Intel's Core 2 Desktop Processors Tested

Steve Kerrison writes "It's early morning here in the UK, but that doesn't stop us from being around to see the launch of Conroe and friends, Intel's newest desktop chips. Even a $180 Intel CPU can beat an Athlon FX-62 in a number of tests. Now that's bound to get the fanboy blood pumping, right? We've also taken a look at a pre-built system that's powered by the Extreme X6800 CPU, along with an nForce 4 SLI chipset. As you'd expect, it's quick."

Loss Leader?

Gotta wonder if intel can legitimately deliver at this price or if they are going with loss leader tactics to try and regain marketshare.

Re:Loss Leader?

Loss or not, they at least gave stockholders a little more confidence than the slaughtering over the last two years. This is good news even if they take a loss.

Re:Loss Leader?

Well, for now, the yields of the fastest Core CPUs is probably low enough that the average price of manufacturing could be higher for the cheapest chips, as they are a necessary part of the process, anyway. On the other hand, I think that the pure manufacturing costs for a (desktop) CPU tend to be quite a bit lower than this -- the big costs are the onetimers in development and investing in fab infrastructure. When that's already in place (for the current chip generation), it makes sense to use the resources available fully.

Re:Loss Leader?

Bearing in mind that for the first time ever, the silicon for the cores of their laptops, desktops and servers are capable of coming off the same wafer, and I'd say they're onto a cost-saving exercise here. Heck, even the Core Solo is a Duo with a core disabled.

Thing is; these are miles ahead of AMDs current crop, Intel could double the prices on them and they're still good value for money. If they're a good product, market share will come without trying.

Re:Loss Leader?

That has to be one of the most entertaining, yet informative, reviews I've read in a long time!

From TFLA (The Fine Linked Article):
"[Intel's] P4 chip has largely been having its ass handed to it on a silver platter by the Athlon64 family of CPUs from AMD."

and then later:

"But this is where their [(Intel's)] little parade comes to a screeching halt - why? Because in the most simplistic of terms, Conroe (dubbed Core 2 Duo) kicks the Athlon64 right in the balls and doesn't look back."

Now, my friends, *that* is how you write a review!

Re:Loss Leader?

Now, my friends, *that* is how you write a review!

And on which of the 20 pages the review is divided into, should I insert these witty remarks.

Typical "hardware enthusiast" article

The basic fact is that "hardware enthusiasts" are neither good writers nor are they all that technically knowledgeable. Their saving grace is that they are willing to spend days on the drudgery of benchmarking. We are willing to forgive their artistic pretensions in exchange for a valuable service. You just have to read the reviews with your BS filter on.

It is true that this is not the first time that Intel has focused on IPC, that integrated memory controllers are not evil, and that few people fully understand the detailed workings of SSE (definitely not me). These are all instances of marketing BS. But they don't really mean anything. The benchmarks show that the Core architecture has much better IPC than the P4, regardless of whether this is due to the extra pipeline, shorter pipelines, better cache, lower memory latency, etc. And the benchmarks also show that the Core has better memory latency than P4 despite the external memory controller. And lastly Intel has drastically improved the floating point performance of the Core processor over its predecessor, the Pentium M, thanks to improvements in the SSE unit, whatever those improvements may be.

This is always going to happen when a journalistic organ is supported by sponsors from the industry it covers. The editors are obligated to include a bunch of marketing BS. You can get valuable information from these compromised sources, but you have to read between the lines.

I doubt they'd do a loss leader thing

Most likley they are just having good yeilds. They've pretty much got the kinks worked out of their 65nm process with the Pentium Ds they made on it so it wouldn't supprise me that Core 2s are having high yeilds. High yeilds = low cost per unit. This is espically true if the yeilds are high, but mostly at lower speeds. Say 90% of chips work, but of that 90% 50% only work at the slowest speed. Well, just knock the price down on that and get it back in volume, hike it up more on the rarer fast chips.

If you look at their current pricing, it's not real supprising. You find you can get a Pentium D 65nm for as little as $175. That gets you a 3GHz one on their old 90nm technology. The price creaps up on the first incriment, a 3.2 is $217. However it takes a sizable jump then to $317 for 3.4GHz. The 3.6GHZ, if you can find it, is $500 or so. Past that, well there's only the "extreme" edition and that's over $1000 for 3.73GHz.

The jumps like that are normal. They can easily produce low speed chips and there's a large market for them so they are cheap. Maybe a couple incremental upgrades. Then you hit a knee and prices start jumping fast.

Based on their current pricing for their current high end, I don't see anything out of the oridinary for this new pricing.

first PC's?

so when will the first PC's come out with these?

Re:first PC's?

so when will the first PC's come out with these?

WWDC is August 7th.

The question asked when the first PCs will come out with Core 2 desktop processors. You gave an answer based on an unconfirmed and very uncertain Mac rumor (linking to Apple's conference/lovefest), then got modded up.

Unbelievable.

Here's the answer the GP was probably looking for (from Anandtech's conclusion [anandtech.com]):

According to Intel:

Intel Core 2 Extreme processor based systems and boxed product are expected to be available on the day of launch, 27 July. Intel Core 2 Duo processor based systems and boxed product [through places such as Newegg] are expected to be available from 7 August. Each OEM has their own product introduction / transition cycles based upon their target market segment and current product offerings. We expect some to offer product in August with more introductions extending through September. Check with the OEMs of your choice to get their specific message on system availability.

From what Intel is telling us, you shouldn't be able to so much as purchase Core 2 processors until after the first week in August, although you'll be able to get complete systems before then. At the same time, we're hearing that distributors already have some Core 2 parts in stock and will begin shipping very soon. While we tend to believe Intel's assessment of availability, we're hoping it's conservative.

Re:first PC's?

I agree with you. Way back when, in my first years in college (mid 80s), I had an Atari 130XE. A math teacher asked me if I had a PC, and I said "Yes." I was very annoyed when she berated me in front of the class, explaining that a PC meant "IBM PC".

So I stoop up and I asked "So you don't think a computer for personal use is a personal computer? So what should I call it? An egg timer?", and thankfully got a lot of chuckles from the other nerds in the class.

So she explained that PCs were used for things like word processing and spreadsheets. "I have a word processor and a spreadsheet."

She was like "oh, I didn't you could get those on an Atari."

I loved my Atari (in a strictly platonic way, of course).

More Detailed Review Here -

There's a much more detailed review up at HotHardware.com [hothardware.com]

So...

will this be enough to run Vista?

Re:So...

Now THAT was a stupid question...of course not. But it will be fast enough to run the latest in spyware in the years to follow.

Re: Just don't run Office 2007

>I wonder if we will ever see a Core Pentio?

Probably a year or two before Vista ships.

Kyle Bennet seems to disagree...

"Real World" [hardocp.com] testing of the new core2 duo's over at HardOCP seems to suggest that the hype is, well, Bullshot (Penny Arcade). He also savages... no, investigates, the other benchmarks with his normal subtle-but-robust manner :) It seems that the top of the line Core2Duo just barely beats an FX-62 numerically in actual game performance; statistically there is no difference whatsoever... As with all things, it comes down to perspective. I have no doubt that Intel are catching up to AMD, may indeed have caught up. However, I simply do not believe they have gone from lagging significantly to leading significantly at the same clock speed; Time, I suppose, will tell.

Re:Kyle Bennet seems to disagree...

One of the things with HardOCPs review shows one extremely interesting thing...

If you have a single high end card (7900), there isnt a whole lot of difference between the FX62 and the X6800, or even the E6700. Most games are GPU limited now, and will be until the next generation of cards is released in 3+ months (FEAR is really the only exception to this).

They didnt run any benchmarks at 800x600 or whatever, because those results are more or less useless. Who spends $500+ on a processor and $500 on a video card and plays games at that low resolution.

What matters if you're going to buy a new rig now is the price performance ratio. If you're a midrange gamer, your best bet is probably a E6600 and a $250 video card. Or an AM2 setup, it all depends on the prices AMD cuts their X2 line to. We'll find out closer to the end of this month what the deal is. Come August 1st we'll have a very good idea of which platform is on top.

That's almost always the case

The processor is generally the thing I upgrade the least because it simply has the least increase in demands. Video cards you can upgrade once a year and not be doing it too often given the advances they have. Throwing lots of RAM at your system is also a good idea. Processor? Well for gaming and most apps it just really isn't that big a deal. Get a good dual core of prett much any design you like and call it good. Hell if all you are worried about id gaming and not doing things in the background while you game get a good single core, games still don't make any use of a second core to speak of.

