Overclock.net banner

1 - 20 of 22 Posts

·
Old to Overclock.net
Joined
·
1,751 Posts
Discussion Starter #1
<a href="http://www.tomshardware.com/2006/09/10/four_cores_on_the_rampage/" target="_blank">http://www.tomshardware.com/2006/09/...n_the_rampage/</a> <br />
<font face="Comic Sans MS"><font size="4"> How many of you have read this? What do you think?</font></font>
 

·
Registered
Joined
·
2,264 Posts
F'ing amazing?
 

·
Registered
Joined
·
119 Posts
Yep, good stuff. Still LGA775 tho (i've heard), so upgrading won't be too bad..
 

·
Registered
Joined
·
2,671 Posts
Damn. If intel eliminates AMD we are going to have to double our spending on processors since there will be no competition...just like microsoft is overpricing all their *****.
 

·
Registered
Joined
·
288 Posts
If I am not mistaken Intel and AMD are both US based companies and thus by law they cannot possess a monopoly on the processor market.
 

·
Premium Member
Joined
·
8,095 Posts
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>Larethian</strong>

</div>
<div style="font-style:italic">If I am not mistaken Intel and AMD are both US based companies and thus by law they cannot possess a monopoly on the processor market.</div>

</td>
</tr>
</table>
</div><br />
Ya, but putting out massively better stuff doesn't mean that AMD can't go under...lol.<br />
<br />
If it just happens...then Intel can't be held responsible...
 

·
Overclockiit
Joined
·
3,413 Posts
i read about the Quad Core chips intel plans to release soon and ive seen some unofficial benchmarks relating to it and its pretty amazing the 3DMark06 CPU score i beleive was over 5000 alone so its gonna be a good seller im not sure what AMD plans to do with there 4X4 but they better do something pretty soon or im afraid AMD may not have any market shares left to sell, also intel plans to retail the Quadro for around $1kusd not bad at all.
 

·
Registered
Joined
·
4,046 Posts
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>|2A|N</strong>

</div>
<div style="font-style:italic">i read about the Quad Core chips intel plans to release soon and ive seen some unofficial benchmarks relating to it and its pretty amazing the 3DMark06 CPU score i beleive was over 5000 alone so its gonna be a good seller im not sure what AMD plans to do with there 4X4 but they better do something pretty soon or im afraid AMD may not have any market shares left to sell, also intel plans to retail the Quadro for around $1kusd not bad at all.</div>

</td>
</tr>
</table>
</div><br />
I am fairly certain that new reports say there will be two versions of quadro. The first is the K6700 and the other is the 6600. One should be at the 999 price point and the other at the 510 price point. Still more expensive than a single 6600, but still, not bad at all for FOUR CORES and only 2XX megahurtz less speed. rock on, thats what I am getting.... <br />
<br />
<a href="http://www.xbitlabs.com/news/cpu/display/20060919051744.html" target="_blank">http://www.xbitlabs.com/news/cpu/dis...919051744.html</a>
 

·
Banned
Joined
·
2,177 Posts
damn 510 is a lot still but for 4 cores thats an awesome deal, once I get a job I'm going to be getting one of theses unless 4x4 turns out to be awesome (like 16 operations per cycle <img src="/images/smilies/biggrin.gif" border="0" alt="" title="Big Grin" class="inlineimg" />)
 

·
Registered
Joined
·
4,046 Posts
Ah, right NOW. But in the near future my dear friend, in the near future we will be TRIPPING over the multiple threads they put in games. UP TO OUR EARS in threads that developers lay at our feet. And you need to ask yourself. Do I want to be completely and utterly LACKING in FPS when that happens?<br><br>
I am terribly sorry, its late....
 

