another quad (on the way)

  • Thread starter Thread starter Flasherly
  • Start date Start date
F

Flasherly

(Best I can figure on the scenario below is I'll take a hit on some
bus-architecture speeds. Ideal, crossed-fingers scenario...).

Thought I was stuck to a dualcore, but. . .things can happen.

AMD Phenom X4 9550 2.2 GHz Quad-Core HD9550WCJ4BGH CPU Processor

sent off for one (used, fair but a little more than Intel quad I'm
running at same 2.2Ghz on another MB, both Gigabytes built pretty
close together;- still too pricey for 65watt variants, or, for getting
into 3Ghz core speeds. Reckon I do like getting CPUs, nice ones, for
$20 or under.)

As is, running a dualcore AMD Athlon 64 X2 4200+ (second update,
this'll be -better be- to a third).

6watts difference between the two (89/95watts on the Phenom). Can't
especially see myself going thru what (at bottom of this post) another
person experienced for a measly 6watts;- besides, that might be a
125watt Phenom that he's dealing with.

The MB...

GA-M61PME-S2
CPU Socket Type AM2
Athlon 64 X2 / Athlon 64 FX / Athlon 64 / Sempron

which is a little strange, as there was a GA-M61PME-S2 P specifically
for a quad (along with, probably, a broader range of AMD offerings,
too). Apparently two revisions to each of these two boards, in a
redundant sort of manufacturing manner of keeping abreast with the AMD
2 to 3 socket transition types.

Mine, per se, isn't a socket 2+ or 3, however, Gigabyte does
specifically state (as a stipend to my model): "(Note) If you install
AMD AM2+ CPU on AM2 motherboard, the system bus speed will downgrade
from HT3.0(5200MHz) to HT1.0(2000 MT/s) spec; however, the frequency
of AM2+ CPU will not be impacted. Please refer "CPU Support List" for
more information."

and, on the quoted listing, there's an provision for the Phenom I got
off Gigabyte's support page.

Guess I'll doublecheck my BIOS revision and wait for it to come in,
see what happens. (Also snoop around, see if I can't cut/grind down
some HS material and fashion an augmented chipset cooler. *see notes.
Last I poked around with an IR temp sensor I couldn't find the sucker,
although it most definitely will climb up to 135F during summertime.)

-
couple notes noted from roundabout the time I bought it (Newegg cust
reviews) ...

For a dollar more, you can get the GA-M61PME-S2P.
The only difference I notice, and the only reason I gave it four eggs,
is that the CPU socket on that board is compatible both with AM2,
which is for dual-core AMD processors, AND AM2+, which is for
quad-core processors.

Cons: The chipset runs hot, all the time. Everyone says this is ok,
but I have started to see blocks of color during post that do not
belong. The solid state caps burnt on the first board. Seemed to to ok
with the Athlon, but have had nothing but trouble with the Phenom. It
randomly will not boot. When Grub does start, you can get Linux to
load, and nothing else.
 
Flasherly said:
(Best I can figure on the scenario below is I'll take a hit on some
bus-architecture speeds. Ideal, crossed-fingers scenario...).

Thought I was stuck to a dualcore, but. . .things can happen.

AMD Phenom X4 9550 2.2 GHz Quad-Core HD9550WCJ4BGH CPU Processor

sent off for one (used, fair but a little more than Intel quad I'm
running at same 2.2Ghz on another MB, both Gigabytes built pretty
close together;- still too pricey for 65watt variants, or, for getting
into 3Ghz core speeds. Reckon I do like getting CPUs, nice ones, for
$20 or under.)

As is, running a dualcore AMD Athlon 64 X2 4200+ (second update,
this'll be -better be- to a third).

6watts difference between the two (89/95watts on the Phenom). Can't
especially see myself going thru what (at bottom of this post) another
person experienced for a measly 6watts;- besides, that might be a
125watt Phenom that he's dealing with.

The MB...

GA-M61PME-S2
CPU Socket Type AM2
Athlon 64 X2 / Athlon 64 FX / Athlon 64 / Sempron

which is a little strange, as there was a GA-M61PME-S2 P specifically
for a quad (along with, probably, a broader range of AMD offerings,
too). Apparently two revisions to each of these two boards, in a
redundant sort of manufacturing manner of keeping abreast with the AMD
2 to 3 socket transition types.

Mine, per se, isn't a socket 2+ or 3, however, Gigabyte does
specifically state (as a stipend to my model): "(Note) If you install
AMD AM2+ CPU on AM2 motherboard, the system bus speed will downgrade
from HT3.0(5200MHz) to HT1.0(2000 MT/s) spec; however, the frequency
of AM2+ CPU will not be impacted. Please refer "CPU Support List" for
more information."

and, on the quoted listing, there's an provision for the Phenom I got
off Gigabyte's support page.

