THG (P)reviews "Core 2 Quadro" - aka Kentsfield!

Toggle sidebar Toggle sidebar
No more damned topics about "Kentsfield bottlenecked by FSB"... there is not much difference between the 1067 and 1333 Kentsfield at 2667MHz.
 
Does this mean the price of the X6800 will drop?
Click to expand...
I doubt it. Intel generally only maintains 1 Extreme Edition price slot at a time which means that the X6800 and Kentsfield will likely co-exist at the same price. I believe, the price of the X6800, 965EE, 955EE, and 840EE are all still officially $999. In this case, it'll make some sense since in some cases the extra clock speed of the X6800 may be more important than extra cores. Whatever's going on with the 3.2GHz X6900, I don't know.
 
lawl


It's funny how these reviews go straight to the forums in a matter of mins. This threads probably gonna get huge and everyone is gonna say, "Oh, I gotta have a core 2 quadro! It's so good, and way better than what I have now. Now should I do this, or this, or this....lol

Next year its gonna be some octo core or something with similar responses.



Techie
 
it pains me to say this but, AMD has been sleeping on the job. usually I like to see AMD with higher performance but intel has pulled ahead for the moment...
 
THIS IS WHY I LOOOVE THG! 😀 that is sooo cool. im so glad i waited and didnt buy a conroe.

my saved money will go towards the new quadro cpu and then i will oc and oc and oc and oc even a little more 😀!!! and maybe buy a ocz phase change unit for 300 and keep my temps just below zero 8O lol.


*though i didnt catch if the new processors are the same size or not?? does anyone know
 
THIS IS WHY I LOOOVE THG! 😀 that is sooo cool. im so glad i waited and didnt buy a conroe.

my saved money will go towards the new quadro cpu and then i will oc and oc and oc and oc even a little more 😀!!! and maybe buy a ocz phase change unit for 300 and keep my temps just below zero 8O lol.


*though i didnt catch if the new processors are the same size or not?? does anyone know
Click to expand...

same size as in? die size? lol no there 2 core 2s next to eachother so they are 2 times the size in that aspect.
 
No more damned topics about "Kentsfield bottlenecked by FSB"... there is not much difference between the 1067 and 1333 Kentsfield at 2667MHz.
Click to expand...


EDIT: Whollly crap, did you see the utter butt thump the multi-media math benches are putting on the Athlon/P4 line..... jiminy --- now we see why all those encoder benches seem to take off. I mean look at the SiSoft Multimedia Integer bench, a 3.33 GHz Kentsfield is 7x (700%) faster than an FX-62.

Looking at the spread in the matrix of data, looks like a few companies have lazy software writers and are not writing well optimized code for their customers (quad core or not).


Jack

8O
I almost fell off my chair.
had to put my reading glasses on and clean the monitor.
I new I was holding off on core2 for a reason. :)
 
http://www.tomshardware.com/2006/09/10/four_cores_on_the_rampage/index.html
Those extra cores do make a difference! Also, THG noted that the top-of-the-line will be ~$1000 - the price of the X6800 now. Does this mean the price of the X6800 will drop? We can all hope... 😀
Click to expand...

Word.
 
No more damned topics about "Kentsfield bottlenecked by FSB"... there is not much difference between the 1067 and 1333 Kentsfield at 2667MHz.
Click to expand...
Its bottlenecked by the video card! Same performance as C2D in all game benchmarks.
 
Want to know how a 4x4 AMD will peform? Hint: It cant be any faster than a quad core Opteron setup e.g. http://www.digitalvideoediting.com/articles/viewarticle.jsp?id=61808 ... which is almost beat by a SINGLE Core 2 running at a stock 3 Ghz.

Kentsfield is way, way ... waaaay faster than anything AMD can offer in its 4x4 offering.

4x4 is just a marketing slogan for a dual processor Opteron rig ... which I have been running for a year already.
 
AdvancedMicro what?
Click to expand...

Well, this is a truly exciting preview of what to expect out of intel in the coming weeks (about 4-6 if you believe their mid-october blurb).