·
Administrator
Joined
·
6,696 Posts
Personally I do not like what I am seeing with Core 2 Quadro.<br />
There are several problems as far as I am concerned with this particular quad core processor.<br />
<br />
<b>1. The bus speed is not great enough for full data transfer to and from a quad core processor. </b><br />
<br />
A 1066MHz/1333MHz bus is not great enough to allow great sums of data to be sent from/to each processor.<br />
<br />
Core 2 Duo = 1066MHz, 533MHz each<br />
Core 2 Quadro = 1066MHz/1333MHz, 266MHz/333MHz each<br />
<br />
In the real world this would never happen as there is no limitation in the bus pipeline. In retrospect means that one core can use 100% of the bus speed while the others use nothing, again it could be a 50/50 split. There is no value that each processor is limited per bus.<br />
<br />
<b>2. Ability to process SMP SSE</b><br />
<br />
A proportion of the Complex Intel SSE extensions are not able to run in a SMP environment as they are limited to single thread execution. Therefore a proportion of instructions would be limited to 1/4 processors (25% of overall processing power + memory store). In this scenario there would be no speedup at all if these complex instructions were to be used.<br />
<br />
<b>3. Thermal Displacement Ability</b><br />
<br />
The ability of this chip, with four cores, is not able to displace heat as well as I would have hoped, therefore overclocking will be limited from a thermal point of view.<br />
<br />
<b>4. Ability of Developers To Create a SMP Game For a Four Core System</b><br />
<br />
Creating a game that uses advanced SMP techniques to support multiple cores is very complex to write. The ability of most game developers at the current coding speedup will not be great enough to create many games that can use a four core system before eight cores are used.<br />
This is a problem for all quad core processors <img src="/images/smilies/smile.gif" border="0" alt="" title="Smile" class="inlineimg" />
 

·
Registered
Joined
·
4,046 Posts
The manual -- To some extent I agree. I also think that the Bus issue is something that may definitely have an effect on overall performance of the chip. I also fear for heat issues, as this is essentially two dual cores sandwiched together. However, if the benches at XS have proven anything, overclocking thresholds on Kentfield are still in the stratosphere when compared to prescotts and high end pentium Ds (were not talking 805, here).<br><br>
As far as programming for multiple cores, I dont think we are far off from being able to program for X cores. Meaning, that hopefully developers will unlock a process by which they can program applications to utilize the number of cores available, not just hard program applications for two cores, for four cores, etc. I see this as the future of MP programming. and while I have nothing to back up whether this is possible, I didn't think multicore processing was much of a possibility 2-3 years ago, so I dont want to put developers "in the box".
 

·
Administrator
Joined
·
6,696 Posts
<div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">As far as programming for multiple cores, I don’t think wee are far off from being able to program for X cores.</td>
</tr></table></div>
I disagree with that statement <img alt="" class="inlineimg" src="/images/smilies/smile.gif" style="border:0px solid;" title="Smile"><br>
The ability for us to program an application to allow processing of a multiplier core system is very complex.<br>
The SSE instruction set can support two cores to an extent, however many of the instructions are not built to handle a four way split. Due to the extreme complexity of this issue the SSE instruction set will take a long time to port to a four core instruction processing system.<br>
Therefore as long as this instruction set can not be ported correctly the ability to create an application which needs SSE (like most games) is not an easy possibility. This is the same with MMX to an extent, which again, is another important instruction set for gaming and 3D applications.<br><br>
Creating a full four core supporting application will take very many years. Don't forget we only have a small handful of applications that can fully use a two core system. Most can only partially do this i.e. 3DMark, PCMark, Quake 4 etc.<br><br><div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">However, if the benches at XS have proven anything, overclocking thresholds on Kentfield are still in the stratosphere when compared to Prescott’s and high end Pentium Ds (were not talking 805, here).</td>
</tr></table></div>
That is logical considering that the TDP of this chip is within level of the Pentium D series of processors, therefore overclocking should technically be within a certain range of those processors on the Netburst architecture. However following my statement there will be a limitation due to the thermal equilibrium status of the processor, just like there was with the Pentium D. Which is exactly what my statement implied <img alt="" class="inlineimg" src="/images/smilies/smile.gif" style="border:0px solid;" title="Smile">
 

·
I remember when......
Joined
·
3,812 Posts
I agree that the bus speeds taken under the most extreme situation could cause some bottlenecks, however, it is unlikely that there will be a regular occerance of times where all four cores need the full bus bandwith. Many times one core can be happily working out of it's cache or on die and then shoot information through the bus while another core just finished using the bus and is working on die, ect, ect....of course, this is happening rather fast, and could be likened to a WW1 plane firing it's machine gun through the propeller (before the invention of the interuptor gear).....certain there will be some impacts, but not constantly.....and the performance drops from bandwith bottleneck would probably not be noticed by most simply due to the large jump in performance over what we have now.<br><br><div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">It certainly makes sense to increase the front side bus to FSB1333, since each Core 2 "Conroe" processor has to use this bus to communicate with the other one and fetch or write L2 cache data, especially when many applications run simultaneously (aka multitasking). Speeds such as those pose no problem for 90-nm chipsets.</td>
</tr></table></div>
<br>
Will Kentsfields be all that they can be? No.<br><br>
Will they be faster than anything out now? Yes.<br><br><br>
Also, while it is true not a lot of applications use multi-core, there are some ([email protected], BeyondTV, ect....), and these chips will be VERY nice for people that use them. Too often the benifit of new hardware is compared to it's effect on gaming and single application speed....and while I understand there are a great many people who find that most important, it is not the end all. I find much more usefullness in the ability to do more things at once......<br><br>
I for one am looking forward to having 4 cores......I can fold for the cure(2 cores), while recording live TV and watching video stored on my hard drive......or I could play a game, fold, record live TV, ect, ect......
 

·
Registered
Joined
·
1,890 Posts
I am not up to speed on the conroe make up but I wanted to give my 2 cents anyway. I see people talking about AMD going under due to market share. AMD will not be going anywhere. They are doing just fine. They may be behind for now but they will be back the CPU war will continue and hopefully that will benefit us all!! <img alt="" class="inlineimg" src="/images/smilies/biggrin.gif" style="border:0px solid;" title="Big Grin">
 

·
Administrator
Joined
·
6,696 Posts
*sigh*<br>
First error:<br><div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">Many times one core can be happily working out of it's cache or on die and then shoot information through the bus while another core just finished using the bus and is working on die</td>
</tr></table></div>
How does the processor transmit data to the onboard memory sub-system? The answer is the bus system also communicates with onboard devices contained within the processor.<br>
In the real world the FrontSideBus does not exist, it has not existed since about 1990 or so in Intel processors. FrontSide was the name given when the system operated with two bus systems, a FrontSide bus and a BackSide Bus <img alt="" class="inlineimg" src="/images/smilies/smile.gif" style="border:0px solid;" title="Smile"><br>
This is logically (well I hope) why you do not know that the bus system communicates within the processor.<br>
I take it you have heard or interrupts and buffers? (about age 16 or so you should have been taught about those basic elements). With a multicore system the interrupts can not all be assigned to different locations, therefore at least two cores will want to access the bus system at any one time, regardless.<br><br>
Error 2:<br><div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">I agree that the bus speeds taken under the most extreme situation could cause some bottlenecks, however, it is unlikely that there will be a regular occerance of times where all four cores need the full bus bandwith.</td>
</tr></table></div>
Taking the points above into account you may now realise that the full bandwidth of the system will be used far more than originally expected as data will not just be transferred out of the processor but also to the internal memory sub-system.<br><br>
Error 3:<br><div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">Will they be faster than anything out now? Yes.</td>
</tr></table></div>
You can not justify that statement at this current moment in time.<br>
With the current benchmarks released you will see that in many games the performance is less as they can not use SMP to the full extent <img alt="" class="inlineimg" src="/images/smilies/smile.gif" style="border:0px solid;" title="Smile"><br>
An accurate answer can not be justified until the release of substantial evidence prior to the release of the processor.<br><br>
Error 4:<br><div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">Also, while it is true not a lot of applications use multi-core, there are some ([email protected], BeyondTV, ect....), and these chips will be VERY nice for people that use them.</td>
</tr></table></div>
[email protected] is unable to fully use resources on SMP systems with a greater CPU count than 2 due to limitation in SSE/MMX instructions.<br><br>
Couple more errors, but I think they are the most important ones <img alt="" class="inlineimg" src="/images/smilies/tongue.gif" style="border:0px solid;" title="Stick Out Tongue"><br>
Hope that helps, have a nice day <img alt="" class="inlineimg" src="/images/smilies/wink.gif" style="border:0px solid;" title="Wink">
 

·
Registered
Joined
·
303 Posts
my question is that even though Intel seems to be whooping up on AMD, I don't understand why the Intel Corporation stock is hardly increasing. Anyone have an answer, I don't know much about stocks so maybe you can help inform me?
 

