Intel's Core 2 Quadro Kentsfield: Four Cores on a Rampage

[chaosgs]

i only see 1 use for a quad core as of right now and thats servers or big buisnesses.

 

[strata]

You said there's a dual core patch for doom 3?? Could you link it to me please? i can't find it anywhere

 

[blargman]

Tom's, lets see some x264 benchmarks! pretty please!

that processor encodes like a mad man

 

[kinneer]

The home user is going to require the 4 cores as high def becomes more mainstream.

Nope.



I disagree. That is like saying "Who needs more than 640k". If the resource is available, it will be used. Games or otherwise. Take Vista for example.

 

[ElPolloDiablo]

What magical butt did you manage to pull out this cpu all of a sudden? Not provided by Intel!?


LIES!

 

[Eurasianman]

IMO, quad core is over kill for windows xp.

When is a overkill not good for us

overkill is more than we need,and the pc has not yet attained that status.
when we are rendering games in real time,and have 60% overhead ;that will be overkill;untill we play 2 such games at one time.

heres your future gamer,playing online in several different games,all renered in real time.

what in that scenario would be overkill?64 cores?sokolum is right,when is overkill not a good thing? 😛

when people don't have an OS that supports it correctly :?:

 

[MplsBob]

Has anyone seen any pricing information on the Quadra?

 

[prolfe]

Boe I want to say something about anomolous benchmark results on the VGA charts. Have you checked the entire tested system at the other sites to confirm that they are the same? For example, if THG VGA chart has a 7900GT that scores XXXX FPS an a given game, but anandtech or madshrimps or firingsquad has a different FPS, you have to check the system specs. If firingsquad used a C2D, but THG used A64, there would be a different score. The concept behind the VGA charts, as I see it, is to compare the cards to each other with identical systems. So if the older GPU's were benched with A64 processors, the new GPU's would have to be benched with A64 also to get an accurate reading of the cards compared to each other; this would cause THG to not match other sites, who are probably using C2D. If you have checked the entire testbeds and they match, but THG still gets different readings, check the resolution and quality settings!

Hope this helps.

 

[TheeBadOne]

Hey, more performance is always good news. Keep things fresh, not stale, keep looing for ways to move ahead. Sometimes it turns out to be a dead end, but in the end progress always occurs.
As for myself, I do a lot of Photoshop CS2 and run multi apps all the time, how can this not be good news?

 

[MplsBob]

verndewd

You were quite right. I printed the article and read it propped up in bed, trying to keep my eyes open, before I turned the lights out. I missed that sentence that had any info regarding pricing.

 

[E]

Let's be honest, you really don't need this for gaming at least not for a few years. Even dual-core is a bit overkill. Moer of your gameplay is impacted by your video card at this point.

Unless you are doing animation, editing, or especially 3D rendering, with a muti-threaded renderer, you are wasting your money. Actually I don't know why at this point Tom's hardware insists on using 3DS Max as a render test. I don't believe it's a multithreaded renderer is it? Unless they are referring to Mental Ray, which is a perfect example of what they should use to test.

 

[bga]

Let's be honest, you really don't need this for gaming at least not for a few years.

If everybody looked at the benchmarks -instead of saying that quad is the fastest CPU yet - it can be seen that quad core does not add anything to gaming performance. A dualcore is much faster (and cheaper), and due to thermal performance can be overclocked much more. If you look at recent workstation quad core tests with the Xeon 51XX series/5000X chipset (with lame FB-DIMM RAM), it is even more obvious - quad core does not benefit game playing. If gaming is your main application, then stay away from these chips until new games are written to take advantage of this - and this will take a few years. The "State of the art" chip for gaming will be the highest clocked dual core Intel chip for at least another year. Kentsfield/Woodcrest/AMD 4x4 are far from the pinnacle of gaming performance.

- Don't fall for the: "If it is newer, it must be better".

Actually I don't know why at this point Tom's hardware insists on using 3DS Max as a render test. I don't believe it's a multithreaded renderer is it?

On TH's tests (and Anandtechs), the 3DSMax renderer is one of the very few apps which DO scale very well when adding cores. So it must be a good multitreader. Also my experience from a HP xw9300 quad core Opteron workstation.

 

[boe]

Boe I want to say something about anomolous benchmark results on the VGA charts. Have you checked the entire tested system at the other sites to confirm that they are the same?

