Sign in with
Sign up | Sign in
Your question

Intel's secret weapon - Turbo .. what?

Last response: in CPUs
Share
a b à CPUs
September 18, 2009 2:15:42 AM

We all know what Intel is claiming to fame right now is that thier cpus are such great overclockers, but what they are really hiding is thier true stock speeds. How can this be? easy, call it turbo boost when its running at its stock speed, but claim thier stock speed as being much less.

Think about it, are you buying an I5 750 because its a 2.66ghz cpu or because it will normally run every day without overclocking at actual speeds of 3.2ghz? So what does this really mean. Simply put, overclockers can brag about getting a 2.66 ghz cpu to run at 3.8-4.0 ghz. Actually they are overclocking thier 3.2 ghz to 3.8-4.0.

Those second numbers should sound familiar, because thats exactly what you can do with the PII 955. So what is this all about? The ability to sell you something amazing by stating it as being much less than what it actually is to begin with. If you don't believe me, go search out some results on the Intel I series processors with results with TURBO DISABLED. Its all about marketing to specific groups IE overclockers.

Basically you can look at the AMD cpus the same way, they just named it differently. Cool N Quiet will allow the cpu to run much lower than its everyday potential. AMD just marketed it for what it actually is, slowing the cpu down.

So here it is in a nutshell. buy a turbo boost cpu that will run above what its sold as, or buy a cpu that can run slower than its rated speed.

See the catch? If you didn't, read the last statement again.

a c 96 à CPUs
September 18, 2009 2:39:33 AM

noob2222 said:
We all know what Intel is claiming to fame right now is that thier cpus are such great overclockers, but what they are really hiding is thier true stock speeds. How can this be? easy, call it turbo boost when its running at its stock speed, but claim thier stock speed as being much less.

Think about it, are you buying an I5 750 because its a 2.66ghz cpu or because it will normally run every day without overclocking at actual speeds of 3.2ghz? So what does this really mean. Simply put, overclockers can brag about getting a 2.66 ghz cpu to run at 3.8-4.0 ghz. Actually they are overclocking thier 3.2 ghz to 3.8-4.0.

Those second numbers should sound familiar, because thats exactly what you can do with the PII 955. So what is this all about? The ability to sell you something amazing by stating it as being much less than what it actually is to begin with. If you don't believe me, go search out some results on the Intel I series processors with results with TURBO DISABLED. Its all about marketing to specific groups IE overclockers.

Basically you can look at the AMD cpus the same way, they just named it differently. Cool N Quiet will allow the cpu to run much lower than its everyday potential. AMD just marketed it for what it actually is, slowing the cpu down.

So here it is in a nutshell. buy a turbo boost cpu that will run above what its sold as, or buy a cpu that can run slower than its rated speed.

See the catch? If you didn't, read the last statement again.


CPUs today are almost all binned by thermal dissipation characteristics rather than by maximum attainable clock speed like they were in the not-so-distant past. The i5 750 certainly can run at 3.20 GHz on all cores, else it would not be guaranteed to be stable when Turbo Boost kicks in fully. However, it is only allowed to run two of those cores at that speed at any one time to limit thermal dissipation. Turbo Boost is simply a way to distribute the energy allowable by the thermal limits to maximize performance. If you can use 95 watts with your quad-core CPU, would you rather:

1. Determine how fast your CPU can run with all cores loaded and not dissipate more than 95 watts. Once you have figured out that speed, do not let any cores run faster than this speed.

2. Determine how fast your CPU can run with all cores loaded and not dissipate more than 95 watts. Repeat the test with three, two and one loaded core and then allow the CPU to run cores at those (higher) speeds during those loading conditions.

The Nehalems' (except the Xeon E550x series) approach is #2, while all other x86 CPUs use #1. The first approach will have the CPU running well under its TDP limit when it is not fully loaded, while the second approach will have the CPU always running nearer its TDP limit with full load on at least one core. If you value minimizing energy usage over performance, then you pick #1. If you value performance over minimizing energy usage, then you pick #2. Each approach has its merits in certain situations, for example, I'd want to minimize energy usage and thermal dissipation if I had a blade server in a colo facility as power and cooling are at huge premiums. I'd also want to minimize energy usage in, say, a firewall or other always-on, low-load server or an HTPC. In a desktop or workstation, thermal/energy issues aren't much of an issue and I'd want the performance instead.

Oh, and one last thing. Turbo Boost is not overclocking as Intel guarantees it can run at that speed under certain loading and temperature conditions. It's only overclocking if YOU set a CPU to run at speeds not guaranteed by the manufacturer. I think somebody should make a sticky about Turbo Boost in Nehalems as I am seeing this come up a LOT.
September 18, 2009 3:22:42 AM

noob2222 said:
We all know what Intel is claiming to fame right now is that thier cpus are such great overclockers, but what they are really hiding is thier true stock speeds. How can this be? easy, call it turbo boost when its running at its stock speed, but claim thier stock speed as being much less.

Think about it, are you buying an I5 750 because its a 2.66ghz cpu or because it will normally run every day without overclocking at actual speeds of 3.2ghz? So what does this really mean. Simply put, overclockers can brag about getting a 2.66 ghz cpu to run at 3.8-4.0 ghz. Actually they are overclocking thier 3.2 ghz to 3.8-4.0.

Those second numbers should sound familiar, because thats exactly what you can do with the PII 955. So what is this all about? The ability to sell you something amazing by stating it as being much less than what it actually is to begin with. If you don't believe me, go search out some results on the Intel I series processors with results with TURBO DISABLED. Its all about marketing to specific groups IE overclockers.

Basically you can look at the AMD cpus the same way, they just named it differently. Cool N Quiet will allow the cpu to run much lower than its everyday potential. AMD just marketed it for what it actually is, slowing the cpu down.

So here it is in a nutshell. buy a turbo boost cpu that will run above what its sold as, or buy a cpu that can run slower than its rated speed.

See the catch? If you didn't, read the last statement again.


Not another AMD fanboy who doesn't know what he's talking about... :sarcastic: 
Related resources
a b à CPUs
September 18, 2009 3:32:34 AM

Well said MU.

"We all know what Intel is claiming to fame right now is that thier cpus are such great overclockers" No.
"but what they are really hiding is thier true stock speeds" No.
"call it turbo boost when its running at its stock speed, but claim thier stock speed as being much less." No.

"Think about it, are you buying an I5 750 because its a 2.66ghz cpu or because it will normally run every day without overclocking at actual speeds of 3.2ghz? So what does this really mean. Simply put, overclockers can brag about getting a 2.66 ghz cpu to run at 3.8-4.0 ghz. Actually they are overclocking thier 3.2 ghz to 3.8-4.0." Once again no.



"Those second numbers should sound familiar, because thats exactly what you can do with the PII 955. So what is this all about? The ability to sell you something amazing by stating it as being much less than what it actually is to begin with. If you don't believe me, go search out some results on the Intel I series processors with results with TURBO DISABLED. Its all about marketing to specific groups IE overclockers." Again, No. All of you fanboys need to stop it already with this turbo disabled BS. It is a feature of the chip. It should NEVER be turned off if the cpu is running at its stocks specs. NEVER.

Basically you can look at the AMD cpus the same way, they just named it differently. Cool N Quiet will allow the cpu to run much lower than its everyday potential. AMD just marketed it for what it actually is, slowing the cpu down. Again. NOOOOOOOOOOOO. Intel also has this. Its called speedstep. AMD cpus at DO NOT do the same thing and no its not just a different name.

"buy a turbo boost cpu that will run above what its sold as, or buy a cpu that can run slower than its rated speed." =EPIC FAIL. Your entire post was a epic fail.


"See the catch? If you didn't, read the last statement again." You need to read MU_Engineer entire post again, and again.



a b à CPUs
September 18, 2009 4:35:10 AM

this was never about the architecture or how intel cpu works. This is about Intel's marketing. Its sheer genious to sell something and advertise it as being less than what it is. Go to best buy and talk to their sales people and see if they don't advertise it as a 2.66 ghz cpu being faster than amd's 3.2 ghz.

Who cares about how it works or why. thats not the point. Intel sells a 2.66 ghz cpu that never runs that slow. Thats marketing.

Go to newegg and see how many people put 1 egg on an amd cpu because they had no clue that cool n quiet enabled would show the cpu as a 1.6 ghz on bootup.
a c 123 à CPUs
September 18, 2009 4:57:24 AM

Or it could be what Intel ha sbeen doing all along and just selling CPUs that were able to OC damn well.

Hell even the C2Qs (Conroe) OCed pretty good for a first gen quad and then the C2Qs (45nm) OCed even better at less heat output.

Its a way to satisfy both the regular joes with a CPU thats powerful enough to do anything thats thrown at it by a normal person and satisfy the enthusiasts need for speed.
a b à CPUs
September 18, 2009 5:06:20 AM

noob2222 said:
this was never about the architecture or how intel cpu works. This is about Intel's marketing. Its sheer genious to sell something and advertise it as being less than what it is. Go to best buy and talk to their sales people and see if they don't advertise it as a 2.66 ghz cpu being faster than amd's 3.2 ghz.

Who cares about how it works or why. thats not the point. Intel sells a 2.66 ghz cpu that never runs that slow. Thats marketing.

Go to newegg and see how many people put 1 egg on an amd cpu because they had no clue that cool n quiet enabled would show the cpu as a 1.6 ghz on bootup.


It will run at 2.66GHz sometimes though. On a 4-threaded program that fully loads all FP and integer units, it will only run at 2.66. It only clocks higher when it is below the rated thermal envelope, and therefore has some additional room to bump up the clocks. Selling it as a 3.2GHz CPU would be quite dishonest, since it does not run at 3.2GHz most of the time (as can clearly be seen by comparing its multithreaded benchmarks to an i7-965 with HT and turbo turned off). It is called a 2.66GHz CPU because it will always run at a minimum of 2.66GHz, regardless of the load. Sometimes it will run faster, yes, but that doesn't make it a "3.2GHz" CPU.
September 18, 2009 5:06:22 AM