Guess I'll doublecheck my BIOS revision and wait for it to come in,
see what happens. (Also snoop around, see if I can't cut/grind down
some HS material and fashion an augmented chipset cooler. *see notes.
Last I poked around with an IR temp sensor I couldn't find the sucker,
although it most definitely will climb up to 135F during summertime.)

-
couple notes noted from roundabout the time I bought it (Newegg cust
reviews) ...

For a dollar more, you can get the GA-M61PME-S2P.
The only difference I notice, and the only reason I gave it four eggs,
is that the CPU socket on that board is compatible both with AM2,
which is for dual-core AMD processors, AND AM2+, which is for
quad-core processors.

Cons: The chipset runs hot, all the time. Everyone says this is ok,
but I have started to see blocks of color during post that do not
belong. The solid state caps burnt on the first board. Seemed to to ok
with the Athlon, but have had nothing but trouble with the Phenom. It
randomly will not boot. When Grub does start, you can get Linux to
load, and nothing else.

You should verify, when the system is running, that the bus
clock isn't out of spec. Your high temperatures reported,
might be caused by the bus transceivers running at 4000MT/sec.
Which they weren't intended to do. As you note in your copy/paste
above, that board is going to turn down the system bus clock, when
a modern processor is installed. The 6100 would not have
been designed for that. So the bus multiplier value
must be kept lower than normal. The system bus clock must
be set to the lower frequency defined by the limits of the
chipset bus interface and the CPU bus interface. The processor
might say "5200" but the chipset "2000", so the board must
be run at 2000 or less. For testing purposes, people have run
the bus multiplier as low as 1x. Just to prove the clock synthesizer
can handle that.

On the initial batch of 6100 chips, some seemed to be
leaving the factory poorly tested. As the colored block
problem was seen in those days, on brand new motherboards.
That problem seemed to disappear later, as NVidia got
their production process under control. I no longer
saw complaints about the colored blocks after the
initial batch of boards.

Paul
 
You should verify, when the system is running, that the bus
clock isn't out of spec. Your high temperatures reported,
might be caused by the bus transceivers running at 4000MT/sec.
Which they weren't intended to do. As you note in your copy/paste
above, that board is going to turn down the system bus clock, when
a modern processor is installed. The 6100 would not have
been designed for that. So the bus multiplier value
must be kept lower than normal. The system bus clock must
be set to the lower frequency defined by the limits of the
chipset bus interface and the CPU bus interface. The processor
might say "5200" but the chipset "2000", so the board must
be run at 2000 or less. For testing purposes, people have run
the bus multiplier as low as 1x. Just to prove the clock synthesizer
can handle that.

Basically, [ If you install AMD AM2+ CPU on AM2 motherboard, the
system bus speed will downgrade from HT3.0(5200MHz) to HT1.0(2000
MT/s) spec; however, the frequency of AM2+ CPU will not be impacted.
], it's saying an automatic routine pre-established, that chipset
support is intended to fall-back presumably within AMD specs, both for
chips capable of adapting to AMD S2-3 (this Phenom is a AMD socket
AM2+);- screwy, that -- processors scattered all over the socket(s)
skew.

It's a Phenom -X4 9550- AMD. Not 6100 Intel. Updating my other
Gigabittle board, Paul. But, I hear you -- if it isn't designed,
doesn't automatically adapt to the bus speed, not without *manually*
adapting, so as -- 'The system bus clock must be set to the lower
frequency defined by the limits of the chipset bus interface and the
CPU bus interface.'.. yeah, then I'm stuck swapping back processors in
order to get the thing to boot, possibly in order to drop back the
BIOS timings before attempting another CPU install.

Nasty. Used to have that with a PCI-E videoboard if installed and the
BIOS was set to boot from its vidchip (regardless, if/and how the BIOS
was intended to handle a contingency for a seated PCI-E board);-
Except wouldn't need any thermal paste, heh, to [repeatedly re]seat a
vidboard.

All excellent points. Just hope Gigabyte plays nice and I don't have
to deal with them, per se, "timings." Of course, perhaps it's better
now, more automated in BIOS menu timing selections (than MB used to
be, older than this 6yr-old M61pme-s2).