AMD though has their card to play yet, we still need to see what the 4x4 will do. Gaming wise, I expect it to do well with a quad SLI setup, I hope we can see a quad SLI 4x4 up against a Quad SLI Core 2 Duo rig to get a good compare; ultimately the best comparision will come with a 4x4 against a Kentsfield setup with quad SLI. We will need to see.... I am sure there are a few HW sites out there just crazy enough to try it. :)

Interestingly Toms was treated to another AMD dog and pony show (last page of preview) with the 4x4 where they ran a bench for them to see it work. :) .... no numbers, no data but hey they got Halo (the telecoference not the game) to see it in action --- sadly they did not actually see it though.

Jack

4X4 will probably= $x$ though. I still feel it was a knee jerk reaction and will not be competitive in terms of cash per unit of performance, especially compared with the Intel quadcores. AMD needs to get their 65nm quadcores out right fast. One can only hope that AMDs early release rumors are true and that they are well into 45nm work ups. If Intels quad core does come out in october, I think AMDs momentum is in danger of hitting the proverbial immovable object.

What an explosive year in the CPU arena. I cant help but wonder if Intel isnt pushing quadcore out sooner than planned in response to stock prices. There is no competition from AMD right now, and early release of quad core can only hurt C2D sales, so there doesnt seem to be another reason for dropping quad core on the market right now. Additionally, it leaves one wondering what (if anything) Intel has waiting in the wings, not to mention the possiblity that they may be adversely conditioning consumers to expect major advancements every 6 months.

Lots of stuff to ponder
 
4x4 is just a marketing slogan for a dual processor Opteron rig
Click to expand...
Ditto. Though I do certainly hope the 4x4 boards are quite a bit less than Opteron boards. But if Intel gets out Kentsfield before 4x4, I can't see 4x4 catching on. For AMD you would need special cpus and board (which I'm sure price will be reflected there) then you're going to be cooling 2 chips instead of 1 with Kentsfield. Which was something I was happy with when dualcores came around, it gave me the opportunity to run 2 cores at close to the same performance as dual single core Opterons and without the price tag.
 
http://www.tomshardware.com/2006/09/10/four_cores_on_the_rampage/index.html
Those extra cores do make a difference! Also, THG noted that the top-of-the-line will be ~$1000 - the price of the X6800 now. Does this mean the price of the X6800 will drop? We can all hope... 😀
Click to expand...

When Pentium 4 EEs, Pentium D 840EE, 955EE, and 965EEs are all still a grand, I think not. Only AMD ever seems to let its top-dog CPUs drop in price over time. What you WILL get is a multiplier-locked E6800 with the same 2.93 GHz core clock and 4MB L2 that the X6800 has. And yes, that will be much less than $1000.
 
Can someone give a firm answer on what Kentsfield's actual FSB will be? THG quotes both ways:

http://www.tomshardware.com/2006/09/10/four_cores_on_the_rampage/page4.html

Intel zipped up the FSB speed for the Core 2 Quadro to 333 MHz compared with the 266 MHz for Core 2 Duo.
Click to expand...
http://www.tomshardware.com/2006/09/10/four_cores_on_the_rampage/page5.html

The sample was factory-set to a speed of 2.66 GHz and a bus speed of 266 MHz.
Click to expand...
So which is it?

I know their results don't show a decent advantage for a 1333MHz FSB, but that doesn't mean there isn't one if they implement it properly. I've heard that the 1333MHz FSB setting in the i975X chipset uses relaxed latencies compared to the 1067MHz setting which means that although you have higher potential bandwidth you need higher clock speeds to overcome the latency disadvantage. That is likely why the 2.67GHz 1333MHz FSB results aren't much better than the 2.67GHz 1067MHz ones.

I'm unclear whether the 1333MHz FSB latencies can be corrected with a BIOS update or whether they are hardwired and require a revision. The i975X chipset was originally designed for Netburst chips and just introduced the 1067MHz FSB for dual die dual core chips, so the 1333MHz setting was just a test feature which was why things weren't tightened properly.

Now I'm unclear what motherboard THG used. They say a P965 based one although their test setup page said Intel systems used i975X based ones. Maybe that's just for the other Intel chips. Now I'm not sure that the P965 has a 1333MHz FSB setting, so THG might have achieved it using overclock which should have avoided the i975X 1333MHz latency issue. However, with the immaturity of the P965 memory controller and BIOSs, I wouldn't trust it's scaling as indicative of good performance.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom