OSNews

You can't. The mini uses mobile chips, of which the new Yonah ones will not be as dramatic of an improvement.

Merom you mean? The improvement should still be pretty good... the opened up pipeline and bigger cache should offer some readily discernible speed increases.

I'd say Core(1) is probably about even with the Athlon X2s, so you'll probably see a significant boost by waiting. 20-30%? It will also be 64 bit, so I'd wait if you are going to keep it for years. If you upgrade every year or two, though, I'd just get something now. There's always something coming out in a few months that will be faster.

Yes... I recently bought a 2.0 yonah for my mac mini. I will wait 6 months and pick up a merom.

sure you can

http://www.macrumors.com/pages/2006/06/20060613185240.shtml

No, I was referring to him wanting to drop in a Conroe, which is the chip being reviewed. He can only use a current Yonah, Sossaman, or the upcoming Merom.

What about this one:
http://enthusiast.hardocp.com/article.html?art=MTExMCwxLCxoZW50aHVz...

Focussed on multimedia.Furthermore there's a comparizon with a AMD FX-63:-)

the performance edge over AMD in multimedia, the new chip obviously has increased that performance. Props to Intel.

I got a airplace engine as a CPU (If you could hear it, you would understand)...

Sounds like you need a new heatsink or a new power supply, not a new CPU.

P4 Prescot generates a lot of heat. To cool down a CPU, you need a good heatsink, good thermal paste & good fan. When the CPU gets at 60 celsion, the fan try to do faster to compensate... The fan going faster will create a lot of noise... The noise annoys me...

Core 2 Duo uses less power and therefore produce less heat (According to all reviews at least)... Having a CPU that generate less heat will allow me to have fan going slower... You see where I am going?

But what about the difference between Core 2 and Core?

The Core 2 has shared L2 cache.The Core has unshared L2 cache.

Which means one single core can consume up to 100% L2 cache (Core 2) when only one core is used.

see:http://www.overclockers.com.au/article.php?id=489587&P=2

Which means one single core can consume up to 100% L2 cache (Core 2) when only one core is used.

Also, i suppose that when a process is migrated from one core to another, this process can reuse directly the shared cache instead of refilling the cache of its core.
That is a great thing.

Edited 2006-07-14 15:55

The Core 2 has shared L2 cache.The Core has unshared L2 cache.

No, both of them have shared L2.

Anyway, what matters is price, performance, and power, not microarchitecture.

Anyway, what matters is price, performance, and power, not microarchitecture.

Agreed,$500 prize difference between the X6800 and E6700 is a lot of do whereas the performance is generally pretty the same.

No, the Core architecure uses separate L2 caches (2x 1MB and 2x 2MB), one specifically for each core. It's similar to putting two normal CPUs together into one chip package. Depending on the CPU, they even have to go across the FSB to access data in the other core's L2.

The Core 2 architecture uses a single, shared, 2 MB or 4 MB L2 cache. Each core can access the full 2 or 4 MB. The cores dynamically assign the L2 cache as needed to eachother. And they can access data that the other core put in the cache.

Big difference.

The only great performance change is that Merom will be 64-bit; Yonah is not.

Who voted this down? I was being realistic with that analysis - you have a problem with it, back up your vote with some reasoning.

Tomshardware gives a much better reivew of this.

All of the tests on THG show every model of the Core2 beating every model of the X2 and Fx, in every benchmark.

Thanks a lot for the link. I just saw another review at Techreport and so on, boy oh boy the number differences are quite amazing! Intel needs to be truly applauded for this quick turnaround. THeir Israel team is truly amazing.

HardOCP is intentionally choosing to GPU-limit their game benchmarks to demonstrate that there is little substantive difference in practice for the gamer. Even NetBurst processors will show little difference if you start playing at 1600x1200 with a single GPU. If you start looking at the overall benchmarks you will see that the Core 2 is clearly offering higher performance per clock with lower power consumption. If what you'll be doing is GPU-bound then it doesn't matter what you have performance-wise. Compared to Intel's performance with NetBurst, though this is quite an improvement and given the intended pricing makes Intel's offerings very competitive. That's now, and not at the end of 2007 or 2008. With the reports of air overclocking around the 4GHz mark, there is clearly much room for the processor to scale. The biggest problem with the processor until next year will be availability.

With a single GPU, pretty much every Core 2 Duo and X2 are going to be even in games. Even most of the Pentium Ds are fast enough. If you move to SLI/Crossfire, though, you'll start noticing differences.

A sad day for home users, gamers, and hobbyists, perhaps. Or, it could just mean cheaper Athlon64s for everyone.

Definitely a sad day for AMD-based laptop makers, but then, AMD has never really had a strong laptop CPU.

But, until Intel scraps their shared-bus chipset/memory layout, they will remain second choice for all multi-processor server setups. There's still nothing in Intel's lineup that can even match a lowly Opteron 2xx setup using single-core CPUs.

Hmmm, looks like you're right. I was confusing the Pentium D with the Core.