Sure, technically it's marketing. It's another feature that they can put on the box and make it look all pretty. However, you can't say that it's not useful and resourceful. Turbo chooses how the CPU should be binned at different work loads to maximize performance under the wattage spec. And maybe you should look at charts where turbo is disabled again. The i5 750 may not win even half the tests, actually it loses by quite a lot, but the fact that in even some of the tests it can keep pace with the pII 965 is incredible. Not to mention that once they are at the same clock the i5 750 beats the pII 965 in almost every benchmark. Therefore, with turbo disabled and with the clock equalized, we can definitively say that the i5 architecture, or the nehalem architecture in general, is superior to the pII's architecture. There is no smoke and mirrors when it comes to this fact because it is simply true.
a b à CPUs
a b å Intel
September 18, 2009 6:15:36 AM

cjl said:
It will run at 2.66GHz sometimes though. On a 4-threaded program that fully loads all FP and integer units, it will only run at 2.66. It only clocks higher when it is below the rated thermal envelope, and therefore has some additional room to bump up the clocks.
Whether it ever runs at 2.66GHz depends on the heatsink. I've run my Core i7 920 under Prime95 for several hours and it's never dropped out of turbo mode because the temperatures have never hit the throttle point. In fact I've never actually caught my system running at 2.66GHz, it's either under load in turbo mode or it's idling along at 1.6GHz...
a b à CPUs
September 18, 2009 12:49:49 PM

I'm sorry but I fail to see what the OP is on about here. What exactly is the problem? Intel's Nehalem chips have the ability to ramp their performance up for short bursts when they need it. That's not overclocking, that's not marketing, that's not lying or deprecating their product.

It's just a feature.

BMW say their M5 has a 360bhp engine, but engaging sports mode ramps it up to 500bhp with computer-adjusted suspension. Are BMW lying or overclocking their car?

No. It's a feature.

So unless you actually have a point to make or a genuine grievance, please clarify exactly what the issue is.
September 18, 2009 1:13:31 PM

Hmmm, cool n quiet in here today? I bet if you use that, your cpu downclocks as well lol
The only problem I have is, if the cpus are the same as they were from last gen, same price as well, same clocks, and THEN have turbo? its all good, but, having a more mature process and having lower clocks that turbo up to and slightly past last gens clocks for same pricing, thats a no no
September 18, 2009 1:28:13 PM

I think there's a term for that... hold on... I can't remember now... what's the phrase..

Oh yeh, its called "technology advancement". Get over it.
a b à CPUs
September 18, 2009 4:42:57 PM

Oh stop this nonsense. So now its just a marketing gimmick. And its intels job now to actually teach people at best buy about the products in the stores. Do not bring in the entire stupid person in the store point. Do not bring up the entire how a 3ghz cpu is slower then a 2.66 either. In P4 days Intels where clocked way higher and where slower. Then core2 and so on...It is nothing new.

And with turbo Intel MUST say its 2.66ghz cpu. With your logic what do want them to do, put 3.2ghz on the box? And yes it does run that slow. It will ALWAYS be able to run at that spec. Where as 4 cores will not be...Nevermind i give up.. Waste of time.
September 18, 2009 7:04:47 PM

It seems to me itd be just as easy to give it max clocks UNTIL it has to downclock, isnt that what it does?
So, why name it at the lowest? I mean, itll give max cloxks first, then everything else next, or its not working properly.
So, instead of saying its using the minimum of say 2.66, show me where thats all it ever runs at ever, maybe Im mistaken here? If it always runs higher than 2.66, then it should be listed that way? Whether its 1,2 or even 4 cores?
a b à CPUs
September 18, 2009 8:06:20 PM

Jaydee. It doesnt give max clocks until is has downclock. It basically does the opposite. It starts at the bottom. Idle with speedstep and all the other power saving features. When the cpu is given a workload it will adjust its clock speed upwards at the same time turbo may kick in. How high the clocks go depends on workload. Factors such as how many cores the app uses. How much stress the app puts on the cpu. Which determines how much power/heat the cpu uses/generates. Now it does give the max clocks until is has to downclock as you stated. But even then thats a tricky way of putting it. As you "downclock" you may actually end up raising the actual clock speed. In the case of the i860. 2 cores loaded max turbo is 3.33ghz. One core is 3.46ghz. Now if this cpu needs to throttle down what will it end up doing? It can either clock back down to its base clock of 2.8ghz or "downclock" to one core active at the higher clock speed.


Basically think of it like a power saving feature. Depending on the workload it will do whatever it can to get the work done in the shortest time possible without going over the TDP.


When AMD cpus hit the market with this feature I will be surprised if they dont market it basically the exact same way.



September 18, 2009 8:22:36 PM

MU_Engineer said:
CPUs today are almost all binned by thermal dissipation characteristics rather than by maximum attainable clock speed like they were in the not-so-distant past. The i5 750 certainly can run at 3.20 GHz on all cores, else it would not be guaranteed to be stable when Turbo Boost kicks in fully. However, it is only allowed to run two of those cores at that speed at any one time to limit thermal dissipation. Turbo Boost is simply a way to distribute the energy allowable by the thermal limits to maximize performance. If you can use 95 watts with your quad-core CPU, would you rather:

1. Determine how fast your CPU can run with all cores loaded and not dissipate more than 95 watts. Once you have figured out that speed, do not let any cores run faster than this speed.