I wouldn't presume my reported temps -- via, hm, SpeedFan Ver4.5, for
up to 135F during the summer (yeah, it's pretty hot here) -- are other
than indicative of poorly implemented temperature/specification
designs, on the part of Gigabyte. Got it hooked up into a couple amps,
one being top-notch studio-grade soundamp, ART rack stuff, via PCI
soundcard, btw;- that ART can tend run even hotter, especially when
disconnecting an internal, dinky little DC fan;- the whole of its
sides are massive 12" finned heat sinks for four output transistors,
which I've seen get hotter than hell.
On the initial batch of 6100 chips, some seemed to be
leaving the factory poorly tested. As the colored block
problem was seen in those days, on brand new motherboards.
That problem seemed to disappear later, as NVidia got
their production process under control. I no longer
saw complaints about the colored blocks after the
initial batch of boards.

Well, this Phenom -X4 9550's- extra cores will be MOST WELCOME for
additionally another half-dozen virtual sound processing stages,
potentials I'm presently locked out of, due to the AMD Athlon 64 X2
4200+, that's presently socketed, and incapable of rising up with
enough -UMPH!- to match a present generation of studio sound
processing software.

Proved the processor takes as Gigabyte promises. It'll be heaven.
It'll be the best sound system around hereabouts, for miles around me,
which I can theoretically guarantee in writing. (Thing's so damned
sensitive already, I'm near to having to adjust for efficiency on
every professionaly prerecorded track played. But it's sweet once
there, heh.)
 
MB m61pme-s2 (with a Athlon x2 4200 -core brisbane 2.2Ghz)
F2 BIOS

CPU-Z reports :

mem 368Mhz
BUS 200(x11)
HT(Link) 1004Mhz

Basically assembled, stuck in a couple CPUs, turned it on and it's
been on now for 6 years now. Haven't had much else to do with the
BIOS, other than couple simple tweaks to the HDs/setup, power
management, &etc.

It's been a great board for handling "supported CPUs." So far.

I'll have the
AMD Phenom X4 9550 2.2 GHz Quad-Core HD9550WCJ4BGH CPU Processor
shipped in by this, maybe next week.
 
I'll have the
AMD Phenom X4 9550 2.2 GHz Quad-Core HD9550WCJ4BGH CPU Processor
shipped in by this, maybe next week.

Fast for Ebay. Got here quicker than expected, two days, this
afternoon's USPS mail. About 30 minutes to put it in tonite. BIOS
rev. identified the processor and itself as patched for it. [Careful
of the connections were the same, but] Windows nevertheless did a
rescan on the MB's chipset drivers, to realign them.

So, I've got four cores now to screw with some nifty sound processing
stuff (bogging down the last one with two). Sizzle up some bacon,
presumably.

Rock my soul came up when powered up, so guess I've no choice but to
dedicate EBiship's _I'm Struttin' My Stuff_ to all you ladies out
there following in swiftly abated breaths.
 
so the board must
be run at 2000 or less. For testing purposes, people have run
the bus multiplier as low as 1x. Just to prove the clock synthesizer
can handle that.

There's not much on that board's BIOS, other than the multiplier and
auto. And it's lower slightly than 2000.

Found the culprit, some help - AMD's 'Cool & Quiet' wasn't engaged.
Once on, there's some improvement, a few degrees cooler, nor any
degradation hits offhand/practically for AMD's "performance" option.

I still don't trust that Gigabyte chipset, but won't know unit it gets
hotter, in the summer. A squeeze to access the HDs, tho it's possible
to fashion a chipset cooler from an old CPU HS, cut one down and try
layering it over Gigabyte's. Might work for a quick-take improvement.

Not bad, worth an easy pop-in, a $25 Ebay used quad Phenom. I've
greater stability from engaging more cores/options in sound stream
processing, I hadn't with the dualcore. Cheap and lucky for a 5+ yr.
old MB - long last and now maxed out. Original order...

1/11/2009
Inv #
Order #

AMD Athlon 64 LE-1640 Orleans 2.6GHz Socket AM2 45W Single-Core
Processor ADH1640DHBOX
Item #: N82E16819103239 512M L2 cache (not 1M)
--Processors (CPUs) Return Policy $35.99

G.SKILL 2GB (2 x 1GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual
Channel Kit Desktop Memory Model F2-6400CL5D-2GBNQ
Item #: N82E16820231098
Limited Non-Refundable 30-Day Return Policy $31.99 2

GIGABYTE GA-M61PME-S2 AM2 NVIDIA GeForce 6100 / nForce 430 Micro
ATX AMD Motherboard
Item #: N82E16813128333
30 Day Return Policy
Gigabyte (626)-854-9338 option 4 www.giga-byte.com $48.99
 
Back
Top