Athlon64 3000 - who's taking the $215 plunge?

Mercutio

Fatwah on Western Digital
Joined
Jan 17, 2002
Messages
22,269
Location
I am omnipresent
Mercutio said:
Mercutio said:
Bought a couple DFI K8M800-MLVF boards last week (u-ATX integrated everything). We'll see how these are. DFI boards aren't terribly exciting, but this one has a comfortable number of big caps and a VERY sensible layout.

Very sensible and utterly unable to boot windows, no matter what I try. Wow. What a great product.

Just to follow up...

After some investigation, I discovered that the problem on both of the DFI boards I bought was caused by extra-long bits of "flash" from the soldered traces on the back of the board. These little spikes were just touching the motherboard tray on the chassis.

Since I wasn't absolutely confident in my ability to remove the long bits without damaging anything, I mounted an anti-static bag behind the nuts the board normally screws into, and now the little guys work just the way I'd want them to.
 

Adcadet

Storage Freak
Joined
Jan 14, 2002
Messages
1,861
Location
44.8, -91.5
Mercutio said:
Mercutio said:
Mercutio said:
Bought a couple DFI K8M800-MLVF boards last week (u-ATX integrated everything). We'll see how these are. DFI boards aren't terribly exciting, but this one has a comfortable number of big caps and a VERY sensible layout.

Very sensible and utterly unable to boot windows, no matter what I try. Wow. What a great product.

Just to follow up...

After some investigation, I discovered that the problem on both of the DFI boards I bought was caused by extra-long bits of "flash" from the soldered traces on the back of the board. These little spikes were just touching the motherboard tray on the chassis.

Since I wasn't absolutely confident in my ability to remove the long bits without damaging anything, I mounted an anti-static bag behind the nuts the board normally screws into, and now the little guys work just the way I'd want them to.

I had a motherboard a few years ago, and the same thing happened. IIRC it ways a Tyan board, surprizingly. I may not have had the board perfectly mounted in, however. Regardless, with an antistatic bag behind part of the board, it worked great. To this day I always wonder if I should consider this very simple safeguard when installing new MB's.
 

LiamC

Storage Is My Life
Joined
Feb 7, 2002
Messages
2,016
Location
Canberra
Time,

the GB board has the correct chipset for Gig-E LAN, but the manufacturer only equips it for use as a 10/100 LAN connection. My reading of H's post was that he was looking for a S939 solution with Gig-E. The GigaByte board doesn't have it. Or did I read it wrong?
 

time

Storage? I am Storage!
Joined
Jan 18, 2002
Messages
4,932
Location
Brisbane, Oz
You read it wrong. :)

The board has two 'ethernet' ports, but Gigabyte chose to equip just one with gigabit capability - I think that's what the author was whinging about.

The benchmarks show the gigabit port delivering what I would have said was adequate throughput for a workstation.
 

CityK

Storage Freak Apprentice
Joined
Sep 2, 2002
Messages
1,719
time said:
The board has two 'ethernet' ports, but Gigabyte chose to equip just one with gigabit capability - I think that's what the author was whinging about.
If I was reviewing the board I'd whine about it too - why the helll did Gigabyte couple the fast ethernet on the integrated (GigE supporting) controller and then bind GigE on the PCI ??? Is their some sort of expensive/restrictive Nvidia licensing issue involved in utilizing the GigE capabilities of the chipset?

I would also whine a lot about all the stupid dinky little fans they put on board.
 

time

Storage? I am Storage!
Joined
Jan 18, 2002
Messages
4,932
Location
Brisbane, Oz
CityK said:
If I was reviewing the board I'd whine about it too - why the helll did Gigabyte couple the fast ethernet on the integrated (GigE supporting) controller and then bind GigE on the PCI ???
Perhaps this is a workaround for the throughput issue? In any case, if you believe the TechReport test, the Marvel PCI solution worked quite satisfactorily, with very low CPU utilization to boot. I can't see that the board is intended as a server solution, and the measured network throughput is more than the storage system could sustain anyway.

I would also whine a lot about all the stupid dinky little fans they put on board.
Yes, but there's only one and as Merc said, just rip it off and throw it in the bin - it's only for overclockers.
 

time

Storage? I am Storage!
Joined
Jan 18, 2002
Messages
4,932
Location
Brisbane, Oz
CityK said:
I count two time :D

Whoops! I thought the Dual Power System was just an option like it has been in the past. I don't know that it's worth having if it really needs that fan. :(

Buck, I ordered a GA-K8VT800 Pro (because of the lower price) but the supplier was out of stock. They claimed the GA-K8NS Pro (also out of stock) was far more popular. Still a tad expensive, I think, but I feel better that I'm not inflicting the Realtek network chip on a customer - 80% CPU utilization really is over the top!
 

time

Storage? I am Storage!
Joined
Jan 18, 2002
Messages
4,932
Location
Brisbane, Oz
Regarding TechReport's review of the GA-K8NSNXP-939:

The reviewer measured extraordinarily low 1394 throughput. He stated that no driver was included and that therefore the generic WinXP driver must be adequate, then concluded that the hardware was to blame.

Having just set up a GA-8NS Pro, essentially the same chipset with the same TI controller, I note that it does come with a specific IEEE1394 driver.

Are there any competent reviewers left out there? (Aces Hardware excepted).
 

CityK

Storage Freak Apprentice
Joined
Sep 2, 2002
Messages
1,719
time said:
Perhaps this is a workaround for the throughput issue? In any case, if you believe the TechReport test, the Marvel PCI solution worked quite satisfactorily, with very low CPU utilization to boot.
BTW, what thoroughput issue exists with the nF3 GigE?
 
Top