2. Determine how fast your CPU can run with all cores loaded and not dissipate more than 95 watts. Repeat the test with three, two and one loaded core and then allow the CPU to run cores at those (higher) speeds during those loading conditions.

The Nehalems' (except the Xeon E550x series) approach is #2, while all other x86 CPUs use #1. The first approach will have the CPU running well under its TDP limit when it is not fully loaded, while the second approach will have the CPU always running nearer its TDP limit with full load on at least one core. If you value minimizing energy usage over performance, then you pick #1. If you value performance over minimizing energy usage, then you pick #2. Each approach has its merits in certain situations, for example, I'd want to minimize energy usage and thermal dissipation if I had a blade server in a colo facility as power and cooling are at huge premiums. I'd also want to minimize energy usage in, say, a firewall or other always-on, low-load server or an HTPC. In a desktop or workstation, thermal/energy issues aren't much of an issue and I'd want the performance instead.

Oh, and one last thing. Turbo Boost is not overclocking as Intel guarantees it can run at that speed under certain loading and temperature conditions. It's only overclocking if YOU set a CPU to run at speeds not guaranteed by the manufacturer. I think somebody should make a sticky about Turbo Boost in Nehalems as I am seeing this come up a LOT.



^^^ = pwned

Forgive the laziness, but he simply stated all that I would've, but was too tired to type.
a b à CPUs
September 18, 2009 8:46:00 PM

+1 for MU_Engineer. The OP is a troll. I vote for the use of a ban hammer.

a b à CPUs
September 18, 2009 9:05:29 PM

Remember the good old days of the Intel 486 ? Just push the turbo button on the front of the case, and I got from 400 to 433 Mhz in a blink :D  ha well, times change :) 
a b à CPUs
September 18, 2009 9:25:06 PM

jimishtar said:
Remember the good old days of the Intel 486 ? Just push the turbo button on the front of the case, and I got from 400 to 433 Mhz in a blink :D  ha well, times change :) 


lol yeah. 33 mhz doesn't sound like much know but back then that was a lot of power. Now where the turbo button on the new cpus? :lol:  jk.
September 18, 2009 9:28:37 PM

yomamafor1 said:
Not another AMD fanboy who doesn't know what he's talking about... :sarcastic: 



*SARCASM NOTED

Another narrow-minded single sentence from a Intel Fanboy loser. PII can OC to 6Ghz or more and yet it costs
a fraction of what Intel's offering does. AMD gives you more bang for your buck. It is proven.

Speeds now are fast enough to do virtually anything a user wants. For a majority of computer users, COST/PERFORMANCE is
the deciding factor for purchasing a new machine. Especially now, since the economy is in the drink.

http://www.youtube.com/watch?v=dwcKyrHHQac
(January 08, 2009)

Intel is a joke! Just like you are! Wasting money is not something I enjoy doing.

AMD is the only company that LISTENS to its customers and saves them money.

How many of AMD's innovations has Intel copied? LMFAO

BTW
Sarcasm is a sign of a weak mind. It's no wonder all you offer to the argument is a single sentence.
a b à CPUs
a b å Intel
September 18, 2009 9:45:07 PM

Shadow703793 said:
The OP is a troll.
Actually, I disagree. His point is a valid one, with modern processors that run at various frequencies, the frequency on the box is purely a marketing label. So what is the best strategy for marketing your processor?

Take disk drives as an example. The transfer rate of a disk drive depends on whether the drive is accessing the outermost tracks or the innermost tracks. Do the disk manufacturers advertise the lower transfer rate, or even the average transfer rate? No - they advertise the "Peak" transfer rate.

So why wouldn't Intel market it's chips with a "peak" clock speed? It's an interesting question, IMHO.
a b à CPUs
September 18, 2009 9:52:04 PM

Ahh I see this thread has brought out one the old AMD trolls.
a b à CPUs
September 18, 2009 9:52:19 PM

#1. if Intel sells an I5 at 2.66 GHz, and the cpu is actually 3.2, that is better for the customer, u get more and payed for less :D  thats a good thing in my book

#2. be realistic. intel is the Empire. Leader. Amd just tries to keep up with intel, and its strategy is lower prices. both companies just want our money, so amd fanboys, wake up, amd cares for your money, they care u dont give it to the Empire.

#3 The Empire ALWAYS strikes back :D  lol
a b à CPUs
September 18, 2009 10:00:56 PM

"For a majority of computer users, COST/PERFORMANCE is the deciding factor for purchasing a new machine."
Yeah that because amd cant win in performance and even then amd losing in the performance front due to intel core i5.

For that video you posted on here, It's a fail. I wont even have to pay that much to get that kind of quality on a i7. Most like 1,500.


"Intel is a joke! Just like you are! Wasting money is not something I enjoy doing. "

Thats funny. last i checked BOTH companies listen to there costumers.


