Core Duo the new Prescott?

I was reading yet another thread on here last night created by a Core Duo user who was concerned about operating temperatures. It looks like there's yet another one this morning.

Has the power of the Core Duo blinded people to the fact that it's a bit of a toaster? Does dual core cause more heat that standard notebook cooling can't handle? Does the X2 do the same?

I certainly can't recall any other cpu consistently warranting such worried posting.

 

Hard to say.

I suspect the problem is Intel's overly optimistic way of measuring TDP.
They report TDP's that are a good 15% lower than the maximum amount of heat the cpu can actually generate. So if heatsink manufacturers take them on their word, you'll get overheating systems, especially if the CPU often runs close to peak. (The P4 used vastly more power under load than when idling. Core Duo may come closer to peak even when idling, so it'll exceed the TDP more often)

I also expect this is why the TDP on Core 2 is reported to be higher. Not because it produces more heat (from what I've seen, it doesn't), but because of all the reports of Core Duo machines overheating. (The Macbooks come to mind as prime examples).

The Core Duo doesn't pump out all that much heat (which makes it unlike the Prescott). But it doesn't have to, in order to reach high temperatures. It just has to pump out more than the notebook manufacturer expected for temperatures to build up.

I haven't heard of problems with X2's running hot, but that could just be because AMD reports a higher (actually valid) TDP to begin with, so manufacturers know what they're dealing with, and can fit a sensible cooling system.

That's just my guess though. It's the only way I can make sense of the facts that a lot of Core Duo systems do run very hot, at the same time as benchmarks shows it to use very little power, and produce not that much heat.

 

Jalfs's explanation is very accurate. The TDP of Core Duo and Core 2 Duo is what the notebook manufacturers design their cooling on. So, that figure needs to be accurate to a cool running processor.

Adding more cores to a processer certainly does increase heat output, if it isn't done properly. The new Core Quad that has been released has twice the TDP of the regular Core Duo. This is because Intel has simply stuck two Conroes onto a processer package. In the future, however, Intel will make an effort to intergrate four cores into one core, and not two seperate ones. This will help reduce TDP. I know that these don't refer to mobile processers, but I think they give an idication of how things are like.

 

Very good explanation Jalf, although I find it distressing that notebook manufacturers like Apple don't bother to actually test the actual amount of heat generated before mass producing them and shipping them off to retailers.

 

just for the record, it wasn't an explanation. It was a guess...

 

Very good guess Jalf, although I find it distressing that notebook manufacturers like Apple don't bother to actually test the actual amount of heat generated before mass producing them and shipping them off to retailers.

 

I have an X2 and this thing can run cooler than my old celeron M, at about 51-53 celcius right now with media player and fire fox running

when gameing the computer temps go up to a peak of 77 celcius after a couple hours of graphics intensive gameing.

when i encode video(ripping and converting to WMV) with magic dvd ripper i get 70 celcius

the computer does not get hot but the air coming out the side of the computer is very hot when gaming, quite warm in encoding but bearable, and word proccessing like now warm kinda nice if it is a cool day.

 

If my laptop got anywhere remotely near 70 degrees C I'd be off for a cooling pad faster than light, let alone 80. Honestly I don't think I'd ever buy a laptop which had had any reviews citing heat problems.

 

Well, 80 degrees isn't a heat problem. Cpu's usually run fine at that temp. But if they get much higher, you'll start getting problems...

 

You kidding? NHC is usually set for a warning at 75 degrees and shut down at 95 unless I'm mistaken. I know mobile cpus can operate that high but honestly I'd be very wary of it. I'd much prefer temps that dont get above 70 for a laptop and not above 60 for a desktop.

 

NHC is very conservative. I got a CPU temp warning the other day, apparently I managed to break 80C while gaming, but this is within the normal range for Core 2 Duo's apparently so I just adjusted it to be slightly higher.

 

Well, the thing is NHC is set to play it safe.
As you said, it gives a warning at 75 (which makes sense, because as you say, most people prefer notebooks to stay under 70), but it doesn't shut down until 95. Which is because there's no reason to shut down before that. If you get above 75, you might want to pay attention to the temperatures, and to your system's stability, but it'll probably still work just fine. It's only when you reach 90 or so that you usually start getting problems with stability.

 

80c standard - no way - do you have a link?

