Looking at trying to build a new machine basically only for gaming (FPS and
MMOs), is there any advantage in this scenario to go quad over duo? At first
my plans was for a E44/4500, but it seems I can get a fairly cheap E6600
from eBay which makes me think that might be a better choise (double cache
and all). But how about Q6600? For a desktop machine doing very light word,
excel, some homepage editing and mostly games, would this processor justify
its higher price?
Thomas Andersson wrote:
> Hi!
>
> Looking at trying to build a new machine basically only for gaming (FPS and
> MMOs), is there any advantage in this scenario to go quad over duo? At first
> my plans was for a E44/4500, but it seems I can get a fairly cheap E6600
> from eBay which makes me think that might be a better choise (double cache
> and all). But how about Q6600? For a desktop machine doing very light word,
> excel, some homepage editing and mostly games, would this processor justify
> its higher price?
>
> TIA
> Thomas
In my mind, the justification for the Q6600 2.4Ghz quad (G0 stepping),
is that it overclocks to 3GHz without too much trouble. It is the
same price as a E6850 3GHz dual, but gives you quad cores. There
are at least two steppings for Q6600, and G0 draws 10 watts less
than the other one, and has a slightly better overclock.
Right now, the main value is bragging rights, as even games that
spawn multiple threads, are not capable of loading all cores
equally.
There are some multimedia applications that exhibit "perfect
scaling", because all the threads have equal functions and
equal workloads. Cinebench, for example, is a good one for
that. So it is possible for multimedia programs to load things
equally.
In the case of games though, you wouldn't expect the threads
to have equal functions (AI, rendering, physics), so there is
no reason for all threads to load their core to 100%. Only one core
will be 100%, and the other three are some lesser percentage
loading, when in a game.
If you only wanted to run the Q6600 at stock 2.4GHz, then
I'd have a lot more trouble making the decision between
it, and an E6850 dual core at 3GHz. The Q6600 can still beat
it, but perhaps only in something like Cinebench. Running the
Q6600 at 3GHz, makes the decision easier.
"Paul" <nospam@needed.com> skrev i en meddelelse
news:fj397e$l58$1@aioe.org...
> Yes, the Q6600 G0 results I've been reading, suggest a virtual
> guarantee of 3GHz. As good as my old Celeron 300 going to 450 :-)
Wasn't it 454?...Those were the times..50% overclock :-) AFAIR I got at at
514, but not that stable...
'Per Nielsen' wrote:
| > Yes, the Q6600 G0 results I've been reading, suggest a virtual
| > guarantee of 3GHz. As good as my old Celeron 300 going to 450 :-)
|
| Wasn't it 454?...Those were the times..50% overclock :-) AFAIR I got at at
| 514, but not that stable...
_____
In days of yore, the limitation for a good Celeron 300a was the PCI bus
speed. The IDE interface could begin to throw errors when the PCI bus speed
was above 37 or 38 MHz. That works out to about 504 MHz with a FSB of 112
MHz because there was no ability to lock the PCI bus frequency with the
chipsets available when Celeron 300a first came out. They had no 133 MHz FSB
with a 4:1 PCI divider capability. Consequently 504 MHz was about it for
the Celeron 300a unless you did not use a IDE hard drive. Maybe a SCSI
drive would have allowed a higher overclock B^)
If I remember correctly, the Celeron 300a came out at about the same price
in US dollars as the current price of the E4300. And the US dollar has
inflated by what, maybe 50% since then?
According to the invaluable Wikipedia, the Celeron 300a [Mendocino, 250 nm
feature size, 300 MHz speed, 66 MHz FSB, 128 KByte L2 cache, 4.5x
multiplier, 2.0 volt core, 19 Watt TDP] came out in August 1998.
Does anyone have a Celeron 300a computing power comparison that could be
extended to compare a Celeron 300a to a Core 2 Duo CPU?
Phil Weldon
"Per Nielsen" <dazzler53@hotmail.com> wrote in message
news:47581bca$0$21932$157c6196@dreader1.cybercity. dk...
| "Paul" <nospam@needed.com> skrev i en meddelelse
| news:fj397e$l58$1@aioe.org...
| > Yes, the Q6600 G0 results I've been reading, suggest a virtual
| > guarantee of 3GHz. As good as my old Celeron 300 going to 450 :-)
|
| Wasn't it 454?...Those were the times..50% overclock :-) AFAIR I got at at
| 514, but not that stable...
|
|
Somewhere on teh interweb Phil Weldon typed:
> 'Per Nielsen' wrote:
>>> Yes, the Q6600 G0 results I've been reading, suggest a virtual
>>> guarantee of 3GHz. As good as my old Celeron 300 going to 450 :-)
>>
>> Wasn't it 454?...Those were the times..50% overclock :-) AFAIR I got
>> at at 514, but not that stable...
> _____
>
> In days of yore, the limitation for a good Celeron 300a was the PCI
> bus speed. The IDE interface could begin to throw errors when the
> PCI bus speed was above 37 or 38 MHz. That works out to about 504
> MHz with a FSB of 112 MHz because there was no ability to lock the
> PCI bus frequency with the chipsets available when Celeron 300a first
> came out. They had no 133 MHz FSB with a 4:1 PCI divider capability.
> Consequently 504 MHz was about it for the Celeron 300a unless you did
> not use a IDE hard drive. Maybe a SCSI drive would have allowed a
> higher overclock B^)
>
> If I remember correctly, the Celeron 300a came out at about the same
> price in US dollars as the current price of the E4300. And the US
> dollar has inflated by what, maybe 50% since then?
>
> According to the invaluable Wikipedia, the Celeron 300a [Mendocino,
> 250 nm feature size, 300 MHz speed, 66 MHz FSB, 128 KByte L2 cache,
> 4.5x multiplier, 2.0 volt core, 19 Watt TDP] came out in August 1998.
>
> Does anyone have a Celeron 300a computing power comparison that could
> be extended to compare a Celeron 300a to a Core 2 Duo CPU?
CPU for gaming rig, duo or quad? 
Linear Mode