"How many of AMD's innovations has Intel copied? LMFAO "
How many times has both sides copy each other? Seriously everyone copies each other.
a b à CPUs
September 18, 2009 10:08:12 PM

Not sure where you get your information from MU, but from what I read on Anandtech, Intel do not guarantee even 1mhz of 'turbo'.
September 18, 2009 10:28:33 PM

Unless Im mistaken again, as I said, cool n quiet does trhe same thing, starts out low, below listed clocks, nothing new there.
My point is, once its needed, turbo kicks in, period.
Show me where it doesnt?
You have links? Im not so certain it NEVER kicks unless its disabled, so therefore, its truly higher than the 2.66 its labled at.
Lets look at the Intel marketing approach to it, anyone with links regarding turbo?
If the approach is when needed, and the reviews/benches shows its always working, then its NOT when needed, its all the time, and therefore should be listed and promoted as such.
Since its not, its giving the buyer a supposed feeling of getting something for free, in a cool sorta , never been nerdy sorta way.
So, again, if anyone has links showing me where turbo ISNT and cant be used, Id like to see them.


PS Personally I really dont care, but this caught my attention, and would like to know now
a b à CPUs
September 18, 2009 10:36:31 PM

jimishtar said:
Remember the good old days of the Intel 486 ? Just push the turbo button on the front of the case, and I got from 400 to 433 Mhz in a blink :D  ha well, times change :) 


warmon6 said:
lol yeah. 33 mhz doesn't sound like much know but back then that was a lot of power. Now where the turbo button on the new cpus? :lol:  jk.



Asus Turbo-V can retask the power button for a quick OC. Looks like the Turbo button is back!
a b à CPUs
September 18, 2009 11:18:01 PM

http://www.intel.com/technology/turboboost/

It is NOT the same as cool and quiet. Cool and quiet can not do what turbo does.

It will kick in when it can.

What is so hard for you guys to understand? You can't list the cpu at the any of the various turbo speeds/core configs.

Turbo is on all the time. Its a feature of the chip. That does not mean the frequency of the chip is always higher then its base block.

The customer is not getting anything for free with turbo boost. It is a feature of the chip.

Maybe Intel should label the i5 like this. Quad core at 2.66ghz but depending on the conditions it can be a quad core at 2.8ghz a tri core with 2.8ghz a dual core @ 3.2ghz and a single core @3.2ghz

September 18, 2009 11:24:23 PM

MU_Engineer said:
CPUs today are almost all binned by thermal dissipation characteristics rather than by maximum attainable clock speed like they were in the not-so-distant past. The i5 750 certainly can run at 3.20 GHz on all cores, else it would not be guaranteed to be stable when Turbo Boost kicks in fully. However, it is only allowed to run two of those cores at that speed at any one time to limit thermal dissipation. Turbo Boost is simply a way to distribute the energy allowable by the thermal limits to maximize performance. If you can use 95 watts with your quad-core CPU, would you rather:

1. Determine how fast your CPU can run with all cores loaded and not dissipate more than 95 watts. Once you have figured out that speed, do not let any cores run faster than this speed.

2. Determine how fast your CPU can run with all cores loaded and not dissipate more than 95 watts. Repeat the test with three, two and one loaded core and then allow the CPU to run cores at those (higher) speeds during those loading conditions.

The Nehalems' (except the Xeon E550x series) approach is #2, while all other x86 CPUs use #1. The first approach will have the CPU running well under its TDP limit when it is not fully loaded, while the second approach will have the CPU always running nearer its TDP limit with full load on at least one core. If you value minimizing energy usage over performance, then you pick #1. If you value performance over minimizing energy usage, then you pick #2. Each approach has its merits in certain situations, for example, I'd want to minimize energy usage and thermal dissipation if I had a blade server in a colo facility as power and cooling are at huge premiums. I'd also want to minimize energy usage in, say, a firewall or other always-on, low-load server or an HTPC. In a desktop or workstation, thermal/energy issues aren't much of an issue and I'd want the performance instead.

Oh, and one last thing. Turbo Boost is not overclocking as Intel guarantees it can run at that speed under certain loading and temperature conditions. It's only overclocking if YOU set a CPU to run at speeds not guaranteed by the manufacturer. I think somebody should make a sticky about Turbo Boost in Nehalems as I am seeing this come up a LOT.


Sound logic from a person with a track record of neutrality and in-depth technical analysis.
a c 96 à CPUs
September 19, 2009 12:29:57 AM

Shadow703793 said:
+1 for MU_Engineer. The OP is a troll. I vote for the use of a ban hammer.



I think we need to make a sticky about Intel's Nehalem Turbo Boost, to tell the truth.
September 19, 2009 12:47:54 AM

jennyh said:
Not sure where you get your information from MU, but from what I read on Anandtech, Intel do not guarantee even 1mhz of 'turbo'.


Not sure where you get your information from (probably AMDzone), but from what I read on Anandtech, Intel does guarantee Turbo. That's why it is sold as a feature.

On the other hand, AMD does not guarantee "4Ghz overclocking", despite its previous Newegg advertisement suggested so.
September 19, 2009 12:49:36 AM

enigma067 said:
*SARCASM NOTED

Another narrow-minded single sentence from a Intel Fanboy loser. PII can OC to 6Ghz or more and yet it costs
a fraction of what Intel's offering does. AMD gives you more bang for your buck. It is proven.