All my AMD x2 (this is desktop wise) run around 40c, I would never want my machine running above 60c under load - time for spring cleaning and adding some fans!

Jalf - I would agree that running temp I'm sure can be ABOVE 70 - 75 as warning - but lets look at how long that laptop (or even desktop for that matter) lives running at those temps

Overall I beleive cooling is a very effective way to increase stability of all components of the PC (not only the CPU) and increase the life.

 

Uh as you may have noticed, your desktop is much much larger than my s96j. And heat tends to build up in a smaller space a lot faster than in a larger space. Also, the dual core Turions may run cooler as well, I never checked their op temps because I did not give them as much attention when getting my laptop.

 

still dont see a link for 80c standard. . . wasn't talking about turions, and as I said in () THIS IS DESKTOP WISE -

 

I wasn't aware that you were referring to another post other than my own. I thought you were replying to my post about my laptop reaching 80C, which is within the standard operating temps for Core Duo's and Core 2 Duo's in laptops. You do bring up a good point about heat shortening the lifespan of PC's overall however.

 

So 80 degrees C is considered normal load temps for Core Duos?

Think my next machine might be a Turion!


Talking desktops for one second I would consider hitting 50 degrees C to be hot for a desktop. With good cooling you shpuld be able to get idles in the 30s on a desktop machine with little trouble.

It seems my Celeron runs a bit hotter than I thought, although that could be because it's sitting on a chair. When it's on the desk it's low 50s idle and low 60s under load and I wouldn't want to go any higher than that!

 

Gator - all good, I was just wondering core 2 duo standard temps for adding to my knowledge database

Jess - I would agree that 50 C is warm, not "hot" but for a celeron or p4 I would see 45 - 55 C as standard, the older intels all run warmer stock -

My standard PC's with AMD x2 4200+ run around 38 - 40C standard (with stock fan) with 120mm fan in front of case, and 80mm fan in back - The Athlon XP's and Athlon64's prior to dual core ran stanard around 30 - 35C with 1 80mm rear case fan -

Anything above 60C unless it is a laptop I would say time for cleaning or heatsink replacement / Case cooling issues or adding -

 

80 C for a Core Duo is a safe temperature. Yes, it may be a little high for most peoples liking, but it will not physically damage the processor. Most notebooks, however, don't run that hot. The moority of them sit idling at about 40 to 50 C, and 55 to 60 on load. A few of the Sony models easily get their temperatures up to the 80 C level. For a desktop system, 60 C is quite high, as jpagel has pointed out.

 

What does TDP stand for, and why the heck would intel underreport it ???

 

My guess would be this:

Thermal Dynamic Power or Thermal Design Power

and my guess would be overporting would use to much power or produce to much heat -

 

Intel's TDP rating scheme is obnoxious for its lack of accuracy. I'm not trying to sound like a fanboy, but AMD rates their chips at their maximum possible power draw and heat dissipations. Intel isn't as aggressive in determining their TDPs. Luckily both AMD and Intel have been pushing energy requirements down over the past few years (after the great Prescott run, that is). Now if only graphics cards would start consuming less power AND offer more performance (I'm pointing at the G80s and 1950XTXs).

Anyone else looking forward to 35W 65nm parts from AMD in the next few months? I am. Media center PC, ftw.

 

I am looking forward to AMD's x4 TRUE Quad-core CPU (not a crappy core duo glued with another core 2 with EMULATED 64 bit processing) with Vista Ultimate w/ MCE built-in -
Cant wait for Intel to be shutup again just like when they said you couldn't make a 64bit CPU that was backwards compatible, then AMD did that - and then they can stop their "fastest cpu ever" commercial.
Until Intel can stop cutting corners and put REAL time into their CPU's will I start to respect their CPU's / company. I think intel is ridiculously over rated.
oh boy what have I started now

 

Well, this is off one of my older posts, but I think it is relevant information that I should dump here for all you guys to look at and pounder about.

( Original post here)


Regarding the temperature, I've posted up a response to a topic/thread right here, but I'll cross-post it here again (for the people who aren't aware of that other topic/thread):


According to Intel's specifications for the Core 2 Duo T7200 (sSpec: SL9SL), the thermal specifications is set at 100°C/212°F.

In fact, Intel lists 100°C/212°F as the thermal specification for their entire Core Duo and Core 2 Duo line of processors.

Intel also lists the catastrophic processor temperature for their Core 2 Duo range at 125°C/257°F (Section 3.4 - Catastrophic Thermal Protection of the Core 2 Duo Mobile Processor datasheet avaliable from Intel (.pdf file format))

Thermal specification is defined as:
"The thermal specification shown is the maximum case temperature at the maximum Thermal Design Power (TDP) value for that processor... For processors without integrated heat spreaders such as mobile processors, the thermal specification is referred to as the junction temperature (Tj)."

Catastrophic processor temperature is when the temperature has reached a level where potential irreversible damage might occur to the processor's silicon die, and that the THERMTRIP# signal should be activated to cut off Vcc supply "within 500ms to prevent silicon damage due to thermal runaway of the processor".

Simply put: as long as the processor stays below the thermal specificaion value, the built-in protection logic won't kick in and you won't get throttled back in speed. And as long as the catastrophic processor temperature is not reached there won't be any damage to the silicon die.

Also, almost all semi-conductor devices these days (i.e. CPUs) have an intended design life of at least 10 years, therefore, as long as it is below the quoted thermal specification value (100°C/212°F in the case of the Core Duo/Core 2 Duo), it will last that long.

Remember, most mobile processors these days are designed to run/handle higher operating temperatures, to rely on passive cooling for as long as possible without the fan kicking in to make for a quieter running system (which is desired in a notebook computer). People sometimes starts to panic when their mobile processor hits 80°C+/176°F+ running temperature, but to be honest, there is nothing to worry about. Plus Intel has for years implemented an excellent self-protection circuit which will immediately throttle the clockspeed to manage thermal dissipation levels before any long-term damage is dealt to the CPU core.

I'll only start to worry when the CPU is shooting up to past 85°C/185°F for prolonged periods, which usually means the heatsink or fan isn't doing its job.


F.Y.I., on my M1210 with the T7200, I get idle temperatures of around 45°C-55°C (113°F-131°F), and the fan only kicks in (at low speed) when it reaches 65°C (149°F). Under constant heavy loading it'll go as high as 80°C (176°F) on passive cooling before the fan kicks in and brings it back down to around 70°C-75°C (158°F-167°F).


In my opinion, if Intel have rated their mobile processors to have a useful thermal working enevlop upto 100°C/212°F for its intended design lifespan, do I want to manually force the fan on high just to have it stay, say, below 50°C/122°F but have to endure the fan noise/whine all the time? I'd say its not cause for concern and I'd rather have a quieter working system and let the BIOS handle when to regulate the temperature.

 

Judging by how hot they often seem to run I think Core Duos are a bit overrated too!

Liking the look of the Aurora m9700 as my next machine, I just hope they keep them Turion based!


Edit: in response to the post above, the cpu may be fine at stupid temperatures but that won't stop it causing damage to every other major component in your laptop when it heats up like that. I don't care what intel say, cpus should not run that hot!

 

My T7200 hits around 80c on some games (for some strange reason, HoMMV is one of these, on the other hand, NWN2 and such run on significantly lower temperatures), which is when the NHC warning triggers off (it might be default higher level warning for CD -processors?). On the other hand, the laptop doesn't get particularily unstable or choke/slow down, and the heat gets spread very well around (it's a little tense around the exhaust hole, and it sprouts out quite warm fumes), so it doesn't get uncomfortable or impossible to use which was the case on a P4-build of my friends'.

While I don't experience performance drops, choking, or any other problems, and the laptop is nice and far from too hot to use, I'm good to go.

The AMD fanboy throws the rock in the glasshouse, eh? Most people would agree that current Core2Duos are better than Turion X2s, what comes in future is a whole other story. It's an endless race I suppose.

 

More powerful yes, but not necessarily better. I'd love to see data on how cpu operating temperatures affect performance and degredation of other hardware. I can imagine IGPs and hard drives suffering from excessive heat produced by a baby belling cpu.