I moved from a P4 2.4GHz to a Pentium D 2.8GHz when I did a system overhaul not too long ago. Why such a minor processor upgrade, you might wonder? Well because the processor wasn't the issue. That 2.4 was plenty fast, for games at least. The graphics card was the issue and I wanted PCIe which my board didn't support. Had the board had the same socket, I would have just kept the processor. It was fine (though because of teh audio work I do I'm appreciating the dual core). I just got a dual core because they weren't that much more expensive and it has geek appeal to me.

The real useful thing, in my book, is that the Core 2s run cooler. Current processors have tended towards too hot. AMD is much better than Intel but even they put out quite a bit of heat at the high end. It sounds like the Core 2s are quite efficient for the performance they give. That's good because I value a quiet system and frankly, it's as good as I'm willing to make it at this point cooling wise. I'm not going water cooling and there's just no more air cooling I can do short of making the fans speed up.

I don't think I'd recommend these as an upgrade to anyone who already has a dual core AMD or Intel system. Unless you are doing simulations or rendering or something I just can't see the minor increase as worth it. Certianly not for games. However if you need to upgrade anyhow, these look like winners.

Re:Kyle Bennet seems to disagree...

Responding to anonymous... I must be mad.

But your point is accurate. Gaming is mostly GPU limited; my gaming system, an s939 amd64x2 3800+ with a pair of old GPU's (7800GTX 256MB) achieves equal or better gaming results than all of these.

I suppose the point is are such prices for CPUs currently justified when they wont have much impact on user experience?

No doubt the new entry level core 2 duo's seem to be the upgrade of choice to maintain near cutting edge; but a high end GPU seems to be a wiser spend than a new CPU for gamers.

As for video encoding et al, HardOCP had the same results in their "real world" testing as others, but at least they make an effort to simulate the way the "average" person might use the things; either way, I'll reserve judgement here until I see some 64bit results, since encoding in native 64bit will be the telling tale IMHO.

In any case, I think we are reaching the point of dimishing returns, a year old 2GHz processer already rips music as fast as the drive can deliver it, already transcodes video as fast as the drives can burn it etc... GPUs control gaming... It is nice to see intel returning to the game in a serious fashion and no doubt this will have positive results for the consumer if AMD try to match price performance. I was mainly trying to point out that the "benchmarks" aren't nescessarily useful in describing the performance of these beasts in operation.

err!
jak

Re:Kyle Bennet seems to disagree...

You have to be carefull with the HardOCP benchmarks. I only read the first several pages, but they're doing their usual "real-world" stuff... which means leaning on the video card to do most of the work. Naturally, CPU differences aren't going to show up much here. I appreciate them doing something to put a real-world perspective on things, but what I read doesn't change the fact that the Core2Duos (I hate typing that) are really stinking fast. But playing games that do most of the work on the video card won't matter, big surprise... it really looks like a clever yet still desperate attempt to be a raving AMD fanboy and prop them up.

Uh-oh, rant ahead, I tried to avoid it, I swear... ;)

I am a raving AMD fanboy, but I'm a raving AMD fanboy because they've made the best CPUs for a long time. They also have a wonderful motherboard architecture that makes very high bandwidth applications much easier to deal with. I find myself wishing I could plug the Core2Duos into an AMD motherboard... on-chip motherboard controllers would help Intel also. Ah, what do I care, I want to see a real motherboard built around a Cell, the overall system bandwidth is almost as exciting as the cpu. Too bad that means buying everything from Rambus... :(

Kyle Bennet is an AMD whore...

(cribbed from my post in anothe rplace).HardOCP are complete AMD whores here.

They do the power tests with power saving settings turned off. This gives AMD the edge at idle, mostly due to a lower transistor count. As other sites have shown, turning the power saving settings on (as one would expect) puts Intel far out front at idle.

How do they end that article?

" I would highly suggest keeping your eyes on AMD low wattage / energy efficient processors for those projects that require a noiseless solution."

So they make Intel look worse than they are, and yet Intel still wins at under load. What's the takeaway? Buy AMD.

In the gaming, after the Intel gets done smoking the FX-62, what do they say?

"It is very interesting that in all of our testing, both "what is playable" testing and "apples-to-apples" testing, the Intel Core 2 Extreme X6800 and Intel Core 2 Duo E6700 are very close in performance. In fact, in some games they are dead even. The price difference between the two is very extreme with the Core 2 Extreme X6800 costing $999 and the Core 2 Duo E6700 at $530. Does it look like the price is justified between the two for gaming? We can safely say "no" as far as gaming goes with this gameplay testing we have performed."

