2 Intern

 • 

2.3K Posts

July 17th, 2007 17:00



SR45 wrote:
This is the link to the final conclusion, which is not to good for such an expensive system  :smileysad:
 


And I quote:
Dell's images of an XPS 720 H2C clocked to 4 GHz caused quite a stir. I would love to see those images? These kinds of settings were impossible to achieve on the system they sent, however - Shelton was only able to increase the stable processor speed from 14x 266 to 14x 271. The CPU topped out at 3.794 GHz, in the same neighborhood of what many unnamable-but-experienced sources told us would be this processor's limit.
 
We were lead to believe that the factory-overclocked XPS 720 H2C had more than enough room for determined overclockers to take it to the next level of performance, but this simply wasn't the case with our sample. Restrictive BIOS settings and components that were already near their limits prevented us from seeing any noticeable performance gains, so this system should be considered an as-delivered solution like any other Dell.
 
 
Peace
 
 

1.9K Posts

July 17th, 2007 17:00

Here is all you need to know.
 
We were lead to believe that the factory-overclocked XPS 720 H2C had more than enough room for determined overclockers to take it to the next level of performance, but this simply wasn't the case with our sample. Restrictive BIOS settings and components that were already near their limits prevented us from seeing any noticeable performance gains, so this system should be considered an as-delivered solution like any other Dell.
 
They were able to tighten the timings a bit over the factory OCed 5-5-5-15.

2 Intern

 • 

12.1K Posts

July 17th, 2007 17:00

This is the link to the final conclusion, which is not to good for such an expensive system  :smileysad:
 

1.2K Posts

July 17th, 2007 20:00



pointguard1122 wrote:
Here is all you need to know.
 
We were lead to believe that the factory-overclocked XPS 720 H2C had more than enough room for determined overclockers to take it to the next level of performance, but this simply wasn't the case with our sample. Restrictive BIOS settings and components that were already near their limits prevented us from seeing any noticeable performance gains, so this system should be considered an as-delivered solution like any other Dell.
 
They were able to tighten the timings a bit over the factory OCed 5-5-5-15.


Wow, what a let down!  In other words, the 680 motherboard in the 720 is no better (in actual use) than the 590 motherboard in the 700.
 
Edit:  I also note the memory bandwidth results for the 720 are not good at all ... a high-end system with ddr 1066 ram supported by a qx6800 at 3.7+ Ghz should score in the 8000-10000 range on that Sandra test where the XPS scores 6200.  So, even when equipped with fast Corsair memory, the motherboard is slow and loose.


Message Edited by Aivas47a on 07-17-2007 04:23 PM

Community Manager

 • 

56.9K Posts

July 17th, 2007 20:00

All,

It would have been more interesting if they had published system temps and fan speeds while doing this. I have passed it on.

1.9K Posts

July 17th, 2007 22:00



Aivas47a wrote:

Wow, what a let down!  In other words, the 680 motherboard in the 720 is no better (in actual use) than the 590 motherboard in the 700.

I hope the news isn't that bad.  We will really know the differences when we test our current setups with the new mobos.  I think the main drag of the article is that they had no headroom over the factory OC.  I think they were lucky to get one that ran stable at that speed.  The 20,000+ 3DMARK score and 4.0 image was all garbage.  I never trust anything from Dell marketing anymore.  They are the biggest cause of trouble in Dell at the moment.

2 Intern

 • 

3.3K Posts

July 17th, 2007 22:00

   Michael,
Great find.

2 Intern

 • 

2.3K Posts

July 17th, 2007 22:00



DELL-ChrisM wrote:
All,

It would have been more interesting if they had published system temps and fan speeds while doing this. I have passed it on.


Chris... Thanx, I would love to see the data and also the the 4.0 images/data too.
Thanx Budd.... this is just to quench my curiousity.
 
Peace
 
 

2 Intern

 • 

3.3K Posts

July 17th, 2007 23:00

   PG,
 I agree with you. They were led to believe there was more room.
A QX6800 @ 4GHz, ???, maybe with nitrogen, but I'd have to see it to believe it, now an X6800, thats different. I've got this one running @3.636 as we speak!!!
 
 
Stable as a rock I might add

Community Manager

 • 

56.9K Posts

July 18th, 2007 01:00

gdwrnch3,

Can you run CPU-Z and tell if it shows the CPUID? Mine does not but I have an older proc.

Message Edited by DELL-ChrisM on 07-17-2007 09:49 PM

2 Intern

 • 

2.6K Posts

July 18th, 2007 01:00

i think we should see what they come up with tomorrow. The speed of the 720 is already very high and the qx6800 does not have too much room left. They may get higher FSB and lower divider but raw speed I do not think will be much better.

2 Intern

 • 

3.3K Posts

July 18th, 2007 09:00

   Chris, it does not, it just shows X6800.
I have some screen clips if you like.


Message Edited by gdwrnch3 on 07-18-2007 08:12 AM

24 Posts

July 18th, 2007 11:00

A quad core at 3.7ghz is already an impressive overclock. There's no reason to be disappointed.

1.2K Posts

July 18th, 2007 13:00



tphillips63 wrote:
i think we should see what they come up with tomorrow. The speed of the 720 is already very high and the qx6800 does not have too much room left. They may get higher FSB and lower divider but raw speed I do not think will be much better.

That is right ... in fact in some cases 3.73Ghz may already be too high!
 
A much more interesting test would be to drop the cpu multiplier to 7 and see how far the fsb can be raised for a quad core processor.  This would help answer the question whether folks can expect to get a high level of performance from the 720 using a $260 q6600 instead of a $1000+ qx6800.
 
FSB overclocking of quad core processors has been a weakness for many 680 motherboards, but some models (evga new revision with p27 or later bios) have been able to reach 400+.  That's sort of a "magic number" for running a q6600 at 3.6ghz.
 
It would also be interesting to see how much further the memory timings could be tightened.  There should definitely be more headroom in a pair of new corsair ddr2 1066 sticks.

2 Intern

 • 

2.6K Posts

July 18th, 2007 16:00

It would be great but ... wait Dell did not expose FSB options to even extreme CPU's so it's a no go on the Dell.
I do think it would be better around the 3.6GHz area and a better FSB and low latency DDR2 800MHz RAM.
 
 
 
I know all the arguments against it but for crying out loud Intel branded motherboards have pretty good OC capability now and have for several years.  Why?  Because of consumer

Aivas47a wrote:


tphillips63 wrote:
i think we should see what they come up with tomorrow. The speed of the 720 is already very high and the qx6800 does not have too much room left. They may get higher FSB and lower divider but raw speed I do not think will be much better.

That is right ... in fact in some cases 3.73Ghz may already be too high!
 
A much more interesting test would be to drop the cpu multiplier to 7 and see how far the fsb can be raised for a quad core processor.  This would help answer the question whether folks can expect to get a high level of performance from the 720 using a $260 q6600 instead of a $1000+ qx6800.
 
FSB overclocking of quad core processors has been a weakness for many 680 motherboards, but some models (evga new revision with p27 or later bios) have been able to reach 400+.  That's sort of a "magic number" for running a q6600 at 3.6ghz.
 
It would also be interesting to see how much further the memory timings could be tightened.  There should definitely be more headroom in a pair of new corsair ddr2 1066 sticks.



No Events found!

Top