 

As far as the Notebook CPU games goes, I think the AMD/Intel platforms pretty much are equal contenders from what I have seen / worked on - but I have seen neither in x64 action and would assume the turion would outperform in that aspect, thats a different ball game -
I am talking Quad-core, which mostly would tend to Desktop computing for the current moment - Also my comment points to intel as a company OVERALL, as in when they jumped to saying 64 bit computing was not available until MS wrote a x64 platform - which they were wrong, I already have lost respect for their quote unquote "quad" core CPU, which like I stated above is only 2 dual cores glued together - I stick by, if you going to do something, do it right - otherwise in the end it is going to crap out - Also like mentioned above, they can't even produce a TRUE 64bit cpu - lots of glass just broke everywhere - I used to be an Intel fan, back before the Athlon XP came out - but once they started doing deals with Dell and Best Buy that they would sell DIRT cheap to them if they promote / mostly sell THEIR products and I learned about all of the above as it was happening just more and more shined me away from their company -
Sorry all you Intel Die hards out there

 

Great point, almost forgot that !

 

On the desktop side, Intel has finally got their act together. They finally have an architecture that can actually compete with AMD. For the past 4 or 5 years, all they have had is the NetBurst architecture, and we all know how unsuccessful that was. Now, most people recommend the Core 2 DUo range over the AMD models, for good reason. They are cool running, fast processers. However, I agree with jpagel when he says that the Quad-core isn't real. All they have done is join two Conroes together, and the result certainly isn't something that is efficient. The TDP has risen to what is used to be in those old days on Pentium 4. Having said that, they have produced the first 4 cored processer.

The K8L architecture from AMD has yet to be released, and we'll have to see what that brings to the table. It should be built on the 65 nm process, so it will run quite cool, and consume less power. It is much more of a fair comparison with Conroe. Its performance should also be quite good, although we don't know how it will compare with Conroe.

On the Mobile market, Intel has been doing quite well for the past few years. They have been able to compete consisteny with AMD, and it is in no small amount due to the fact that they have used some very efficient architecture. They should continue to do well, if they can keep the power consumption low and the performance at a high level.

 

jpagel:

While there weren't many benchmarks or comparisons done under 64-bit Windows between AMD's K8 and Intel's Core micro-architecture processors, there have been sites which have taken on this endevour, for example:

X-bit labs - Conroe and EM64T: Is There a Problem?

Even though it is between the Athlon 64 FX-62 and Core 2 Extreme X6800, it is extremely relevent as their mobile counterparts, the Turion 64 X2 and Core 2 Duo (Merom) are exactly the same (just with slower speed, smaller L2 cache and slower FSB speeds).

You can read the detailed benchmarks and explanations yourself, but let me quote from their findings in the conclusions:
"As we have expected, nothing serious has happened. CPUs with Intel Core microarchitecture and EM64T technology work normally in 64-bit modes. No dramatic performance drop has been detected in most benchmarks.

Yes, even though AMD "won" in 1 benchmark under 64-bit and even though the K8's x86-64bit implementation is better (well, it was afterall AMD who developed and pushed x86-64), Intel's EM64T is by no means "broken" or not working.
"AMD K8 turns out 6% mode efficient in 64-bit mode than Intel Core."
"However, this difference cannot compensate for the 20% performance advantage of the Intel Core 2 Duo over the Athlon 64 X2 working at the same clock speed, which we have pointed out in our previous articles."
"Therefore, we will not change our conclusions about the performance of the new Intel processors even keeping in mind the upcoming launch of 64-bit Windows Vista OS family."


Also, with regards to the "hacked" Quad core that is Intel's Clovertown and Kentsfield. Let me quote you what my Engineering Head of Department said to all of us during his annual speech at orientation to first year students:

"There will always be many ways upon which a problem can be solved, however, there will never be the right or perfect solution. As a professional engineer, you will learn to solve problems with solutions that works best given your current constraints. Searching for the perfect or right solution will not only be costly in terms of money and time, but it might cost you your job in the future due to missed deadlines and ripping effects up and down the entire project tree."

In my many years while doing my design courses, I found that statement to be absolutely true.

While Intel's current Quad core is not perfect, or as many AMD enthuists put it, "a complete hack", it works, and is a perfectly viable solution, and it is something that is out on the market NOW. And its performance is proven in many tests and benchmarks (such as the Anandtech benchmarks here comparing with the QuadFX, or between the Xeon (Clovertown) Vs Opteron platforms).

Does it matter to me if its a "true" monolithic die quad core or a multi-die package quad core? As an engineer, if it works like a quad core and gives me the necessary processing power boost, I wouldn't care less one way or the other what's under that heat spreader on the CPU package.

Obviously, Intel is not going to be stagnant either, after AMD launches their K8L products, Intel will move to a "true" monolithic die quad core with their 45nm process. Of course, by then I'm sure we'll get the same argument once Intel puts 2 quad-core 45nm die together to produce an "Octo-core".

Bottom line is, competition is always good, and if not for AMD coming up with the K8, Intel will have no incentive to really push R&D into something better for a resonable price. I too, look forward to the K8L not only because of what AMD might bring to the table, but also see what Intel will come up with to counter that.


Oh, and as for business ethics, well, each business is there to make money for their shareholders, and once you are in a dominant solution you'll do whatever you need to do to stay there. Every business, company, corporation is guilty of that. It'll be extremely naive to think that if the situation is reversed (e.g. AMD in Intel's position in this example), they will behave otherwise.

 

All of your statements are absolutely true and correct besides the fact that you cannot assume what another company "would have" done - Technically speaking from your engineering standpoint - absolutely agree - I speak from a end-user / reseller standpoint - I am the one selling the product you or intel or AMD designs, and to rush a quad core CPU onto the market to "beat the competition" does not impress me, nor make me want to sell their product, especially since it is more expesive to me since I am not "dell". To take the time to do it properly, is impressive (in my book). None of know what will happen till release in June, whats the point of rushing it out? Marketing scheme, making the end-user think that your company is better, lieing. Of course every company need advertising / marketing - Intel in my eyes are schemeing the public into beleive they are the best, as does every company try to do, some do it with honesty, some do it by cutting corners - in the end the best will prevail, thats why AMD has come as far as they have and take as much of the market as they have. Competition of course is 100% healthy, and sure maybe a 6% loss of EMT64 is not much since the CPU power can counter that, but what happens when the true quad core 64 bit CPU comes out? that 6% can go up with more cores and come right BACK into the picture. I am agreeing to disagree with you - You are very intelligent, you can tell by your post, your research and your posts - I care less what another website says, I care about what I see in performance, in the tests that I run, how much it costs me and my customers. Look forward to your reply.

 

and of course, as you know, on top of that Quad core can't even be utilized right now - is that advertised? of course not - I would expect them to, it is the point of the matter, technically from where I stand.

 

"On the Mobile market, Intel has been doing quite well for the past few years. They have been able to compete consisteny with AMD, and it is in no small amount due to the fact that they have used some very efficient architecture. They should continue to do well, if they can keep the power consumption low and the performance at a high level."

what the heck?? You make it sound like AMD's leading and Intel is putting up a valiant fight. Try the other way around...up till the Turion the consensus is generally that AMD didn't even HAVE a legitimate mobile part, and the CD/C2D spanks the AMD in both power consumption and performance. And if it has lower power consumption then it follows that it has less energy that needs to be dissipated. Temperatures in notebooks depend heavily on the cooling systems on the notebooks, and on an equal playing field a CD/C2D would offer better performance at lower temperatures.

And insofar as quad cores go, does it really matter? What kind of consumer needs that much power? And while Intel's design isn't "legitimate," it still works, and once again it spanks the AMD performance-and heat-wise in all but one of the benchmarks that the folks at Tom's Hardware ran.

I'm hardly a fanboy for either company, and "quality of engineering" aside, a plain utilitarian comparison of AMD and Intel mobile processors right now puts Intel ahead in the wide majority of cases.

Of course, that said, most people would probably be perfectly satisfied with either family of processor, even though there are important differences between them.

Edit: I should add that my definition of "spank" is a double-digit differential, and in many of the cases I'm describing it's a 30%+ differential

 

Prior to Turion the Athlon64 mobile was as well impressive - just power hungry -

At the current moment that power is not needed, but with vista being released and the capabilities of the OS and applications in the near future will require a lot processing power for optimal performance / gaming as it always does when newer things are released -

Intel of course right now is in the lead (no doubt) the xbit link I must add is quite interesting, some tests AMD wins, others core 2 win, others the xe965 wins, it just depends on the test - there isn't anything much of anything new out from AMD except the AM2 which doesn't even add much performance value, just lower power / heat - obviously it is similar to nvidia and ati, one come out that spanks the other, the other comes out a few months later that tops it just a little bit or in a different way -
E.G. Nvidia is geared for gaming, but they cut corners, skipping a every 7th pixel and blending it with the others, faster performance, ATI doesn't, better picture quality, lower performance but higher specs to push everything - it all comes down to this - Price and what you look to do with your PC -

The last thing I will say about all this is I look forward to the future - It will definitely be exciting -

 

Can go? Is this a fact or merely speculating? What if the 20% also goes up when the amount of cores goes up?

 

Merely Specuating -

 

Just been another thread popped up about the uselessness of dual core! I hope Alienware don't change the Aurora m9700 too much before I start looking for a new machine!

 

wearetheborg: TDP = Thermal Design Power. That's the amount of heat (in watts) that the cooling system must be able to dissipate in order to work with the CPU. It's used to give heatsink or notebook manufacturers a clear idea of how much effort they have to put into cooling a specific chip. The problem is that AMD and Intel reports them differently. According to AMD, it's the *maximum* power dissipation under full load. If you ask Intel, it's "the highest amount of heat that will be produced during typical usage".
And often, Intel's measure works ok, because, as they say, it's typical usage, which happens to be, well, typical. But if you stress your CPU, you'll end up producing more heat than the TDP allows for. So if your notebook has a heatsink that can only just handle the official TDP, and the CPU then produces 15% more heat, temperatures will start to build up.

jpagel: Intel certainly spends a lot of time and resources on their CPU's. Don't make the mistake of thinking they don't care about quality.
But when designing a CPU, you have to predict several years into the future. NetBurst looked like a good idea by the time it was designed (Or at least, it didn't look like a completely ridiculous idea). Then, three years later, when it is launched, it doesn't do as well as hoped. And 6+ years later, they're getting slaughtered by AMD. But predicting the future in this business isn't exactly straightforward. You try guessing what an efficient CPU will look like in 6 years time, and I'm willing to bet that you'll be wrong.
Intel doesn't "cut corners", and they do put "real effort" into their CPU's. The Core 2's also do not run "emulated 64 bit" (whatever that means). They're just as capable of running 64-bit code as the A64's are.

As for their "quad core"? Yes, it's an ugly hack and it has some significant drawbacks, but as Angrymob pointed out, it *works*, and it's out *now*. AMD doesn't have a quad-core system of any kind that works or is out now

About busines ethics? Well, we'll see how the AMD vs Intel lawsuit turns out. But there definitely are some things that sound a bit fishy.

Uh, you got it the wrong way around. From the end-user's point of view, it is a quad-core system, and it works like a quad-core system. It produces quite a bit of heat, but that can be dealt with. And it's not "more expensive". More expensive than what? The competition that isn't there? More expensive than AMD's CPU that will be launched in 9 months?
Their quad-core is priced pretty reasonably for a product that has no competition. That's what AMD did when they were on top too. Obviously, if you have a product that people want, it's expensive to produce, and there are no competing products, you set a good high price. That's not illegal or criminal or unethical.

Uh, the point is to have a superior product to sell? To be able to sell a product for 9 months without worrying about your competitor. To make money? That's business 101, and it's so obvious that both AMD and Intel understand it perfectly well. I fail to see how they're lying.

Yep, because for three or four years, AMD were the best. Now? Core 2 outperforms the best AMD has to offer by 20%, while offering all the same features. That sounds like "best" to me, don't you think? And now Intel has launched a quad-core cpu (it has four cores on a single CPU, hence it is, for all practical purposes, a quad-core), which performs like a quad-core system, while AMD doesn't have anything like it. Does that not means Intel is "best" currently?

If anyone are lying, I'm sorry to say it's you. Accusing Intel of being inferior, of cutting corners, of "faking" their chips is just nonsense.

Uh, I don't think you understand. In 64-bit mode, the Athlon 64 performs 6% better in relation to the C2D than it does in 32-bit mode. In other words, where it's normally 20% slower, it's only ~14% slower in 64-bit mode. That has nothing to do with the amount of cores, or whether a system is "true" quad core or not.
Also, by the time AMD releases their quad core, Intel will be pretty close to launching their native quadcore design as well.

Umm... What do you mean? That not many applications are able to take advantage of it?
True, but that's the same with AMD's upcoming quadcore design, and it's the same with both companies' dualcore designs.
And what matters is that some applications *can* use it.

No and no.

 

For every successful (or not so successful) product launch out there, there are probabally tens to hundreds of different possible products/solutions explored then ditched, and some of those might be a better solution than the one being pushed to market. Witness the Intel NetBurst architechture and the "GHz" push Vs. the more efficient P-M (now Core) uArch. by Intel's Isareli design team. It wasn't until AMD came in and ruined the party that upper management had to do a 180 shuffle and have to pull the plug on NetBurst earlier than they wanted to. Also, if it wasn't for AMD, Intel won't be pushed into exploring multi-core designs that quickly.

