Deep Fritz 10 -- kn/s -- Skulltrail update :

[WuShock]

http://www.tomshardware.com/2008/02/08/ ... age14.html

[Nid Hogge]

That's a nice boost there. Too expensive, but still nice for benching\encoding\rendering.

http://www.jens.tauchclub-krems.at/dive ... marks.html Has one OC'ed to 4788 MHz = 38,91 Relative speed 18674 Kn/s.

[emerson4301972]

Wow, maybe developments like these help put off the Hydra project.

Improvements in conventional hardwares is faster and cheaper than it would have cost Hydra to improve its hardware

Has Hydra finally lost all its head? Hope not

[M ANSARI]

I would like to get my hands on one of those. I can't wait till they are out. I checked Newegg and still nothing ... but should be soon. I have some special preparation for this release.

[tano-urayoan]

The review this system got from Tomshardware is summarized in their conclusion: " Technologically Immature And Lacking Software Support".

The reviews I have read all emphasis the problem of using the FB-DIMM memory which limits the system.

Here are some links to reviews of it.
http://www.tomshardware.com/2008/02/08/ ... index.html
http://www.anandtech.com/cpuchipsets/in ... spx?i=3216
http://www.gamepyre.com/hardwared.html?aid=919&p=1
http://www.hothardware.com/Articles/Int ... X9775_x_2/
http://www.pcper.com/article.php?aid=514&type=expert
http://techgage.com/article/intel_skull ... m_preview/
http://techreport.com/articles.x/14052

[Spock]

Yes, this is a bit like the AMD 4x4 thing - flawed. Although in this case, very fast but flawed. Having to use FB-DIMMS is crazy for example.

[Nid Hogge]

Yes, this is a bit like the AMD 4x4 thing - flawed. Although in this case, very fast but flawed. Having to use FB-DIMMS is crazy for example.

I don't know. too early to conclude anything yet. It ain't flawed because it's basiclly just a dual socket board with OC ability. Just like the dual Xeons Hiarcs and Rybka teams uses frequently in tournaments. only difference is that they cant OC.

Also I don't understand all the fuss about FB-DIMMS. Sure, they use some extra watts, but hey, If you can afford the Skulltrail system, that's probably not going to concern you. Also there's some 800 MHz CL3 FB-DIMMs starting to show up.. It's not limiting the OC percentage. Ive seen some 5.5 GHz+ already.. just a matter of time till it gets steady.

This is nothing like QuadFX btw. QuadFX was a joke. It was just a weak attempt by AMD to counter Intel's dominance with it's Quads until it finally gets Phenom out the door.. it deserved to fail.

Skulltrail is just Intel milking the cow right now. they can play around and try to raise they're ASP's.. that's what happens when there's no competition. It's made to sweep all the World Records with ease.. just for the show off.

Despite the fact that it's going to be the fastest system for a while now.. I wouldn't waste my money on it. there are so few applications that can use 8 cores today, and besides, in a few months, it's going to be obolete when Nehalem debuts not far from now.. So no point spending so much money on a system that's going to be outdated in a few months..
Sweet spot right now is Q6600 OC. Should be more than enough for most people.. and still won't be maxed as there's only a few apps that use all 4 cores yet..

[M ANSARI]

To most of us here ... the more cores the better since most of us are interested in Chess engines which are one of the few applications so far that can benefit from as many cores as possible. Already Rybka and Hiarcs and ofcourse Zappa seem to be quite stable on 8 cores. I am a little disappointed with the platform since from the reviews it seems to have a very immature bios. With regards to FBDIMMS ... really what was Intel thinking ... this is obviously an enthusiast and performance freak over clocking platform ... why on earth have registered memory ???? Registered memory is useful for server applications ... but surely not for what this platform is supposed to be showing.

Hopefully they will sort out the problems ... I think in March a new bug free QX9775 will be out. People will find out ways to push the FSB maybe by pin mods ... there will be BIOS hacks ... but still FBDIMMS are FBDIMMS and they are absolutely horrible. They cannot be pushed very fast because they get extremely (I mean extremely) HOT.

Anyway it should keep me busy March to April. I already have built a special cooling vaporchilled system to handle two hot Quadcores. I am calculating 270watts of heat per processor ... so should be more than engough since the 45nm processors are producing less heat than I expected. I have to figure ways around the memory bottleneck ... but I am confident of getting a good stable 8 cores of 5Ghz.

[Nick C]

> but I am confident of getting a good stable 8 cores of 5Ghz

Good luck with that Majd. You should hopefully have 8x5 ready for the next Freestyle competition ... I hope you enter!

Cheers,
Nick

[M ANSARI]

> but I am confident of getting a good stable 8 cores of 5Ghz

Good luck with that Majd. You should hopefully have 8x5 ready for the next Freestyle competition ... I hope you enter!

Cheers,
Nick

I really doubt it ... but if you want a go at it you have free use of my hardware Nick