·
I remember when......
Joined
·
3,812 Posts
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>The_Manual</strong>

</div>
<div style="font-style:italic"> *sigh*<br />
First error: <br />
<br />
I take it you have heard or interrupts and buffers? (about age 16 or so you should have been taught about those basic elements). With a multicore system the interrupts can not all be assigned to different locations, therefore at least two cores will want to access the bus system at any one time, regardless.</div>

</td>
</tr>
</table>
</div>Didn't really need the condescending sigh and comments there <img src="/images/smilies/wink.gif" border="0" alt="" title="Wink" class="inlineimg" /><br />
<br />
At least 2....but not always 4........which, as I said, will result in some bottlenecks, but not a constant bottleneck. Perhaps with a bigger, and redesigned FSB system the CPU could give 80% bumps in performance across the board.......but that doesn't mean I will ignore the fact that it may give an average of 30-40% increase over the current Core 2 Duo with it's FSB as is. A performance increase is an increase none-the-less, even if there is room for more.<br />
<br />
<br />
<br />
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>The_Manual</strong>

</div>
<div style="font-style:italic"> Error 3:<br />
<br />
You can not justify that statement at this current moment in time.<br />
With the current benchmarks released you will see that in many games the performance is less as they can not use SMP to the full extent <img src="/images/smilies/smile.gif" border="0" alt="" title="Smile" class="inlineimg" /></div>

</td>
</tr>
</table>
</div>Well, looks like we are both full of errors then, because the fact that <u>games</u> do not show greater performance doesn't show anything other than games may or may not benefit.......which is a Staw Man, since my original statement specifically stated gaming performance was not everything, and specifically not what I was looking at......if we are to cherry pick benchmarks to prove a point.....<br />
<br />
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

The future belongs to HD content. If we take our benchmarks into consideration you can no longer get by without a quad-core processor. Test results with the software packages Main Concept with H.264 encoding and the WMV-HD conversion make this very clear. <b>We noticed performance jumps of up to 80% when compared to the <u>Core 2 Duo at the same clock speed</u> </b>(2.66 GHz). A Core 2 Quadro at 2.66 GHz and higher is the answer for HD video (editing and rendering) at full HD resolution (1920x1080).

</td>
</tr>
</table>
</div><div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>The_Manual</strong>

</div>
<div style="font-style:italic">An accurate answer can not be justified until the release of substantial evidence prior to the release of the processor.</div>

</td>
</tr>
</table>
</div>Which means your answer is just as erronious as you state mine is.......<img src="/images/smilies/smile.gif" border="0" alt="" title="Smile" class="inlineimg" /> <br />
<br />
<br />
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>The_Manual</strong>

</div>
<div style="font-style:italic">Error 4:<br />
<br />
[email protected] is unable to fully use resources on SMP systems with a greater CPU count than 2 due to limitation in SSE/MMX instructions.</div>

</td>
</tr>
</table>
</div>Since I am so ignorant, maybe you can clarify this a bit <img src="/images/smilies/tongue.gif" border="0" alt="" title="Stick Out Tongue" class="inlineimg" /> Folding may not be able to use 100% of all 4 cores, but it should be able to use 2 of the cores for Folding, and the other 2 cores are open for something else...as I stated in my example.<br />
<br />
<br />
<br />
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>The_Manual</strong>

</div>
<div style="font-style:italic">Couple more errors, but I think they are the most important ones <img src="/images/smilies/tongue.gif" border="0" alt="" title="Stick Out Tongue" class="inlineimg" /> <br />
Hope that <u><b>helps</b></u>, have a nice day <img src="/images/smilies/wink.gif" border="0" alt="" title="Wink" class="inlineimg" /></div>

</td>
</tr>
</table>
</div>Not really, it's hard to be helped when being talked down to <img src="/images/smilies/wink.gif" border="0" alt="" title="Wink" class="inlineimg" /> .....and when the points given fall far from the mark of my original post. <img src="/images/smilies/thumb.gif" border="0" alt="" title="Thumb" class="inlineimg" />
 
1 - 20 of 22 Posts
Top