Of course, Intel bringing in Quad Core this early on brings no real tangible benefit to current set of software that 90% of average cosumers uses, however, they obviously feel that there is a market niche for it (especially in the server/high performance computing area where applications are already multi-thread native). And that also pushed AMD into launching their 4x4 / QuadFX platform, which simply put, is disappointing (again reviews of which are avaliable from Anandtech, Tomshardware, and many other sites). But, as a business, you have to keep up with the game and your opponent. If you say "rush a quad core CPU onto the market to "beat the competition" does not impress me" for the Intel Quad Core, surely, the same cannot be more true with AMD's current 4x4 / QuadFX platform. But yet AMD has to respond to the competition with what they currently got.

Technology and innovation does not stand still, you either innovate with the times, or you'll be quickly brushed off as a "has been". Anyone still remember Cyrix, or NexGen?


I too, also own a small computer business (although not my full time job), and I recommend whatever fits best and offers the best price/performance/power ratio to my customers. Right now, Core 2 Duo gets my recommendation from the middle mainstream to the high end, and lower end Pentium Ds (805/820/915/925) gets the value-mainstream, simply because AMD is neither price, performance, or price/performance competitive in those market segments. The same is true with mobile/notebooks in that I'd recommend a Core Duo/Core 2 Duo/Centrino platform solution. It is only at the bottom end of the market where I'd recommend a Sempron/Turion (single core) over the Celeron.

I used to recommend and sell many Athlon 64 and Athlon 64 X2s in the Netburst Pentium 4 days, and I hope to recommend an AMD offering in the future again if they can get something much better on the market this year.


jess_paws:

Continuing on from the theme above, it is always a "chicken and egg" game with new products and innovations. Most applications for the masses/mainstream (i.e. games) aren't programmed for multi-core processors simply because to design/write a programme for multi-threaded execution is much more complex, just ask any Computer Science major or Computer System Engineer. And because the majority of the computer market only has a single socket single core processor, there are no incentive to programme multi-threadedly (not to mention it cost more money too for the development).

However, now that dual, or multi-core has deeply penetrated the marketplace, things will change very quickly, and very rapidly. Therefore, holding onto the thought that "dual core is useless" is old and outdated thinking, if a little naive (advance apologies if I offeded anyone here).

One should not hold on to the current and now, instead, should always look forward to the future and move with the times. Sure, "for now" it seems the whole dual (or multi) core is not worth its hype, and won't make your current (or old) programmes run any better or faster, but the trend and wave of multi-threaded programming and multi-core processing has well and truly started.

Remember when a long long time ago when everyone thought the world was flat, before someone come up with the concept that the world is round and that shocked the world? Or when Sir Issac Newton discovered a thing called gravity that ushered in a new wave of thinking into Physics and other sciences that outraged the Catholic Church back then? Well, same with the current dual/multi-core "revolution". Always think ahead, think outside the box, and innovate.

 

Never said Intel was out to lose money, Never stated their CPU right now is worse than AMD, I stated the things that have happened in the past to push me to AMD - The performance, the pride of the company, and the time and care they put and continue to put into their product... INTEL RIGHT NOW IS BETTER, TESTS PROVE IT. Designs are different, the way the companies do things is DIFFERENT, I am basing my conclusions on the things that have happened over the past few years, the exp. with the CPUs - Had Intel came out with a quad core CPU with 4 single cores, had quad core not been released shortly after their core 2 duos came out, I would have different opinions, I might even run one for my PC, but since they dont I will wait for AMD for my PC, and sell intels to the people that want to pay for them (which are very little) Unless of course they get a Dell that I get to fix which they got dirt cheap prices on. I am finished on this thread, picking my words apart and twisting them will not change my opinion, as I have stated in MULTIPLE POSTS.

 

Well I'm not saying dual core is useless I'm just quite proud that I realised that it is not the uber advancement everyone seems to think it is. Future maybe but for now single core seems to be fine. My friend who is a software developer, sysadmin and general uber geek is planning on a new gaming rig which will utilise a single core Athlon FX. He already has an X2 3800 but he's not fooled by it either as there are some games it just wont run. They don't tell you that, do they?

