Intel's 'Overspeed protection' will limit Core i7 overclocks!!!!!!

Page 3 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Yeah we are. I was trying to point out synthetic benches are worthless, and everything except gaming, i7 so far looks good, but unfortunately, very costly as well. Looks like 200 on the bclk, IF the cpu can handle it, could give some good oc's
 
if you don't have a highly clocked core 2 duo or quad... a core i7 is a huge improvement... if you have a highly clock duo or quad... there is absolutely not reason to get a core i7... unless your uber rich 😛

thats fact... for a person with "hard-core" gaming in mind... its not really all that important... for multi card setups it is... but then again... what percentage of the PC market has a high end cpu but a low end gpu (8500 gt ftw) lets just say 20 %... then lets say half of those people have a medium high end graphics card... now we're down to 10 %... high end gpus maybe a quarter of that... because mid-end cards are where the money is at... and then 2.5 % of computers have high end cards... THEN take into consideration that maybe 1 in 5 will have a large monitor (1920 by 1200 or more) so now we are down to half a percent... then at that point... multi card setups would start to be ok... but you cross fire 2 high end cards... and see not that much of a performance increase... but then when you add a core i7 you see a little bit more horsepower... well then what? it won't matter as you'll only see the core i7 REALLY pull away from the crowd in gaming at a resolution of 2560 by 1600 (which like 1 in 10 of large monitor owners have... if even that...) so then you add a THIRD card... what percentage of computers is that? a fraction of a percent... core i7 doesn't bring enough to the table for noobs 😀

fact
 
^Not really. I posted a benchmark done @ 1600x1200 in Crysis @ very high with a 9800GX2 and clock per clock vs C2Q Ci7 was showing 35% better performance. Basically at those settings the Ci7 machine was able to hit just about 40FPS, which is smooth, unlike 28FPs on the C2Q. And the 920 @ 2.66GHz was just 1.09FPS lower than the 3.2GHz QX9770.

And the thing is that multi GPU single slot cards are becoming more normal these days. And it looks like Ci7 can give those cards what they need in order to perform better.
 
Im thinking that the Q series brought out the best in the 4870x2, and sometimes the G280 as well, at higher clocks, and we were still somewhat bottlenecked by cpus using tri sli or quadfire, but the bump in IPC, and whatever else i7 is doing (communicating better?) is bringing out the multi card setups
 
thats why i said for noobs... only pros have multi card setups 😛

i for one am a noob since i can't afford to be pro
 
Ok. So all the boards that support i7 so far are the X58 boards, right? Which are for multi-card setups, enthusiasts, and overclockers. Think the override will still be available in P55 boards? I mean, Intel has the technology to severely limit overclockers, surely they put it there for a reason. And what else could it be? Thermal throttling has done plenty well for C2Q and everything before it, why add transistors for something that's already done well?

I just still feel there's something fishy afloat here.
 
Thermal throttling is designed to protect system from heat stroke. It has nothing to do with limiting OC and can and be disabled if Intel put the option into it. I see no reason that they don't do this for desktops. There's no reason to drive gamers into AMD camp once again.
 
Thermal throttling has done plenty well for C2Q and everything before it, why add transistors for something that's already done well?
Click to expand...
Conroe was pretty robust, but quite a few people have wrecked their Wolfdales. Even if you have very efficient heat removal, you can still run too much current through a processor. Current is the direct cause of electromigration - specifically, current per cross-sectional area of conductor. Overclockers talk about voltage because, in a classic circuit, voltage increases current proportionally.

The i7, by monitoring and restricting current, can stop electromigration in the power delivery lines in spite of threatening voltages. Overcurrent protection is a much more robust way of safeguarding against electromigration than thermal throttling, which can be fooled by good cooling, or plain voltage monitoring, which is not conducive to a negative-feedback loop and can hamper performance unnecessarily.



First of all, a 2.4 Venice was listed as....? A 4800, which meant? A P4 running at 4.8Ghz, not 3.2. Now, if you think Im wrong, think again
Click to expand...
A 2.4 GHz Venice is model 3800+. But that's a slightly optimistic marketing figure; if you browse some benchmarks at http://www.neoseeker.com/Articles/Hardware/Reviews/a64-x2-4200/6.html, you see that it trades blows quite well with a 3.6GHz Prescott.

Scaling the pi times:
Athlon64 San Diego @ 3180 MHz: 27.1s (86200 MHz-s)
-> Athlon64 San Diego scaled to 2400 MHz: "35.9s"
Pentium4 Prescott @ 3980 MHz: 32.8s (130500 MHz-s)
-> Pentium 4 Prescott scaled to 3600 MHz: " 36.3s"

SuperPi doesn't favor the Pentium 4, revealing it for the poor performer it was. It's nearly dead-center among the benchmarks between the Athlon64 and Prescott. It also doesn't favor the Pentium 3, or the initial Pentium M, or earlier processors. That makes it a valid benchmark for processors up to the time, in my book. This is in sharp contrast to what happens starting with the Dothan. If you try to compare SuperPi times of Dothan and successive generations with Athlon64, you'll find SuperPi a greatly optimistic indicator of general performance.