Thanks - I did notice the environments were slightly different and did mention that. However, I mentioned in another thread - different game patch levels, different drivers and different CPU's can only register so much of a difference in the CPU however that to a degree will effect all the benchmarks - there aren't too many patches that only effect the intel processor or the AMD processor. There would most likely be a proportional benchmark result - that is not the case here - yet freakishly enough every other site had that same result - not in FPS but in proportional difference between the cards.

 

[impreza]

Its not a true quad core its two dual cores put in 1 package and the 8mb can't be used by all the cores.

 

[Slobogob]

Its not a true quad core its two dual cores put in 1 package and the 8mb can't be used by all the cores.

So what? :roll:

It doesn´t seem to hinder performance the way it did the Pentium Ds and i bet the manufacturing is cheaper this way than a native Quad on 65nm.

 

[JonathanDeane]

Heck with quad core I want Octo Cores !!! lol Its fun but honestly I can see Intel releaseing those next year some time.... I think cores will be the new Mhz war ! I can see it now, some customer walks in Best Buy and gets the salesmen who says "Yes this is the new Intel chip in it.. and yes it has twice the power of the old dual core systems !"

 

[prolfe]

Boe,
I'd really like to talk more about this but I think we're in the wrong thread to do so. Would you please link the thread you previously mentioned? I'd like to take a more in-depth look at these anomalies, because THG is my primary source of Hardware info and I'd like to be confident that I'm getting good data.

Thanks

 

[boe]

Boe,
I'd really like to talk more about this but I think we're in the wrong thread to do so. Would you please link the thread you previously mentioned? I'd like to take a more in-depth look at these anomalies, because THG is my primary source of Hardware info and I'd like to be confident that I'm getting good data.

Thanks

http://forumz.tomshardware.com/hardware/Red-Hot-VGA-Charts-ftopict198132.html

Please note I'm not exactly the only person to have noticed the issues. I do not post the card which is most incorrectly represented as people might think I'm a fanboy - I tend to buy the other brands cards even though the card that should be at the top for just about everything belongs to a company I only buy 35% of the time (I buy A LOT of cards for my clients).

 

[Eurasianman]

Let's be honest, you really don't need this for gaming at least not for a few years. Even dual-core is a bit overkill. Moer of your gameplay is impacted by your video card at this point.

Unless you are doing animation, editing, or especially 3D rendering, with a muti-threaded renderer, you are wasting your money. Actually I don't know why at this point Tom's hardware insists on using 3DS Max as a render test. I don't believe it's a multithreaded renderer is it? Unless they are referring to Mental Ray, which is a perfect example of what they should use to test.

I never said you did need quad core for gaming. I simply stated that it's over kill due to the lack of an operating system that can use it correctly, as well as applications. *sighs* You know, I really feel I wasted my money on a Conroe setup... the only reason I did it was because my Athlon 64 was S754 and the mobo had an AGP slot and i figured it'd be better to just get a new computer altogether, rather than get a new mobo that supports S754 and PCI Express. Sadly, I feel that operating system wise, my Athlon runs smoother than my Conroe. Game wise, conroe runs very good.

Point being made: Um, I have college and work and girlfriend = no time to play with my new EXPENSIVE toy

 

[erase_me]

Hi stranger.
I completely agree and like to add two points:


1) I've noticed now a couple of times that a fair amount of these "benchmark" batteries in THG appear to be complete shots in the darkness. I would even go as far as saying they are pointless.
Example: The Quake or Call of Duty benchmarks.

Did Toms hardware actually bother to contact the gaming studio and enquire
a) If the particular game is multithreaded?
b) And if yes, how many threads does it use?

Without this information I have a quite hard time to get any useful conclusions out of the results:
In case they don't use more than two theads - then the entire exercise is pointless. Or if they do - what do I see then: A medioce threaded application or a shitty SMP chip or both or something else (like a memory bandwidth problem)?

In this light I slowly wouldn't be surprised to find in the near future the following fictive review in THG:

THG tested a Beowulf cluster consisting of 256 Conroe rigs.
Here the findings: A single instance of "Doom" appeared not to be accelerated, 3D Studio Max (@stranger: supports actually SMP and work sharing in rendering farms) went through the roof AND - that's why our tester was absolutely blown away, we found that we were able to encode 128 MPEG movies together with 128 instances of Half-life quite crisply.
Very nice, but so what?

2) I would intuitively assume that if I have a spanky new quad core CPU in my hands and I'd like to have a look if this thing is a "hot" CPU or not, I would by select benchies that unearth the stengths and pitfalls of the chip compard with other existing four proc SMP soluttions.
Why is Core 2 Quadro not benchmarked against a four proc AMD or IBM rig on some established server or smp number crunching benchmarks?
Because THG got the chip from Intel and liked to hammer out an article asap, so let's get the usual benchies and off we are ...
In case the entire shebang is lacking conclusions, well, we just call the article: "Intel's Core 2 Quadro Kentsfield: Four Cores on a Rampage".
There was a time when THG was different...

I f I just go back and have a second look: You could write the entire conclusion + Editor's Opinion section without actually testing the chip at all, just based on common sense.

 

[Whizzard9992]

I never said you did need quad core for gaming. I simply stated that it's over kill due to the lack of an operating system that can use it correctly, as well as applications. *sighs* You know, I really feel I wasted my money on a Conroe setup... the only reason I did it was because my Athlon 64 was S754 and the mobo had an AGP slot and i figured it'd be better to just get a new computer altogether, rather than get a new mobo that supports S754 and PCI Express. Sadly, I feel that operating system wise, my Athlon runs smoother than my Conroe. Game wise, conroe runs very good.

How exactly does XP not support 4 cores correctly? The windows scheduler has no problem running on 2 or 4 threads.

If your Athlon 64 is running smoother than your conroe, then perhaps you did something wrong (or you have some flaky hardware). Maybe a $350 Conroe with a $50 mobo was a bad idea after all 😉

 

[Eurasianman]

I never said you did need quad core for gaming. I simply stated that it's over kill due to the lack of an operating system that can use it correctly, as well as applications. *sighs* You know, I really feel I wasted my money on a Conroe setup... the only reason I did it was because my Athlon 64 was S754 and the mobo had an AGP slot and i figured it'd be better to just get a new computer altogether, rather than get a new mobo that supports S754 and PCI Express. Sadly, I feel that operating system wise, my Athlon runs smoother than my Conroe. Game wise, conroe runs very good.

How exactly does XP not support 4 cores correctly? The windows scheduler has no problem running on 2 or 4 threads.

If your Athlon 64 is running smoother than your conroe, then perhaps you did something wrong (or you have some flaky hardware). Maybe a $350 Conroe with a $50 mobo was a bad idea after all 😉

Uh, you can get an ASUS P5W DH Deluxe for $50??? Sorry, I don't buy products from street dealers. I guess you missed the signature.

Windows might recognize 4 cores, but does it optimize it's work load across the 4?

 

[Whizzard9992]

Uh, you can get an ASUS P5W DH Deluxe for $50??? Sorry, I don't buy products from street dealers. I guess you missed the signature.

Windows might recognize 4 cores, but does it optimize it's work load across the 4?

Don't take offense: it was said tongue-in-cheek I'm just saying that there's enough info out there stating that the core 2 is [much] faster, so if you're the exception to the rule, there's probably a reason. If it's faster in games and not windows, perhaps it's I/O (i.e. Drivers, controller, etc).

Try this.

Anyway, the Windows XP scheduler (the mechanism that handles scheduling thread quantums) is the same as windows [server] 2000 (and probably windows 2003). The OS will scale to something like 16 or 32 processers very well. Any processor limitations imposed on XP are artificial. For example, XP home limits you to 1 processor (two if hyper-threading). This is a check that is placed on the scheduler, but the multi-proc HAL is the same.

The same goes for 4-core systems. XP will limit the number of physical processors as of SP2, but will recognize and use multiple hardware threads/cores.

In short, yes, XP is optimized for multi-core systems. You can even run a quadro with XP home, which is limited to one processor, and still use all 4 cores (granted you have SP2 installed). The code that distributes the tasks is the same (the kernel's scheduler) whether you have 1 processor with 4 cores or 8 processors with 1 core each.

( link )

 

[Eurasianman]

ic... do they by any chance have a program that I can use to will allow me to assign certain tasks to one of the two processors? Or is it faster for a program to spread across two, like converting from .avi to DVD format...

I was thinking about being able to conver a movie on one core while being able to play a game on the other core...

 

[bga]

Windows might recognize 4 cores, but does it optimize it's work load across the 4?

Yes, Windows is in fact a quite good multi-tasker, multi-processor operating system. I have tried up to 4 cores, and if you have a lot of treads, Windows distributes them very well, with no problem between the different applications. Sometimes a multi-threaded application can have internal problems, due to race, and dead-lock problems - hardly the fault of Windows. If you want a single application to take advantage of extra cores, the application typically needs to be rewritten. It is normally not a trivial task to divide the work between several treads, unless it is for something relatively simpel as a background printing process. There are very few real multi treaded programs on the market today.
A can remember some tests which documents that Windows scales well up to 8 cores, but then gets decreasing returns for every added core.