• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Did Intel Fake it's Dual Core Processor

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
Seems VERY unlikely to me that they did...if so, their reputation would drop tremendously...


BTW: that article sounds like 'written by AMDroid'
 
I see were you are coming from, more less I think they mislead us. It's not a true dual core, more two cores stuck together on a chip instead of integrated into one.
 
ugh...the inquirer...I really have a hard time believing anything they say. I mean, what kind of site says "Flame Author" in the email link?
 
I am using AMD, but I hope Intel didn't or they will appear to have the same creditability as CBS News.
 
dicecca112 said:
I see were you are coming from, more less I think they mislead us. It's not a true dual core, more two cores stuck together on a chip instead of integrated into one.


leads me to the question: how do you define dual core? If i made a chip the size of a mobo and put two cores on each end with a bridge linking them, is that dual core?

The guy in the article says that Intels dual core isnt dual core since hes from AMD....who has the real definition?
 
Sjaak said:
leads me to the question: how do you define dual core? If i made a chip the size of a mobo and put two cores on each end with a bridge linking them, is that dual core?

The guy in the article says that Intels dual core isnt dual core since hes from AMD....who has the real definition?

I think his point is that they just slapped them together with no regard on performance.

Well, if you just encase a second, none functioning core in with a working one, that could be a dual core, right?
 
dual core to me is a single core that integrates two cores in one internal structure, or something like two caches. Not just two cores put on a single chip. Don't get me wrong I am not an AMDdroid, in fact, I may go to Intel soon, and am increasingly leaning towards intel if they fix the prescott heat issue. But I just think it is false advertising in a sense.

EDIT: Ohh come on Texas, lets be sensical here. You understand what I mean in the post.
 
Yeah but since there is already spoken of the cores sharing a bus and L3 cache, what IS the core? Just the processing unit & registers w/o any cache? or does L1 / L2 cache count?

Seems to me that for two cores to run fine, you need double the bus, probably even 4 channels of ddr (2 per core)

2 cores sharing one bus is inefficient..look at Xeon's, their performance in memory lags behind because they share a single memory bus.
 
I would not be suprises if the first gen of Intel dual core chips are going to rather poor compared to the others because from what I have gathers is it is very rushed. They are basicly play catch up and trying to keep up with IMB (G5) and AMD which both are well on the road to daul cores.

I think intel beein playing catch up on there chips since the AMD 64 with the intgerated memory control.
 
I would like to think that they'd spend the time to re-engineer the workings to get the best performance out of it, if I'm gunna hafta deal with the heat from both cores, might as well get all the performance I can out of it
 
ugh...the inquirer...I really have a hard time believing anything they say. I mean, what kind of site says "Flame Author" in the email link?
As funny as Mike Magee is on that site, he is normally right on the money most of the time, if you have been followoing the IT industry as I have been....He is well respected...even in Wall Street. Hard to believe right....
 
First generation dual-core Intel CPUs are just 2 cores slapped together.
(Because they don't have time to re-design and be able to match AMD's projected release date of dual-core A64 FX/Opteron CPUs).

In theory, it should operate identical to today's Xeon setups where the bottleneck is a shared bandwidth between 2 CPUs. Increasing the bandwidth of the bus would help the situation. (Compared to AMD's dual-core implementation where they share the memory controller). No one is really sure which is really better until benchies (not synthetic BS, real apps), are done to confirm.

The 2nd generation dual-core Intel CPU is a complete re-design, from ground up and is far more optimised. If you're planning for Intel dual-core, I suggest you wait for 2nd generation. (Besides, it should be based on the Pentium-M by then!)

Anyway, the article in question.

Because at IDF Sept 2004 (when they demo'ed the dual-core on i915 chipset),

(1) Intel didn't let any Journos get close enough to see or check out the specs.
(2) No "plexi" casing for everyone to see the inside.
(3) Intel guarded the box closely.

The one problem with this demo is...
We aren't sure if its a CPU with SMT,
or a dual CPU
or a real dual-core.

But it doesn't mean a dual-core variant of the P4 doesn't exist. We aren't 100% certain, that's all. But who really cares when it won't be available until halfway through 2005.

The biggest misconception I've noticed about dual-core is heat.

(1) They ain't gonna use standard desktop cores with those voltages! They'll use Low Voltage variants of the cores...So instead of one core that produces 90W+, they use something like a dual-core with 40W+ per core.

At most dual-core will be slightly hotter.

Example : Opteron specs
Single-core = 89W
Dual-core = 95W

This allows existing Opteron mobos to easily work with dual-core variants of Opteron...Of course, they'll still need a BIOS update!

(2) The speeds will be lower than single core.
Instead of a single 2.4Ghz Opteron, you'll have a dual-core 2Ghz Opteron.

I'm not sure how a dual-core will behave if you overclock it, that will have to wait until end of 2005. :)
 
MMMM I wanna see a 12 pipe P4. That thing would be a beast. Just put a massive cache on it and that would solve most of your memory issues. Or you would have to go to a Next gen rdram sort of platform. That cranks huge memory bandwidth. You could run said chip at like 1ghz and it would mop the floor with any current P4. Since each pipe would run at about 75% of the effeciency of a single piped P4. Would be a monster chip requiring lots of new stuff to run it though. But it would be equivalent to an 8ghz machine, give or take a few mhz. Hmmm even at that speed though you would be more likely to run shortened pipes so with alot fewer stages you would make up for the lower effeciency. It would sort of be like a gpu but in a cpu jacket.
 
Back