I don't think there's anything inherently Intel-biased about SuperPi coding. The Gauss-Legendre algorithm is derived from mathematical work around the 1800s and was used in 1995 to calculate Pi on a supercomputer. As far as I know, it employs a lot of integer division and some way of calculating sqrt(2) to great accuracy. SuperPi hasn't even been modified for later instruction sets or multicore. Unsurprisingly, we've known from synthetic benchmarks that Intel's mobile lineage has become extremely strong with integer arithmetic. The consistently improving efficiency metrics only underscore the continual changes they make to their core, rather than producing generic shrinks.
 


Dont waste you time on Thundy. He's our pet troll. Most everyone knows to ignore his posts except for their entertainment value. And this post is even a far cry better than his fudzilla regurgitating "i7 wont overclock".
 


http://www.youtube.com/watch?v=Xi4yxKNWehU

http://www.youtube.com/watch?v=Ivx0XYMCZJw

http://www.youtube.com/watch?v=nq_XG411Lik

If you don't see the difference, you are truly blind.

http://www.youtube.com/watch?v=_l2e0mf3CcA

From these videos you can summarize that Intel continually produces products that do not perform and cost the customer more money then their competitors.

And what about all the innovations that came from AMD that Intel copied? oops

=

Description
Heading back home after opening an R&D center in Bangalore, AMD CEO Hector Ruiz, took some time off to bash Intel.

"If you look at the last five years, if you look at what major innovations have occurred in computing technology, every single one of them came from AMD. Not a single innovation came from Intel" Ruiz claimed. "So I would say that Intel is trying to catch up with us in that respect."

But how come Intel is outselling AMD for more than a year now? According to the AMD CEO, "Intel continues... to abuse their monopoly and that's why around the world governments and regulatory agencies continue to go after them."

The European Union and the South Korea accused Intel of anti-competitive practices this year, but the American Federal Trade Commission recently said it would not charge Intel with anti-competitive practices.

http://www.megagames.com/news/html/hardware/amdweinnovate,intelcopies.shtml

As you said, "Oh Wells....."
 


Yet another common troll. Well I will respond to some things.

First video, made by AMD DUH it will be biased.

Second video. Um I am not taking some random guys word for it when I have seen a Q6600 beating it on official benchmarks. In fact here is some prrof:

http://www.tomshardware.com/charts/desktop-cpu-charts-q3-2008/Cinema-4D-Release-10,835.html

Taken right from THG. Q6600 beats the 9500 easily. It took a 9950 @ 2.6GHz (200MHz advantage) to beat the Q6600. Not exactally what that video shows.....

Third video - Uses a chip thats obviously higer price and offers vastly better performance instead of a equally priced chip (Q6600, Q9550) that offers the same or better performance. Huh. Thats funny isn't it. Biased much? I think so.

Fourth video, um what does that have to do with gaming? I mean sure some ppl will play a few games on the 780GX chipset. Its a great chipset. But why would you not get a discrete card that offers vastly better performance?

As for what the ex crap CEO of AMD said, He is obviously not watching. Lets see: 45nm, first working 32nm SRAM, 80 core CPu that has been made and tested, Hafnium, HK/MG, The first quad core, Turbo Mode that dynamically OCs the CPU without the need to do anything in BIOS, oh and my fave a core to core link system that is as fast as one that AMD will be releasing soon, HTT3.1 that is.

Of course he would never say Intel innovated anything becausehe wants to make AMD look good even though if their innovations were so great I don't think they would have 8 straight quarters in the red.

Face it. Phenoms launch was a bust. Phenom is not up to par with C2Q. Deneb should be but we have yet to see any tests. Of course the NDA doesn't lift until much after Core i7s realeas.

Oh, BTW..... whats your excuse for Core i7? Its naitive, can't use that. They have one thats at $300 dollars to start which puts a total PC system at about $1000 for a pretty good one. Not much left for ya is there?

Bleh come back when you have something non biased or at least non youtube. Can't trust most of the results due to how many idiots post videos there on a daily basis.
 
^^I am still waiting for him to get some new material. Prferably something on Nehalem. He can't use the "its not perfect design and will be hotter on one side even though Phenom uses more power but is perfect" crap anymore.

Its going to be funny really.
 


Hotter on one side? What's that supposed to mean? :??:
 
Don't ask me. He posts it constantly stating that Phenoms design is perfect and even heat. Yet Phenom uses more power than C2Q.

Either way he is a troll and it will be interesting to see what he pulls out of his sock puppet, you know the one I mean, next. Probably the whole "The benchmarks are rigged to like Intel BS".

That and the new president should provide me some entertainement for the next few months.
 
...and yet another troll-feeding thread. when will they end? never.

Well it's good that you can overvolt the i7's cos i was gonna rage to the extreme @ francois cuz i saw an interview saying they thought of overclockers needs from the beginning... saw that you couldn't disable overspeed protection thank god we now can.
 
^thing thats funny is this was a rumor posted by Fudzilla a whil ago. And then Intel came out and said you could disable it. But instead of ppl remembering that they just jumpped the gun and believed that you couldn't.
 


He knows how to just stay on the right side of the line. Although I think he should be covered by an elastic clause of "we reserve the right to do whatever the hell we want; it's our board".
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom