"~misfit~" wrote in message
...
| ~misfit~ wrote:
| Phil Weldon wrote:
|
|
| [snip]
|
| By the way, I'm going to try out diamond dust in an oil soluble base
| for a thermal compound; it's cheaper than 'Arctic Silver'!
| Five grams of 0-2 micron diamond particles for $7.34 US. Actually
| larger particles would likely be better; that's a bit more
| expensive; ~ $20 US for five grams.
| McMaster-Carr at http://www.mcmaster.com
|
| S'cuse my ignorance Phil but why the diamond lapping paste as a TIM?
| It seems a rather strange thing to use to me. Does it have properties
| that I'm not aware of? Are you sure that the paste won't thin out and
| run away at temperature leaving a few "rocks" and air gaps in it's
| place?
| Is it just for bragging rights? ("You've got silver? Loser. Mine is
| diamond." g)
|
| (BTW, I'm aware of your view on commercial TIMs).
|
| Cheers, and good luck with the system, I'll be following your
| progress with interest.
|
| Ahh, OK. having done some research I now know that diamonds are *very*
| thermally conductive. I only hope that your paste has a high percentage of
| diamonds to carrier. Maybe you could heat it and hope that the diamonds
| settle? Probably not at that size.
|
| Anyway, good luck. :-)
| --
| Shaun.
|
|

No worries about that.
Those of us watching this thread can't wait for the results.
Keep 'em coming.
Amir


"Phil Weldon" wrote in message
ink.net...
'~misfit~' wrote:
| Ahh, OK. having done some research I now know that diamonds are *very*
| thermally conductive. I only hope that your paste has a high percentage
of
| diamonds to carrier. Maybe you could heat it and hope that the diamonds
| settle? Probably not at that size.
|
| Anyway, good luck. :-)
Glad to hear from you. Having tried butter as a thermal grease, I guess
I'm
willing to try anything B^) Diamond is extraordinarily conductive for
heat
(also, unfortunately, diamond is also very conductive electrically.) If I
remember correctly, there are two axes for thermal conduction, the less
conductive axis is about twice as conductive as copper and the more
conductive axis is about seven times as conductive as copper. With random
orientation that is still going to be at least twice as conductive as
silver
(or so I suppose.) Diamonds for machining have really dropped in price;
at
McMaster-Carr (which is NOT a low cost tool and material supply house), a
one carat diamond, mounted for use in shaping surface grinder wheels, is
less than $65 US. The fine stuff for lapping should be even cheaper, so
five grams should have quite a bit of diamond powder. I'd guess that the
major expense is separating the particles by size. The finest grade is
likely TOO fine. But for $7 US, I'm willing to try it. Graphite didn't
work out well for me, but maybe diamond will show promise. Then there's
the
possibility that the oil based carrier will be to thin. I've just looked
at
the Intel boxed retail E4300, and it comes with thermal compound already
applied to the heatsink. Evidently these Core 2 Duos have MUCH flatter
and
smoother surfaces than past CPUs. The amount of compound is so small that
not only will the coverage be translucent, it will be darn near invisible.

I do plan to test the 'diamond grease' in a jig before I consider using it
on a CPU.

The motherboard is so densely packed that anything short of chilling the
entire case volume is going to be tricky, especially since there are FOUR
heatsinks in addition to the CPU heatsink; two chipset heatsinks connected
by a heat pipe and using a fan and two DC-DC voltage converter-regulator
strip heatsinks without fans. And then there are the memory heatsink.
All
will need adequate airflow.

So far I have all the pieces except for the two memory modules. I order
everything else from ZipFlyZoom.com (at the same time as I order the
memory
modules from newEgg (ZipFlyZoom was backordered.) Both orders placed
about
2 PM Wednesday. The ZipFlyZoom order arrived about 2 PM Friday via 2nd
day
FedEx - 13# for ~ $10 US. The newEgg order is scheduled to arrive Monday
via 2 day UPS - 6 ounces for ~ $4 US.

I agreed with Ed Medlin about the power requirements and picked up an
Antec
550 Neo HE ( three 12 VDC rails, 18 Amps each) at a CompUSA going out of
business sale.

If this turns into a blog, just shoot me ... please.

Phil Weldon

"~misfit~" wrote in message
...
| ~misfit~ wrote:
| Phil Weldon wrote:
|
|
| [snip]
|
| By the way, I'm going to try out diamond dust in an oil soluble base
| for a thermal compound; it's cheaper than 'Arctic Silver'!
| Five grams of 0-2 micron diamond particles for $7.34 US. Actually
| larger particles would likely be better; that's a bit more
| expensive; ~ $20 US for five grams.
| McMaster-Carr at http://www.mcmaster.com
|
| S'cuse my ignorance Phil but why the diamond lapping paste as a TIM?
| It seems a rather strange thing to use to me. Does it have properties
| that I'm not aware of? Are you sure that the paste won't thin out and
| run away at temperature leaving a few "rocks" and air gaps in it's
| place?
| Is it just for bragging rights? ("You've got silver? Loser. Mine is
| diamond." g)
|
| (BTW, I'm aware of your view on commercial TIMs).
|
| Cheers, and good luck with the system, I'll be following your
| progress with interest.
|
| Ahh, OK. having done some research I now know that diamonds are *very*
| thermally conductive. I only hope that your paste has a high percentage
of
| diamonds to carrier. Maybe you could heat it and hope that the diamonds
| settle? Probably not at that size.
|
| Anyway, good luck. :-)
| --
| Shaun.
|
|

Phil Weldon wrote:
I laid my money down, and did pretty well price-wise
($1000 US for the five core components).

Having waited this long, I don't know why you didn't just
wait the extra month for the Core 2 Duo price drops.

'Amir Facade' wrote:
| No worries about that.
| Those of us watching this thread can't wait for the results.
| Keep 'em coming.
_____

Update, Day 2.

Manuals.

* It turns out to be a very good thing that a manual for the EVGA 680i
motherboard is easily obtainable at the EVGA website. No manual is included
in the box, just a 8" X 10" 'quick install' card. Perhaps a manual in PDF
format is somewhere on the driver disk (though it is not mentioned.)

* The EVGA nVidia 8800 GTS card comes with a nice slick manual;
unfortunately is for AGP cards using Windows 98/ME/2000. Again, perhaps
there is a PDF manual on the driver disk, though it is not mentioned.

Power connections:

* Motherboard power connector are multiplying. The EVGA 680i has eleven:
1 X 24 pin ATX connector
1 X 8 pin +12 VDC CPU power connector
1 X 4 pin CPU fan connector
1 X 3 pin chipset fan connector
6 X 3 pin fan connectors
1 X 4 pin Molex connector (PCI-E aux power?)
* Motherboard rear I/O connectors are multiplying (18)
* Motherboard on-board I/0 connectors are multiplying ( ~ 20)

So far the layout looks good, though quite crowded. I can see right away
that many after market coolers have no chance of fitting unless the main
chipset cooler is replaced. But WHY did they decide to provide machine
screws as small as 1.5 mm ( .06" ) to attach the chipset fan? One heavy
sigh as you try to start the screws and they're gone with the wind.

I think I'll do extensive measurements for clearances before beginning the
assembly. And change the subject line to 'E4300 and 680i overclocking' to
reflect the chipset change.

Phil Weldon





"Amir Facade" wrote in message
news | No worries about that.
| Those of us watching this thread can't wait for the results.
| Keep 'em coming.
| Amir
|
|
| "Phil Weldon" wrote in message
| ink.net...
| '~misfit~' wrote:
| | Ahh, OK. having done some research I now know that diamonds are *very*
| | thermally conductive. I only hope that your paste has a high
percentage
| of
| | diamonds to carrier. Maybe you could heat it and hope that the
diamonds
| | settle? Probably not at that size.
| |
| | Anyway, good luck. :-)
| Glad to hear from you. Having tried butter as a thermal grease, I guess
| I'm
| willing to try anything B^) Diamond is extraordinarily conductive for
| heat
| (also, unfortunately, diamond is also very conductive electrically.) If
I
| remember correctly, there are two axes for thermal conduction, the less
| conductive axis is about twice as conductive as copper and the more
| conductive axis is about seven times as conductive as copper. With
random
| orientation that is still going to be at least twice as conductive as
| silver
| (or so I suppose.) Diamonds for machining have really dropped in price;
| at
| McMaster-Carr (which is NOT a low cost tool and material supply house),
a
| one carat diamond, mounted for use in shaping surface grinder wheels, is
| less than $65 US. The fine stuff for lapping should be even cheaper, so
| five grams should have quite a bit of diamond powder. I'd guess that
the
| major expense is separating the particles by size. The finest grade is
| likely TOO fine. But for $7 US, I'm willing to try it. Graphite
didn't
| work out well for me, but maybe diamond will show promise. Then there's
| the
| possibility that the oil based carrier will be to thin. I've just
looked
| at
| the Intel boxed retail E4300, and it comes with thermal compound already
| applied to the heatsink. Evidently these Core 2 Duos have MUCH flatter
| and
| smoother surfaces than past CPUs. The amount of compound is so small
that
| not only will the coverage be translucent, it will be darn near
invisible.
|
| I do plan to test the 'diamond grease' in a jig before I consider using
it
| on a CPU.
|
| The motherboard is so densely packed that anything short of chilling the
| entire case volume is going to be tricky, especially since there are
FOUR
| heatsinks in addition to the CPU heatsink; two chipset heatsinks
connected
| by a heat pipe and using a fan and two DC-DC voltage converter-regulator
| strip heatsinks without fans. And then there are the memory heatsink.
| All
| will need adequate airflow.
|
| So far I have all the pieces except for the two memory modules. I order
| everything else from ZipFlyZoom.com (at the same time as I order the
| memory
| modules from newEgg (ZipFlyZoom was backordered.) Both orders placed
| about
| 2 PM Wednesday. The ZipFlyZoom order arrived about 2 PM Friday via 2nd
| day
| FedEx - 13# for ~ $10 US. The newEgg order is scheduled to arrive
Monday
| via 2 day UPS - 6 ounces for ~ $4 US.
|
| I agreed with Ed Medlin about the power requirements and picked up an
| Antec
| 550 Neo HE ( three 12 VDC rails, 18 Amps each) at a CompUSA going out of
| business sale.
|
| If this turns into a blog, just shoot me ... please.
|
| Phil Weldon
|
| "~misfit~" wrote in message
| ...
| | ~misfit~ wrote:
| | Phil Weldon wrote:
| |
| |
| | [snip]
| |
| | By the way, I'm going to try out diamond dust in an oil soluble
base
| | for a thermal compound; it's cheaper than 'Arctic Silver'!
| | Five grams of 0-2 micron diamond particles for $7.34 US. Actually
| | larger particles would likely be better; that's a bit more
| | expensive; ~ $20 US for five grams.
| | McMaster-Carr at http://www.mcmaster.com
| |
| | S'cuse my ignorance Phil but why the diamond lapping paste as a TIM?
| | It seems a rather strange thing to use to me. Does it have
properties
| | that I'm not aware of? Are you sure that the paste won't thin out
and
| | run away at temperature leaving a few "rocks" and air gaps in it's
| | place?
| | Is it just for bragging rights? ("You've got silver? Loser. Mine is
| | diamond." g)
| |
| | (BTW, I'm aware of your view on commercial TIMs).
| |
| | Cheers, and good luck with the system, I'll be following your
| | progress with interest.
| |
| | Ahh, OK. having done some research I now know that diamonds are *very*
| | thermally conductive. I only hope that your paste has a high
percentage
| of
| | diamonds to carrier. Maybe you could heat it and hope that the
diamonds
| | settle? Probably not at that size.
| |
| | Anyway, good luck. :-)
| | --
| | Shaun.
| |
| |
|
|
|
|

'Fishface' wrote:
| Having waited this long, I don't know why you didn't just
| wait the extra month for the Core 2 Duo price drops.
_____

Because I did not pick the Core 2 Duo E4300 for price. Because the CPU
price is now a small fraction of the display
adapter/motherboard/memory/power supply cost. Because a 65 nm quad won't
show me anything that the E4300 won't.

Phil Weldon

"Fishface" ? wrote in message
news:0ZwPh.673$i93.287@trnddc05...
| Phil Weldon wrote:
| I laid my money down, and did pretty well price-wise
| ($1000 US for the five core components).
|
| Having waited this long, I don't know why you didn't just
| wait the extra month for the Core 2 Duo price drops.
|
|

"Phil Weldon" wrote in message
thlink.net...
'Fishface' wrote:
| Having waited this long, I don't know why you didn't just
| wait the extra month for the Core 2 Duo price drops.
_____

Because I did not pick the Core 2 Duo E4300 for price. Because the CPU
price is now a small fraction of the display
adapter/motherboard/memory/power supply cost. Because a 65 nm quad won't
show me anything that the E4300 won't.

Phil Weldon

I do the same. If I wait until the price drops, something new has been
released that interests me and so on. The price on the 4300s began so low
compared to previous offerings, so I doubt that there will be a huge
difference when considered in the whole scheme of things.

Ed

| I do the same. If I wait until the price drops, something new has been
| released that interests me and so on. The price on the 4300s began so low
| compared to previous offerings, so I doubt that there will be a huge
| difference when considered in the whole scheme of things.
_____

Let's see, mmm... I have the time do build it this week, but otherwise not
for the next month ... is that worth the price of a movie, parking, coke and
popcorn to me ... mmm. Yeah.

Phil Weldon

"Ed Medlin" wrote in message
news |
| "Phil Weldon" wrote in message
| thlink.net...
| 'Fishface' wrote:
| | Having waited this long, I don't know why you didn't just
| | wait the extra month for the Core 2 Duo price drops.
| _____
|
| Because I did not pick the Core 2 Duo E4300 for price. Because the CPU
| price is now a small fraction of the display
| adapter/motherboard/memory/power supply cost. Because a 65 nm quad
won't
| show me anything that the E4300 won't.
|
| Phil Weldon
|
| I do the same. If I wait until the price drops, something new has been
| released that interests me and so on. The price on the 4300s began so low
| compared to previous offerings, so I doubt that there will be a huge
| difference when considered in the whole scheme of things.
|
| Ed
|
|

"Phil Weldon" wrote in message
thlink.net...
| I do the same. If I wait until the price drops, something new has been
| released that interests me and so on. The price on the 4300s began so
low
| compared to previous offerings, so I doubt that there will be a huge
| difference when considered in the whole scheme of things.
_____

Let's see, mmm... I have the time do build it this week, but otherwise
not
for the next month ... is that worth the price of a movie, parking, coke
and
popcorn to me ... mmm. Yeah.

Phil Weldon

Lets see........Wife and I are going to the Corrales vs Clottey boxing match
next Sat. Since it is a Showtime fight, tickets and parking costs are close
to an E6600........LOL..........BTW, I think I am going to go the E6600
route in a couple weeks probably starting with air cooling and then water.
Maybe we can compare some benches and see how it all works out. Something
tells me that you might win out if only in the cooling area. It will be an
interesting project anyway.

Ed

'Ed Medlin' wrote:
| Lets see........Wife and I are going to the Corrales vs Clottey boxing
match
| next Sat. Since it is a Showtime fight, tickets and parking costs are
close
| to an E6600........LOL..........BTW, I think I am going to go the E6600
| route in a couple weeks probably starting with air cooling and then water.
| Maybe we can compare some benches and see how it all works out. Something
| tells me that you might win out if only in the cooling area. It will be an
| interesting project anyway.
_____

It'd be great to learn from each other in this experiment. I'm looking
forward to it. Perhaps others will join in.

Hmmm .... let's see, the championship game is here in Atlanta tonight; I
could get a ticket, buy a QX6800 and two 8800 GTX, or watch the game on TV
while I start putting my new system together. (checks wallet)
Guess I'll be watching TV and nibbling fan holes.

My memory arrived a few minutes ago, so everything is in hand for the first
air-cooled iteration.

One thing I hadn't considered before. The 775 socket system is not designed
for more than a few dozen insert/removal cycles. I am putting a label with
check boxes on the system case to keep track.

*****
Details about the EVGA 680i SLI 775 Motherboard

At first sight my new EVGA 680i motherboard seems a solid reference design
implementation.

As best I can tell, EVGA, at present, makes two different 680i models, the
680i SLI 775 and the 680i SLI 775 LT. The 680i LT is less expensive and
drops certain features like the third graphics PCI-E slot (X16 physically,
X8 electrically.)

The nomenclature is a bit confusing. The 680i SLI 775 model comes in EIGHT
different packages; the motherboard is identical in each package (save for a
small modification in the versions ending in '1')
122-CK-NF63-A1
122-CK-NF63-AR
122-CK-NF63-T1
122-CK-NF63-TR
122-CK-NF68-A1
122-CK-NF68-AR
122-CK-NF68-T1
122-CK-NF68-TR.

The four with 63 in the middle are for sale as part of a promotional package
with a Core 2 Duo CPU.
The A on the end indicates a package with a lifetime warranty plus extra
cables and brackets.
The T on the end indicates a package with a 2 year warranty and fewer cables
and brackets.
The 1 on the end indicates a modified motherboard that fixes a limitation on
the FSB speed when overclocking a quad core CPU (free replacement RMA for TR
and AR.)

The manual is the same for all eight packages.
The motherboard in each package has the following connectors
Rear I/O panel data
6 X USB
2 X RJ45 ( 2 X Gigabit LAN)
1 X PS2 keyboard
1 X PS2 mouse
1 X IEEE1394 (Firewire)
6 X stereo 1/8" phone jacks
1 X SPDIF
Mainboard data
2 X PCI-E X16 SLI graphics slots
1 X PCI-E X8 (X16 physically) third graphics slot
2 X PCI-E X1
2 X PCI
1 X Floppy
1 X Parallel ATA
6 X Serial ATA (with Raid 0,1,0/1, 5)
4 X USB header
1 X RS232 header
1 X Front Panel Audio header
1 X IEEE 1394a header

Mainboard fan
1 X 4-pin CPU fan
1 X 3-pin Chipset fan
1 X 3-pin VR fan
1 X 3-pin System fan
2 X 3-pin Chassis fan
1 X 3-pin Aux fan

Miscellaneous
1 X 2-pin header silk screened 'Intruder'
1 X 2-pin header location, no header installed, silk screened 'TPM
Select' / 0* Protect / 1 Access /
1 X 16-pin header location, no header installed, silk screened 'LPC
Header', labeled in manual as 'Debug'


Header, cable, bracket with 4 X USB
Header, cable, bracket with 1 X IEEE1394
Header, cable, bracket with 1 X RS232.

The 'T' motherboard package includes the following cable / connectors
1 X Floppy
1 X Parallel 80 conductor cable
2 X Serial ATA data cables
1 X 4-pin in-line Molex to Serial ATA power connector Y adapter
1 X SLI bridge (pc board)

The 'A' motherboard package ADDS
2 X Serial ATA data cables
1 X header/cable/bracket w/ 4 USB sockets
1 X header/cable/bracket w/ 1 RS232 socket.

I haven't followed boxing in a long time; I think my earliest detailed
memory of a match is listening on the radio to the 1959 Ingmar
Johannson/Floyd Patterson fight in Stockholm. Earlier than that it's all
"Look Sharp, Feel Sharp, Be Sharp ... Buy Gillette".

