So what's going to be the most cost-efficient Conroe?

[Skott]

Originally posted by: Athlongamer
I'm pretty sure physics cards aren't going to be what they're all hyped up to be.......they're are a few benchmarks out there that you can see screen shots of Farcry I think, although it does look somewhat better, I don't think it's that big of a deal. DirectX10 release will play a MUCH bigger role then a phs-x card. But anyway, Iplan on getting the E6700 you guys think it's worth it?? Or are you guys saying I can achieve the speeds of an OC'd E6700 with and OC'd E6600?? Cuz if so thats a hell of an OC....i'm a bit cornfused

Yeah, right now these cards arent performing any better than not having a physics card. In fact in one review I saw it actually under performed so I dont see anything happening till 2007 possibly and only one game even supports it. Only 6 months left in this year so not sure if thats enough time for them to improve these cards. Game developers need to support them as well but I'm guessing they wont till the hardware technology is more improved.

 

[Athlongamer]

Originally posted by: Skott

Originally posted by: Athlongamer
I'm pretty sure physics cards aren't going to be what they're all hyped up to be.......they're are a few benchmarks out there that you can see screen shots of Farcry I think, although it does look somewhat better, I don't think it's that big of a deal. DirectX10 release will play a MUCH bigger role then a phs-x card. But anyway, Iplan on getting the E6700 you guys think it's worth it?? Or are you guys saying I can achieve the speeds of an OC'd E6700 with and OC'd E6600?? Cuz if so thats a hell of an OC....i'm a bit cornfused

Yeah, right now these cards arent performing any better than not having a physics card. In fact in one review I saw it actually under performed so I dont see anything happening till 2007 possibly and only one game even supports it. Only 6 months left in this year so not sure if thats enough time for them to improve these cards. Game developers need to support them as well but I'm guessing they wont till the hardware technology is more improved.

Yeah, I agree

but I still need some clarification because...
I plan on getting the E6700 you guys think it's worth it?? Or are you guys saying I can achieve the speeds of an OC'd E6700 with and OC'd E6600?? Cuz if so thats a hell of an OC....i'm a bit cornfused

 

[Skott]

Too early to say IMO. Need to see actual production chips benchmarks and reviews. So far its all speculation.

 

[secretanchitman]

Originally posted by: Kougar

Eh, unless you're going to be doing stuff like movie editing, you won't really get the jump you're looking for.. Gaming still isn't that heavily CPU dependent, and with innovations like the physics card coming our way (assuming the games are properly coded), your CPU should last even longer..

I'm sorry but I completely disagree. Conroe is not simply bringing a handful of extra FPS to the gaming benchmarks, but more towards 20-50 and even higher above the FX-62. The whole arguement for skimping on the CPU to invest more for the GPU will likely not apply in many cases with the "Core" uArch. A 2.4ghz Conroe is able to tie or exceed a FX-62 in the gaming benchmarks.

Secondly, cache on Conroe is likely to be very different from both AMD's K8 uArch and even the Netburst based designs. While there are no benchmarks (That I know of) as of yet comparing the processor at different FSB's or cache sizes, based off the plethora of balanced and not so balanced benchmarks that are out there I am expecting the L2 cache size to have a noticable impact on performance.

Another way to look at it is to consider this... a processor that can achieve FPS rates 50+ above a FX-62 on the same vid card and same drives and same memory, will very likely be a very data hungry chip. It has to be fed somehow... and it has already been proven that the 4mb cache plays a very large part in helping to hide the increased memory latency from not having an onboard memory controller compared to AMD.

As the CPU mutlipliers are so low, I wouldn't expect the E4200 to OC to well. I would expect that Intel will bin down most of the worst case, yet still running, CPUs to this bin level. If it wasn't for Intel's strict binning standards, I might even say some E4200s might be faulty enough that they might not work at a full 1066FSB speed.

Regardless, I am personally torn between getting the E6600 2.4ghz 4mb part for $316, or waiting until January when I can buy a $999 quad-core Kentsfield, that has two Conroe dies in it, will have an unlocked multiplier, and should already have a high clock speed. With that line of reasoning though I could wait forever... But my own Northwood is holding up fairly well, so...

Edit: Have a linky to a massive compliation of data on Conroe. I hope I won't get into trouble for linking to THG here... :music: THG Forum Thread. While much of it is out of date (A0 Steppings), or simply not balanced benchmarks, or are Intel sanctioned black box benchmark bakeoffs... there are some good links there regarding Conroe.

well said. still, im aiming for the rig in my sig/E6600 2.4Ghz.

 

[Marmion]

but I still need some clarification because...

The E6700 has a 10x multiplier, as opposed to the E6600 having a 9x multiplier.
Its difficult to say at this point whether the 10x over the 9x multiplier is worth it given the extra $200 in 'price.'
Theortically you will hit a higher clock speed with the E6700, but again, it depends on your perceived value of the extra multiplier and extra $200.
Personally, I'd go with the E6600, like most others. In NZ, the difference in price is likely to be ~$400 which is way too much.
I see it like this: On air, looking at XS, on a E6600, I could probably get around 3Ghz probably on stock cooling (9x333). On a E6700, this maybe 3.3Ghz (10x333). In reality, the extra 333Mhz probably won't bring me any extra perceivable performance, and its definately not worth it for me (paying an extra $400 for a possible extra 333Mhz doesn't sound like value for money to me).

Obviously, I'm just speculating, but I doubt the extra multiplier is worth it for most people.