Speeds now are fast enough to do virtually anything a user wants. For a majority of computer users, COST/PERFORMANCE is
the deciding factor for purchasing a new machine. Especially now, since the economy is in the drink.

http://www.youtube.com/watch?v=dwcKyrHHQac
(January 08, 2009)

Intel is a joke! Just like you are! Wasting money is not something I enjoy doing.

AMD is the only company that LISTENS to its customers and saves them money.

How many of AMD's innovations has Intel copied? LMFAO

BTW
Sarcasm is a sign of a weak mind. It's no wonder all you offer to the argument is a single sentence.


Because I don't want to waste my life trying to explain common sense to someone who practically lives in dreamland... the AMDFTWOMGPWNED Land.
a c 96 à CPUs
September 19, 2009 2:17:52 AM

jennyh said:
Not sure where you get your information from MU, but from what I read on Anandtech, Intel do not guarantee even 1mhz of 'turbo'.


They do not guarantee Turbo will work in all situations, but they do guarantee that it will work in certain situations as outlined in their documentation.
September 19, 2009 2:32:33 AM

Jaydee. It doesnt give max clocks until is has downclock. It basically does the opposite. It starts at the bottom. Idle with speedstep and all the other power saving features

Never said it didnt, was just confirming that yes, of course it starts at the bottom, like duh.
My question is, if its on almost always or always, how can there be a 2.66, besides the original timings?
You seem to not understand my question.
Again, if its on ll the time, and being used all the time, what ways does Intel promote it?
Is it straight up marketing, or something a lil more? Since its not added, its always on, just like speed step or as I said cool n quiet, and you have to shut it off for it not to work, Im more interested in how average Joe perceives this, and how its being marketed to him is all, and links would be appreciated.
Id like to see where it isnt kicking in, anyone? Links? We like real proofs right? All Im askin is links, as I really dont know, and not Intels stuff, but benches showing it NOT kicking in....
September 19, 2009 2:46:51 AM

Ive read about the P0 states, but isnt that almost always happening?, and again, Intel talk doesnt mean real world testing
a b à CPUs
September 19, 2009 3:26:24 AM

As far as them marketing it. This demo is on Intel's site. http://www.intel.com/technology/turboboost/index.htm?ii...

I have not seen any commercials or ads targeted at overage joe computer buyer. But in the past they have just been basically, hey its us Intel, we are inside computers...

As far as it not kicking in. Give it time. Hopefully when more people build or purchase a machine with the chips in it we can see. Would be nice to have a OEM machine and run it through benchmarks that stress all 4 cores and see if turbo kicks in. And if it does see how it fluctuates.

The best we can probably find at the moment are benchmarks that are known to max out all available cores with turbo shut off. Compare them with the same benches on the same system with the feature turned on.

Or just through as many things as possible at the cpu to get it work as hard as possible and see what the clocks are. Play a game, encode a video, anti-virus scan creat a rar archive and so on. Factor in how good your cooler is and how good your actual chip is(VID).

Its clear as day that i5s are the worst die from the batch.. Since the top chip the i870 runs stock a 2.93ghz. And can turbo boost all 4 cores+HT up to 3.2ghz. And still stay in the same power/heat range.

Or just get a i5 and run linpack on it. See if turbo clocks up. As I'm sure Intel test its cpu's with most the extreme testing possible to guarantee them.

September 19, 2009 3:46:53 AM

Fiction...

jennyh said:
Not sure where you get your information from MU, but from what I read on Anandtech, Intel do not guarantee even 1mhz of 'turbo'.



Fact...

MU_Engineer said:
They do not guarantee Turbo will work in all situations, but they do guarantee that it will work in certain situations as outlined in their documentation.






Jenny, keep in mind AMD may be implementing a very similar technology, which you will undoubtedly hail as innovative and cutting-edge.
September 19, 2009 3:53:33 AM

TY someguy, its sorta what I think as well, and a shame no ones done this, either here, or in reviews to a large extent.
I mean, as you say, this is a feature, and weve all read about cnq or speedstep, because thats easy. Doing what at least Id like to see, and others as well Im sure, wouldnt be so easy, and itd be nice to know which real world apps have P0 sends andeven synthetics, as itd give a deeper understanding of the tech being used
September 19, 2009 4:42:29 AM

I wonder how many people even know how turbo works on these new CPU,s. I don't but I am familiar on how it works on my Bloomfield and just doesn't work the same depending on setup. For the 21x multiplier I use it as a actual multi rather than as turbo meaning it will stay at 21 no matter the load or temps from a bios setting (before that I noticed it would throttle down when core temp would go over the 85~90° mark).

But I also have all power saving features enabled and for the multi and vcore to drop while idle and remain at a low milti and vcore I have to change the power setting in the OS to reflect that. In XP I use laptop/portable even tho I don't allow the HDs or monitor to shut down or system standby. If I kept it as Home/Office Desk or Always on (any non power saving scheme) at idle the multi and vcore bounce all over the place and is very seldom at minimum.

As for the single core multi of 22 I have only seen that appear with the power scheme set to max battery. I would guess it would show up also with only one core enabled but haven't tried.

