You're right Ulf :--)That also means that with HT disabled, the 2600k can be overclocked far much than 4.6 Ghz with the same temps.That's promising.

That's not necessarily true...if the limiting factor is heat, then that may signal that there's more potential headroom with HT shut down, but there are other aspects of the physics of the chip that can also limit things. My i7-975 won't go much higher than 4.6, even though I can keep it well below 60 deg with my industrial-grade water cooling system. When you get the operating frequency high enough up off the design freq, you can have parts of the circuit start to resonate, inductive coupling between components etc, that can be as much of a brick wall as high temps.I'm a bit sceptical that we're going to see many people make sustained ops at 5 GHz+ work with the SB CPUs. But a $325 CPU that can run at 4.6GHz ain't nothin' to sneeze at, either. RegardsBob ScottColonel, USAF (ret)ATP IMEL Gulfstream II-III-IV-VColorado Springs, CO

As I said, I had some difficulties with memory. I started off with 4x2GB of Mushkin Ridgeback DDR3 1600 (6-8-6-24). However, with 4 modules installed, they would only run at DDR3 1600 (8-8-8-24) regardless of what I did to improve the timings. Once again, I even had a mushkin rep helping me out. This may or may not be a problem with P67 mobos or immature bioses. In Sandy Bridge's defense, it's pretty well known that having all memory slots on any motherboard introduces instability and usually requires looser timings as well as more voltage. With 2 modules, they ran at 1600 (6-8-6-24) no problem. Today I picked up 2x4GB of Corsair XMS3 DDR3 2000 (9-10-9-27) from Frys - figuring I could just return them if I didn't like them. Of course, Sandy Bridge will not run memory at 2000Mhz, so I had to run them at 1866. My computer would not boot with them at 2133, but I will be trying. Anyhow, now you guys get the benefit of seeing the difference between two very different sets of RAM on an identical system with identical CPU clock! Here are the MaxxMEM2 results:2x2GB Mushkin Ridgeback DDR3 1600 (6-8-6-24)4x2GB Mushkin Ridgeback DDR3 1600 (8-8-8-24)2x4GB Corsair XMS3 DDR3 1866 (9-10-9-27)It would appear that the Corsair is best. If I can get 2133MHz out of it, then it should really pull ahead. As for how this will affect FSX... do we ever really know?! All we can do is get the fastest hardware on paper and pray.

I'm a bit sceptical that we're going to see many people make sustained ops at 5 GHz+ work with the SB CPUs. But a $325 CPU that can run at 4.6GHz ain't nothin' to sneeze at, either.

Well said. It's perfectly reasonable with a golden chip, but most of us will not be running 5GHz 24/7 unless BIOS revisions are capable of miracles... or unless it is deemed that 1.5v really is safe. Anyhow, I'm asking for a .4GHz miracle.EDIT: lowering CPU PLL from 1.8v to about 1.76v may allow you to lower vcore by about .03v. Just found this here. I'm currently stress testing my CPU at 4.6GHz at 1.32v (vcore is set at 1.28 in BIOS) and observing better temperatures as well. My miracle is coming true!

In fact there seems to a problem with some 2500K processors not being able to go over 4.4 GHz. Thats why OverclockersUK have changed their guarantee on pre-OC'ed bundles from 4,6 to 4,4 GHz.http://forums.overcl...87#post182272872600K are still going at @4.6 GHz though.Good info OC'ing on page 1 of this thread.@cmeeks: Which voltage did you use to test the 2 RAM sets?Nailing down the best RAM combination of voltage, size, frequency and latency looks to be one of the difficulties with Sandy Bridge.

I believe I was running the Mushkin at 1.54v. I haven't tried lowering the voltage on the Corsair yet. I'm finally trying to fine tune a 4.6GHz overclock, so I don't want to introduce any more variables than I have to. I would recommend 2x4GB DDR3 2000+ with Cas 9. I wouldn't worry about finding low voltage RAM. I think you can run most 1.65v sets below 1.6v easily. I will undervolt my Corsair tomorrow and let you know how it works out.

In fact there seems to a problem with some 2500K processors not being able to go over 4.4 GHz.

I think I'm one of them - it's incredibly frustrating. Has been running FSX nicely at 4.4 though.

Again why would anyone buy a 2500K? It's already started life as a defective 2600k... They disable the HT for the same reason we do. I know "hyper threading" is not needed, but at the same time, you have 30% of your core missing.

Again why would anyone buy a 2500K? It's already started life as a defective 2600k... They disable the HT for the same reason we do. I know "hyper threading" is not needed, but at the same time, you have 30% of your core missing.

Because I'd be paying $100 on something I'd never use.

Really you couldn't use the extra .4ghz? Baseball uses 1 ball too.. but you don't see any one armed players now do you?

Really you couldn't use the extra .4ghz? Baseball uses 1 ball too.. but you don't see any one armed players now do you?

The 2500K has a stock clockspeed of 3.3GHz while the 2600K runs at 3.4GHz.

Hi Max,The 2500k has a stock clock of 3.3GHz, and the 2600k has 3.4GHz, however, the 2500k is limited with it's overclock, with the highest that has remained stable and tested being around 4.4GHz due to voltage limits. The 2600k, meanwhile, can be overclocked to 4.8GHz, whilst remaining within operational voltage limits.

however, the 2500k is limited with it's overclock, with the highest that has remained stable and tested being around 4.4GHz due to voltage limits. The 2600k, meanwhile, can be overclocked to 4.8GHz, whilst remaining within operational voltage limits.

This is not true. P67-based motherboards shipped with BIOSes that did not support PLL voltage adjustment, a critical component when overclocking Sandy Bridge to high levels. Both chips are capable of ~5GHz (give or take) with new BIOSes that address this issue. As for which chip overclocks better, most reviews pit them at about the same level, sometimes the 2500k OCs higher, sometimes the 2600k. It really all comes down to variation between dies. The 2600k may give a slightly better chance of overclocking on average, simply because it is binned higher and therefore likely to be better silicon, but there is no guarantee.

Alex : I'd worry more about what Intel did to that chip to get it to work in the first place.

Alex : I'd worry more about what Intel did to that chip to get it to work in the first place.

One characteristic of the semi-conductor industry is that If a chip ships with a feature disabled it doesn't necessarily mean that portion of the chip is broken. Often times you have more chips that would meet a certain bin level than you actually need, and have a demand for more volume downstream so you turn on a fuse or two and voila, you have a lower SKU. Case-in-point: look at all of the Phenom II X2s and X3s that are actually quad core chips natively with a core or two fused off. Many users have been able to enable these cores through motherboard BIOS overrides. On the Intel side you can look at the recent introduction of a retail "upgrade" program which allows purchasers of qualifying OEM systems using Core i3 processors to unlock 1MB of L3 cache and HyperThreading for a modest fee via software.

Really you couldn't use the extra .4ghz? Baseball uses 1 ball too.. but you don't see any one armed players now do you?

This has nothing to do with this thread, but this is a bad analogy. There was a pro pitcher a few years ago that was missing half his left arm. He was good,too.Bob

This has nothing to do with this thread, but this is a bad analogy. There was a pro pitcher a few years ago that was missing half his left arm. He was good,too.Bob

And he could pitch two balls at once? Thanks for the PC lesson, i'm sure you are the type that would get upset if I called a slow computer chip retarded.