I reckon my next laptop will be a dual core, probably Core Duo, mainly coz I don't think I'll have a choice in the matter!

 

Technology and innovation does not stand still, you either innovate with the times, or you'll be quickly brushed off as a "has been". Anyone still remember Cyrix, or NexGen?

Of course - they didn't last long at all, and had sooo many issues....

 

I had a Cyrix 686! 133MHz but performed like a 166 or so they say! It seemed fine to me but I was about 13 at the time!

 

"And that also pushed AMD into launching their 4x4 / QuadFX platform, which simply put, is disappointing (again reviews of which are avaliable from Anandtech, Tomshardware, and many other sites). But, as a business, you have to keep up with the game and your opponent. If you say "rush a quad core CPU onto the market to "beat the competition" does not impress me" for the Intel Quad Core, surely, the same cannot be more true with AMD's current 4x4 / QuadFX platform."

I agree, I am not impressed with the current release of that CPU

 

True on all accounts. You're right, dualcore is more hype than anything else at the moment. (that, and multitasking)

People seem to believe that if they have more than one program open at a time, multicore will bring noticeable performance improvements. It won't.
A single CPU is quite capable of running dozens of programs at a time. It's been doing that for years until people got dualcore systems.

Unless people run multiple CPU-heavy apps, it just won't make a difference. And people don't often do that. Especially not now that most games are still singlethreaded.

So yeah, you're right, don't assume dualcore is a must-have in any new computer. It's going to come in handy in a couple of years, certainly, but that's not the same as saying that it's "the only thing".

Performance? Intel > AMD.
Pride? Well, right now AMD are the ones who've been resting on their laurels, scoffing at anything Intel does. A year ago, it was very different.
Time and care they put into their products? Well, if they have superior products at the moment, it's quite reasonable to say that they've put *enough* time and care into them, isn't it? Maybe AMD should have put more time into their K8L chips, so they'd be ready faster?

When that is said though, I like AMD. I'm perfectly happy with my A64 desktop, and I don't plan to upgrade to a Core 2 system.

 

Once again, all of this is right NOW, AMD does not even have a available workstation Quad core CPU out right now, so how can performance even be measured? your going off benchmarks from June and Nov. - You run a brand new CPU against the A64 series of course its going to spank it - I am going to wait till 2q 07 till I make my decision for which company I choose for quad core - whether it be Intel or AMD we shall see - You can't compare a x2 or 2 x2 to a brand new Intel Quad core socket and expect the technologies to equalize, espcially when the rival company already has plans for a new CPU - If the x4 was ready it'd be out right now and we might be saying different things, all this is speculation and will be told in the near future

- For the record, I have liked Intel, I like AMD more at the current moment, based on statements and PREVIOUS expeirence with Intel's products - That does not mean I will never go back to Intel, whoever makes a more stable product via the same product line for best bang for the dollar will be my choice and what I suggest to my customers, all will be told on 2Q07 -
Happy Posting!
No hard feelings (hopefully)
Josh

 

Simple. Pick the best each company have to offer, and compare. Intel has a quad core, AMD has dualcore (and 4x4, if that counts). If you need serious processing power *now*, that is the relevant comparison. What comes out in 8 months from now is less relevant.

But of course, if you're looking to upgrade in 8 months from now, you're right, it's pointless to compare anything now.

Pointless to expect them to even out, yes. Not pointless to compare.

So it was also pointless to compare the P4 to Athlon 64? Because Intel had plans for a new CPU? It'll also be pointless to compare the K8L when it comes out, because Intel has plans for a new CPU. Both companies *always* have plans for new CPU's.

No it isn't. It's not speculation because we're discussing how things stand *right now*. If you want to draw comparison to a CPU that will be released next year, then yes, you are speculating. The rest of us are discussing *current* CPU's.

'Cheers

 

Wasn't this thread supposed to be about heat issues?

I think this thread can be stopped.

 

Ahhhhh Jalf, I'd never have pegged you as an AMD fanboy!

Seriously though, I had a Mobile AthlonXP processor in my old laptop. I'm just glad the ****ed thing didn't blow up on me, it was heating up so much. Maybe the Athlon64 is different...and I'm sure you're benefiting greatly from that on-die memory controller.

All the best.

P.S. pkrfan, if all threads that deviated even slightly from the original topic were to be closed, we wouldn't have much of a community here.