I had to do the same with W7 also but don't remember the exact settings since it has far more settings. I just know by default it had the same idle problems with the unstable multi and vcore till I changed it. I also have never gotten the 22x single core turbo in W7, but since I am not running with max power saving settings I am guessing that is why.

Since the power scheme has an effect on turbo with the Bloomfields I wonder how much it will with the new Lynnfield's. None of the reviews I have seen have said anything about that, or even the older reviews on the Bloomfields.

I would guess that the majority of time that the 750 is under load it would run at the 3/4 core turbo of 2.8 and not at the 1/2 core turbo of 3.2.

I would also like to hear from other users on how their experiences with turbo and how it works for them and what loads it takes to get each level of turbo to engage and what power scheme they are running and how it affects the new Lynnfield CPU's.
September 19, 2009 6:00:31 AM

wwo you say a bad word against intel and you are an amd fanbpoy, you say something bad against AMD and you are an intel fanboy, so there no way yo win, this is what happens, the CPUs are guyaranteed (under warranty) to run at 2.66GHZ, they will run on 2 cores at 3.2 GHZ, that is also guaranteed, the i5 and i7 overxclcok themselves using turbo, this IS overclocking but NOT as we know it, we know over clocking as running faster than the manafacturer intended.
September 19, 2009 6:04:28 AM

enigma067 said:
*SARCASM NOTED

Another narrow-minded single sentence from a Intel Fanboy loser. PII can OC to 6Ghz or more and yet it costs
a fraction of what Intel's offering does. AMD gives you more bang for your buck. It is proven.

Speeds now are fast enough to do virtually anything a user wants. For a majority of computer users, COST/PERFORMANCE is
the deciding factor for purchasing a new machine. Especially now, since the economy is in the drink.

http://www.youtube.com/watch?v=dwcKyrHHQac
(January 08, 2009)

Intel is a joke! Just like you are! Wasting money is not something I enjoy doing.

AMD is the only company that LISTENS to its customers and saves them money.

How many of AMD's innovations has Intel copied? LMFAO

BTW
Sarcasm is a sign of a weak mind. It's no wonder all you offer to the argument is a single sentence.



actually no matter if your and AMD or and intel fanboy the fact is this the Phemon II X4 655 BE has the 3rd highest clock speed in history and to this date the core i5 or i7 dont maybe time will tell but right now the PII has the 3rd highest clock speed in history
a b à CPUs
September 19, 2009 10:17:04 AM

lashton said:
actually no matter if your and AMD or and intel fanboy the fact is this the Phemon II X4 655 BE has the 3rd highest clock speed in history and to this date the core i5 or i7 dont maybe time will tell but right now the PII has the 3rd highest clock speed in history

I'm pretty sure that there have been more than 3 P4s that exceeded 7.2GHz...
September 19, 2009 1:06:07 PM

Whats interesting is, the P4's werent single die for the duos, and much smaller and less dense as well.
Never has a cpu the size of P2 gone as fast, just like everyone shoots down IBMs claims of high clocks, only because its but a small section of a cpu when it was done, Id argue the same for the P4 compared to todays cpus as well.
Everyone knows density and overall size plays a huge role in overall clocks, and here P2 stomps the P4, which is barely 14%faster
September 19, 2009 2:41:20 PM

lashton said:
actually no matter if your and AMD or and intel fanboy the fact is this the Phemon II X4 655 BE has the 3rd highest clock speed in history and to this date the core i5 or i7 dont maybe time will tell but right now the PII has the 3rd highest clock speed in history


What's really the point? Yeh it can hit 6Ghz on LN2 with golden samples, but average blokes are struggling to get 4.0Ghz out of it on water.

Core i7 on the other hand can hit 4.0Ghz easy, on air. The new D0s can hit 4.6Ghz on air. So what's the point?
a b à CPUs
September 19, 2009 7:03:55 PM

yomamafor1 said:
What's really the point? Yeh it can hit 6Ghz on LN2 with golden samples, but average blokes are struggling to get 4.0Ghz out of it on water.

Core i7 on the other hand can hit 4.0Ghz easy, on air. The new D0s can hit 4.6Ghz on air. So what's the point?


don't quote worst case scenarios vs best case when you try to defend yourself.

The PII is limited in its overclocking to just over 4.0 ghz in 64bit os (probably due to the way it handles the 64 bit registers), in 32 bit os, they have been reported to reach 4.4ghz + on air cooling.

Water cooling is just as efficient as new air coolers unless your adding ice to your tank, go read some articles.

Sure, I7 can hit 4.6 with perfect ventilation and a massive heatsink while running 85+ Celcius in a temperature controlled room, Go to newegg and see what the average user can achieve without fear of frying his cpu. Most (key word is most) range from 3.5-4.0, same as the PII BE's. I am quoting newegg because the results are REAL WORLD, not lab controlled.

________ sorry, straying off topic..


Some very funny and interesting comments here, esp calling me an AMD fanboy for complementing Intel on thier marketing scheme.

Here is some testing I would love to see on this topic.

Intel only cpu, be it i7 870 or i5 750.

Test it as shipped
Test it turbo disabled
Test it overclocked as its rated turbo speed with turbo disabled (ie all 4 cores running at 3.2 ghz on the 750)

The results would probably be very interesting esp when comparing overclocked vs stock turbo.

Also speculating the results with turbo disabled will not even be close to any other configuration suggesting that the cpu never runs that slow.
a b à CPUs
September 19, 2009 7:40:34 PM

sminlal said:
Actually, I disagree. His point is a valid one, with modern processors that run at various frequencies, the frequency on the box is purely a marketing label. So what is the best strategy for marketing your processor?

Take disk drives as an example. The transfer rate of a disk drive depends on whether the drive is accessing the outermost tracks or the innermost tracks. Do the disk manufacturers advertise the lower transfer rate, or even the average transfer rate? No - they advertise the "Peak" transfer rate.

So why wouldn't Intel market it's chips with a "peak" clock speed? It's an interesting question, IMHO.

I guess you do have a point. But OP's tone made it sound like he's a troll.

Quote:

The PII is limited in its overclocking to just over 4.0 ghz in 64bit os (probably due to the way it handles the 64 bit registers), in 32 bit os, they have been reported to reach 4.4ghz + on air cooling

O RLY? A valid Linky?

Quote:
I am quoting newegg because the results are REAL WORLD, not lab controlled.

...most Newegg reviewers are stupid. Take newegg reviews with a grain of salt.

Look at this: http://hwbot.org/listResults.do?cpuModelId=1926&applica...
Every one above 4.25Ghz is on LN2.

Quote:

Sure, I7 can hit 4.6 with perfect ventilation and a massive heatsink while running 85+ Celcius in a temperature controlled room,

Whaaa??? You can run an i7 920 at 4.8-4.9Ghz with a good WCing loop and motherboard at ~67-70C with P95 load.
September 19, 2009 11:47:02 PM

noob2222 said:
don't quote worst case scenarios vs best case when you try to defend yourself.

The PII is limited in its overclocking to just over 4.0 ghz in 64bit os (probably due to the way it handles the 64 bit registers), in 32 bit os, they have been reported to reach 4.4ghz + on air cooling.

Sure, I7 can hit 4.6 with perfect ventilation and a massive heatsink while running 85+ Celcius in a temperature controlled room, Go to newegg and see what the average user can achieve without fear of frying his cpu. Most (key word is most) range from 3.5-4.0, same as the PII BE's. I am quoting newegg because the results are REAL WORLD, not lab controlled.


4.4Ghz on air with golden samples + suicidal voltages.

http://www.pcrpg.org/pics/computer/octsuicide.png

On the other hand, i7 920 D0 can hit 4.0Ghz easily with 1.2~1.3V.

So its not the worst case scenario vs. best case scenario. Its more about AMD's best case scenario vs. Intel's average case scenario.


And like Shadow said, about 99.5% of the Newegg users are idiots. I used to read Newegg reviews everyday, so I know.

Water cooling is just as efficient as new air coolers unless your adding ice to your tank, go read some articles. said:
Water cooling is just as efficient as new air coolers unless your adding ice to your tank, go read some articles.


If you don't know how to water cool, then yes, its as good as high end air coolers. But XS users are known for their inadequacy in setting up a proper water cooling setup :sarcastic:  .

Looks like the name you made is very fitting...noob.
a b à CPUs
September 20, 2009 1:04:43 AM

yomamafor1 said:
4.4Ghz on air with golden samples + suicidal voltages.

http://www.pcrpg.org/pics/computer/octsuicide.png

On the other hand, i7 920 D0 can hit 4.0Ghz easily with 1.2~1.3V.

So its not the worst case scenario vs. best case scenario. Its more about AMD's best case scenario vs. Intel's average case scenario.


And like Shadow said, about 99.5% of the Newegg users are idiots. I used to read Newegg reviews everyday, so I know.



If you don't know how to water cool, then yes, its as good as high end air coolers. But XS users are known for their inadequacy in setting up a proper water cooling setup :sarcastic:  .

Looks like the name you made is very fitting...noob.



:lol:  i have to agree. If you don't know how to set something up properly, you wont get the same benefits as some who know how to do it.

Quote:
And like Shadow said, about 99.5% of the Newegg users are idiots. I used to read Newegg reviews everyday, so I know.


Yeah. You can only trust them as far as you can throw them.

a b à CPUs
September 20, 2009 2:23:15 AM

yomamafor1 said:
4.4Ghz on air with golden samples + suicidal voltages.

http://www.pcrpg.org/pics/computer/octsuicide.png

On the other hand, i7 920 D0 can hit 4.0Ghz easily with 1.2~1.3V.

So its not the worst case scenario vs. best case scenario. Its more about AMD's best case scenario vs. Intel's average case scenario.


And like Shadow said, about 99.5% of the Newegg users are idiots. I used to read Newegg reviews everyday, so I know.



If you don't know how to water cool, then yes, its as good as high end air coolers. But XS users are known for their inadequacy in setting up a proper water cooling setup :sarcastic:  .

Looks like the name you made is very fitting...noob.

+1000. And many of those 4.4Ghz OCs are done by pro's (look at hwbot rankings) with hand picked golden CPUs. I envy those guys who have access to trays full of CPUs. :cry: 
!