Computex 2005 Day 2 - ULi's new Athlon 64 Chipset, G70, CrossFire & more
by Anand Lal Shimpi on June 1, 2005 1:57 AM EST- Posted in
- Trade Shows
The show goes on...
In this article we've got more pictures of NVIDIA's G70, pictures of all of the CrossFire motherboards from manufacturers at the show, a look at cooling at Computex, benchmarks of ULi's new Socket-939 chipset and much more.
A view of Taipei by the Taiwan World Trade Center and Taipei 101
If you haven't already, be sure to check out coverage from earlier in the week:
Computex 2005 Early Bird Coverage: NVIDIA's G70, Athlon 64 BTX and more
ATI's Multi-GPU Solution: CrossFire
Computex 2005 Day 1 - ATI R520 Sighting, NVIDIA's new Chipset
More from Gigabyte
Gigabyte's i-RAM card generated quite a bit of interest in our Early Bird Computex Coverage, so we decided to bring you all some more pictures and information about the solution.
As we alluded to in our initial article, the i-RAM card features 4 DIMM slots and supports DDR200 DIMMs. You can use faster memory but it will only run at DDR200.
Below you can see the i-RAM card populated and in action:
53 Comments
View All Comments
Tanstafl - Tuesday, June 7, 2005 - link
#52, the reason they chose to use a PCI slot is most likely because that allows them to use standby power to retain the content of the RAM drive. If you check what's been written about the Gigabyte RAM disk card you will find that the battery is only used when the machine is unplugged, or when there is a power outage.erwos - Friday, June 3, 2005 - link
"#50, yeah there is another purpose other than power. The PCI slot holds the card in place so it doesn't rock around inside your tower. :)"Again, you don't need a PCI slot for that. A mounting bracket would solve that just fine, and possibly even give me room for two in a single 5.25 bay. It's solid state - the thing isn't going to move.
mlittl3 - Thursday, June 2, 2005 - link
#50, yeah there is another purpose other than power. The PCI slot holds the card in place so it doesn't rock around inside your tower. :)erwos - Thursday, June 2, 2005 - link
My _theory_ is that the PCI bus is used for something other than just power. Maybe some kind of configuration and/or debugging. I find it tough to believe they made this thing use PCI for power alone, because a simple 4-pin would have solved that pretty easily.-Erwos
KayKay - Thursday, June 2, 2005 - link
Wow, this was an excellent article, one of the best i've seen to date.AGP and PCI-E on the same board, don't see much of that, but im sure people out there would appreciate it if it was widely available
nserra - Thursday, June 2, 2005 - link
#40 What are you talking about I have been installing PC's with ASrock boards (K8 Combo-Z and K8Upgrade-1689) and all super stable and fast (boot time records!!!). I have to say that these are all much faster than the asus boards that I was installing (K8V SE, ...)You must have been unlucky, and don’t forget one thing if mobo makers and bios makers put the same affords they do on intel chipsets, I doubt that Uli and Sis looked so bad...
jiulemoigt - Thursday, June 2, 2005 - link
this is wrong pikhttp://images.anandtech.com/reviews/tradeshows/200...
on p9
http://www.anandtech.com/tradeshows/showdoc.aspx?i...
there is three dead give aways one which i feel like
an idiot for missing is the VGA connector and the
others are the audio jacks are sperated in one pik but not in the cut away and it still supports two ethernet adapters which is also missing in the second pik.
Penth - Wednesday, June 1, 2005 - link
mosoon:You are right with your guess that even the newest and fastest SATA drives fail to max out a SATA I connection. I think my Raptor maxes out at about 70-something MB per second. Not even half the capability of the SATA interface.
I am supremely interested in this card, but I think if I was going to be paying that much per GB, I would wait for a SATA II connection, where the increase in bandwidth would be very apparent with 300MB/s sustained transfers.
monsoon - Wednesday, June 1, 2005 - link
MLITTL3 forgive my ignorance, but how does it come that data transfer through the SATA connector is faster when pumped by the i-RAM instead of a SATA HD ?i mean, ok, i understand HD access latency is lower, but shouldn't we already have the SATA BUS at its max if we buy some kick ass SATA HD ?
or is it that SATA HDs around ( including RAPTORS, etc.. ) come nowhere near the bandwidth possibility of the SATA bus ?
cordially,
JOHN
PS - please bear with my ignorance !!!
DanDaManJC - Wednesday, June 1, 2005 - link
hmmm that ULi mobo, well specifically their chip that allows both agp and pci-express looks awesome!I've been waiting around for a solution like that... now all I need is $500 for the rest of the pc... haha *sigh*