Phil Weldon


"Ed Medlin" wrote in message
. ..
|
| "Phil Weldon" wrote in message
| thlink.net...
| | I do the same. If I wait until the price drops, something new has been
| | released that interests me and so on. The price on the 4300s began so
| low
| | compared to previous offerings, so I doubt that there will be a huge
| | difference when considered in the whole scheme of things.
| _____
|
| Let's see, mmm... I have the time do build it this week, but otherwise
| not
| for the next month ... is that worth the price of a movie, parking, coke
| and
| popcorn to me ... mmm. Yeah.
|
| Phil Weldon
|
| Lets see........Wife and I are going to the Corrales vs Clottey boxing
match
| next Sat. Since it is a Showtime fight, tickets and parking costs are
close
| to an E6600........LOL..........BTW, I think I am going to go the E6600
| route in a couple weeks probably starting with air cooling and then water.
| Maybe we can compare some benches and see how it all works out. Something
| tells me that you might win out if only in the cooling area. It will be an
| interesting project anyway.
|
| Ed
|
|

Phil Weldon wrote:
'~misfit~' wrote:
Ahh, OK. having done some research I now know that diamonds are
*very* thermally conductive. I only hope that your paste has a high
percentage of diamonds to carrier. Maybe you could heat it and hope
that the diamonds settle? Probably not at that size.

Anyway, good luck. :-)
Glad to hear from you.

Thanks. I check this group infrequently these days but always read your
posts. :-)

Having tried butter as a thermal grease, I
guess I'm willing to try anything B^)

Hehee! Yeah, I'm familiar with your TIM experiments.

Diamond is extraordinarily
conductive for heat (also, unfortunately, diamond is also very
conductive electrically.) If I remember correctly, there are two
axes for thermal conduction, the less conductive axis is about twice
as conductive as copper and the more conductive axis is about seven
times as conductive as copper. With random orientation that is still
going to be at least twice as conductive as silver (or so I suppose.)
Diamonds for machining have really dropped in price; at McMaster-Carr
(which is NOT a low cost tool and material supply house), a one carat
diamond, mounted for use in shaping surface grinder wheels, is less
than $65 US. The fine stuff for lapping should be even cheaper, so
five grams should have quite a bit of diamond powder. I'd guess that
the major expense is separating the particles by size. The finest
grade is likely TOO fine. But for $7 US, I'm willing to try it.
Graphite didn't work out well for me, but maybe diamond will show
promise. Then there's the possibility that the oil based carrier
will be to thin. I've just looked at the Intel boxed retail E4300,
and it comes with thermal compound already applied to the heatsink.
Evidently these Core 2 Duos have MUCH flatter and smoother surfaces
than past CPUs. The amount of compound is so small that not only
will the coverage be translucent, it will be darn near invisible.

I do plan to test the 'diamond grease' in a jig before I consider
using it on a CPU.

Wise move. I hope it pans out for you, fortune favours the brave they say.
:-)

The motherboard is so densely packed that anything short of chilling
the entire case volume is going to be tricky, especially since there
are FOUR heatsinks in addition to the CPU heatsink; two chipset
heatsinks connected by a heat pipe and using a fan and two DC-DC
voltage converter-regulator strip heatsinks without fans. And then
there are the memory heatsink. All will need adequate airflow.

Yeah, with the amount of heat being produced in a modern, 'power-spec' PC,
you need the equivalent of a hurricane passing through your case.

I agreed with Ed Medlin about the power requirements and picked up an
Antec 550 Neo HE ( three 12 VDC rails, 18 Amps each) at a CompUSA
going out of business sale.

Nice!

If this turns into a blog, just shoot me ... please.

Hehee! I thought I was the only one who thinks blogs are for perverse
extroverts.

Please, keep us informed of your progress. It's good to hear from someone
who knows what they're doing rather than dreamers or rich-kid guess-and-hope
guys. g

BTW, for some reason I LOL'ed when I read you say this in another post:

"My memory arrived a few minutes ago...."

Just struck me as funny. shrug

Good luck with the build.

Regards,
--
Shaun.