That was the case for older CPU's but the i3/i5/i7 benefit from both, SpeedStep is better for changing the multiplier/voltage but C State has additional benefits on the new Intel CPU's, instead of the whole CPU either being on/off/idle parts of the CPU can now be turned on/off or set to idle and this works in conjunction with intels Turbo Mode.

So basically they did do the same job but there are benefits to having both on when it comes to the new i3/i5/i7 CPU's.

you will want to set CxE Function to C6 to get these new benefits alongside having SpeedStep enabled (they can work independent of each other but its best to have both enabled, be warned though with newer EVGA BIOS's having CxE Function enabled will allow the higher Turbo Mode multipliers to kick in and could make your OC unstable, if this is the case disable CxE Function but you could keep SpeedStep enabled if it still works, on the X58 Classified the voltage part of SpeedStep does not work with a manually inserted Voltage, it does however still work on the E758 3X SLI board with a manually inserted vCore voltage, this is just due to the components used and how the boards are set-up due to the segments they are for, Classified being a primarily overclocking board when power saving features are secondary.  There are still work around for the X58 Classifieds using the ECP, this should allow you to OC the CPU but use an AUTO voltage which would allow the voltage part of SpeedStep to work, for further information on this please PM me)

seems on the P55 boards they have seperated CxState and C1E Funcion, they are both basically the same thing but C6 is the new state that works with the new intel achitecture (i3/i5/i7)

With C-state tech [enabled] and C6 [enabled] i would assume that would mean you had C-state C6 enabled which works in conjunction with turbo Mode to allow the higher turbo Multipliers.

The P55 BIOS is totally different to the X58 in regards to C-States, there is only 1 option related to this function in the X58 BIOS, CxE Function which can be set from C1, C1E to C6, the P55 seems to have alot of sub categories/features for C-States that i am not familiar with yet.

i do have a P55 FTW incoming so hopefully i can brush up on that board.

Maybe someone whos more familiar with your actual boards BIOS can give you a better idea of whats going on.

Turbo Mode is highly dependent on SpeedStep and the C-State packages - it won't function without them on a P55.
 
On my MOBO, there seems to be a 150 MHz BCLK limit to Turbo Mode with this Xeon. I imagine since Nehalem shares common traits across the board this holds true for most of the rest. Anything that reaches this frequency, whether it be a setting in BIOS or the usual frequency drifts in clocks, will disable it. I've been able to get 148 Mhz to stay, but usually use 147 on a regular basis.
 
Everything comes together quite well in Turbo, even with the OC, in terms of power conservation and energy distribution. The slightly over 10% OC that 147 MHz BCLK only raises the CPU's TDP by about 5W - all the way up to a whopping 95 (that's a joke, yes).
 
CineBench probably provides one of the more "in-your-face" metrics when you demonstrate the single-core performance. The MP ratio drops from an average of 5.05 to about 4.35. meaning the single core can now approach almost 25% of the total workload in the CPU's working the lower freqs.
 
Have a thread going over at the nVidia Forums now comparing P55 and X58 performance on a clock-for-clock basis, as well as a couple of other threads, that's delaying me a bit, but I intend to investigate the Turbo Mode at these levels a lot more fully.
 
I'm in no way dissatisfied or disillusioned about the P55's performance. I'm not changing for quite a while from here...

And does it kick into the 26x multi at idle? For that matter, will it give you anything other than perhaps the 22x multi?
 
Does it utilyze SpeedStep's half-multis?
 
How's your power envelope - which is one of the primary goals of Turbo. Bet you're nowhere near even 100.
 
And does it idle like this?...
 
 

And that's what I've been working to determine - is it a MOBO limitation or a CPU?
 
The 162 MHz spot is the EXACT place I'd love to be for lotsa reasons that aren't relevant to this discussion, and it's frustrating to get that close and not be able to grab it.
 
Right now, the only thing I've taken manual control of are the RAM voltages (set at 1.65V) and the BCLK freq, leaving everything else on AUTO or ENABLED to see if I can keep all of the features runnng. Everything but ECC memory is enabled. (edit - save for Spread Spectrums)
 
But that may be part of the problem in my case: right now, until I can get some 1156 DDR3, I'm using carry-over Mushkins. They're rated at 8-8-8-7 for 1600, but at 1.85V. I'm running JEDEC so I can set 1.65V and keep them running.
 
I imagine that you may not hit the 26x multi because of the power envelope's profile - that BCLCK would set the core at over 4.3 GHz, requiring considerable current and voltage.
 
I'd love to know what your settings are so I could test them.
 
Normal OCs are no problem. This chip will pull 22x at 189 for 4.17 and barely hit 80C on LinPack, HT enabled. In fact, I'm making a clock-for-clock comparison against an X58 rig over at the nVidia Forums with no problems.
 
But I really want to take this Turbo Mode as far as I can because I can see its benefits. Especially in a rig running four GPUs and today's limited-thread app market...

A straight OC isn't the objective, and heat isn't the problem. I'm also at least conscious of the CPU classes and hierachies. I asked because I wanted to see if HT was the deciding factor - one more piece to the puzzle as the clues collect.
 
I'm not Shamino or trubritar, in that I don't carry a 3-Way SLI w/PhysX for framerates. I'm more the Image Quality type. Also carried considerable weight in the decision to go P55 rather than X58. Turbo mode gives me both the performance levels I seek and the energy profile I can really use with this rig.
 
What's so bad about that?
 
And I have a Xeon for its long-term sustainability and durability. I'm actually getting tired of replacing rigs.
 
But that's OK. I've learned a lot in the three weeks I've been working with this setup, so I guess I can continue on just as well from here.
 
Thanks for letting me stop by...
 

I'm looking at my post, and it does look a little more like I'm leaving rather than trying to explain where I'm coming from - didn't mean for that to happen.
 
And the biggest reason I still cruise eVGA (half my equipment still) for this issue is that, while ASUS makes some terrific hardware, and even gear a good portion of it towards the enthusiast crowd, their support forums seem to center around installation and hardware diagnoses rather than OCing. Enthusiast questions seem to mostly lie fallow until they eventually fall to the bottom.
 
Over at the now-combined nVidia Hardware/SLIZone Forums, I'm participating in a thread that's exploring the differences between P55 and X58 on a clock-for-clock basis, and it's been almost as much a build log for me and the thread's OP. I've been able to figure out most of this while having a blast with this thread. Jesse's really enthusiastic, and it's hard to keep him bridled for the clock-for-clock bit, but he gets his spurts in between benches.
 
Gotten pretty familiar with the cards, too, and how to get a lot out of them. Even, at times, how to get more out of the driverset.
 
All of this is coming together now in what's turning out to be almost exactly what I was hoping for - a rig that's more like a Formula One than a straight muscle car. For me, it has to be able to do more than just run the oval gauntlet of benches. I need to handle The Ring as much as downtown Monte Carlo, with occasional side trips through Africa.
 
At the same time, it's starting to look like a network server, and starting to consume as much power.
 
For me, in a multi-core/hyperthreading environment I've long ago established that the 3.2-3.3 GHz breakpoint is the sweet spot for me. At the same time, do I really need even four cores going to support most games at the frequencies we all need for those? Not really. Turbo Mode and the use of C-State packages, SpeedStep and the ability to select and power down pieces of the CPU - and all else - is really a remarkable piece of work. That's why I'm focusing on it very hard right now. And trying to glean clues from any available source.
 
As you can see, I could go on, trying to explain, but this is your thread, and you have aims as well. So I'm gonna do what I had in mind when I originally wrote "Thanks for letting me stop by". I'd intended to "lurk" more than carry on as a distraction...
 

No problems here.
 
And thanks again...
 

SpeedStep is also used for thermal throttling and such, so yes - it's still active if enabled...
 

Good question dBurke, I learned alot with this thread. I have been into computers my whole (short) life, but this amazing i5 machine is also mu first build. I can't wait till my step-up 470's come in.

To the guy wondering about speedstep with a high OC, you can leave it on, just the transitioning from that high of a speed down to idle clocks can sometimes cause stability issues, thats why dudes leave it off when going for gold.

And 3.2 isn't a safe max per se, but it is the max you can push and still run a 2:10 memory multi to run 1600 memory at max spec. Any higher and you are either going to have to drop the memory to 2:8 or start overclocking some already very fast memory.

That combined with the performance vs power ratio of the i5 will keep my max at about 3.2 with turbo on, at least for now. I'm just happy with it now and my heaviest games barely load the cpu 30%, so I haven't pushed it past where it is.

And Jaffaman, welcome to what just might be the best forum on the internet.

I believe the article your refering to is Efficiency Explored: What's The Perfect Clock Rate For Your Core i5? The goal of their article was to maintain stock voltage, and Tom's believed to exceed 3.2GHz and keep Turbo enabled, would require a voltage increase above stock.

Deeper C states (C3, C6) include cache flushing routines that allow the cache to be put into park sleep, because the charge cells do not have to be refreshed to hold data. Some of the other C-states are utilized with Turbo ramping, simply because it allows Intel to maintain a target TDP whilst one of the cores can be dynamically ramped in frequency.

Higher OC's with SpeedStep and C1E enabled can work provided the buck controller and PWM are capable of ramping VID dynamically under load (this is actutally VRD 11.1 spec anyway). Although the issue is that with higher overclocks the optimum low power and full load VID levels would need to be programmed on a per CPU basis (no two processors are the same).

Intel employs a small hold time with Dynamic VID to allow the PWM controller to ramp VCC before it is faced with a severe load. That's why having SpeedStep enabled does not generally hurt overclocking, although there is a very slight performance hit at times.

regards
Raja

I second that...

I read this thread only yesterday and i found out yor questions and answers very useful.
So allow me to post a doubt:
monitoring by CPUz some test overclocking my i5 750 CPU at 3200gHz with Turbo, SpeedStep, C1E and C-States all enabled it seems like to be a reduction in multi (from 20 to 9) in IDLE state but not a Vcore reduction too.
Is it possible?
Thx in advance and regards.
P.S. sorry for my language i'm writing from Italy