Then, when speaking of AMD, do they mention even the E6700 ($530) beat the FX-62 and the FX-62 costs over $800? Nope.

"As for the AMD Athlon 64 FX-62, all of our testing shows that it does trail the two new Intel CPUs in gameplay performance. So, if you wanted to point one out as being a "winner" then for sure it is the new Intel Core 2 X6800 and E6700. But, if you look at the amount of difference between the AMD and Intel CPUs, you will see that it isn't enough to amount to anything. The only game that we saw any real-world difference in was Oblivion, and even that was tiny. A little overclocking would clear that difference up."

Any mention of overclocking levels and how the Core 2 Duo overclocks well? Much better than an FX-62 usually. Nope.

What's their takeaway from the gaming section where a $530 Intel beats out AMD's fastest chip (at $800)?

"We have proven here that the flurry of canned benchmarks based on timedemos showing huge gains with Core 2 processors are virtually worthless in rating the true gaming performance of these processors today. The fact of the matter is that real-world gaming performance today greatly lies at the feet of your video card. Almost none of today's games are performance limited by your CPU. Maybe that will change, but given the trends, it is not likely."

and then

"Lastly, I would advise everyone that is thinking of rushing out and purchasing their latest upgrade that we are sure to see HUGE pricing slashes out of AMD before the end of the month."

Way to go HardOCP. Rig your tests, ignore Intel victories and make your summary "buy AMD".

You have zero cerdibility, HardOCP.

Also, you used bullshot wrong. Bullshot is a term for fake screenshots designed for games (like EA uses). It doesn't fit here.

OCAU's view

We have a comprehensive review on OCAU also: http://www.overclockers.com.au/article.php?id=4895 87 [overclockers.com.au] We compare the new high-end 2.93GHz X6800 and the 2.67GHz E6700, with the current Pentium D 955XE and AMD's A64 FX-62. Lots of info, loads of benchmarks and of course, some overclocking.

More in-depth review

http://www.extremetech.com/article2/0,1697,1989036 ,00.asp [extremetech.com]

OCAU also has a review!

http://www.overclockers.com.au/article.php?id=4895 87 [overclockers.com.au]

ads ads ads

this might very well be an interesting article but if they're going to submit me to atleast five flash ads on screen at the same time, this early in the morning, I think I'll pass.

Intel's Core 2 need programmer do morething

Intel transfer the difficult from Hadware to software, for get more power, programmer need more technology.

Re:Intel's Core 2 need programmer do morething

Intel transfer the difficult from Hadware to software, for get more power, programmer need more technology.

I completely agree! The Intel transfer the from hardware to software, get more on the power. In conclusion:

What?

I would have read the review...

..but it seems I need to upgrade to this new Intel processor so that my computer can handle all the ads in the website.

More good news for the consumer!

Even if it's just a shot at getting market share back, the fact that great things like this are being sold at lower prices only mean good things for the consumer. This, for example, is GREAT for me as a system builder because everything besides the Pentium D 805 was expensive. Now, with something like this, I can offer a (possibly) better CPU for not that much more.

More good stuff is coming from both camps, I predict.

Will this be The Return Of The King?

I've been holding my breath waiting for AMD to respond. Anytime now would be a good time for them to announce how they are going to counter the Core Duo. But the reality might be that they need to recoup their costs from developing the AM2 platform before they can make any changes.

I think the competition has been good, but if Intel returns wearing the performance crown then I think there is a real potential that the CPU market will be dominated by Intel more so than it has ever been before, with consoles being the main holdout. If these benchmarks are true, then the introduction of the Core Duo will be a real turning point I think. Keep in mind that these speeds are introductory and that in the past Intel hasn't had much trouble progressing to higher performance out of the same architecture.

What this all means:

What the benchmarks mean is that if you do a lot of media encoding, compiling, etc, you would probably benefit from conroe. HOWEVER, if you play games, regardless of whether or not you are on an AMD/INTEL system currently--if your system is pretty new--Do not upgrade at this time, as you are GPU limited, not CPU limited. Basically conroe: Large performance gains in cpu bound applications Little performance gain in gpu bound applications, obviously. This is good for intel. My systems for the past 7 years or so have been AMD. My next one very well may not be. The good news for everyone is that AMD is now the underdog again. Remember what happened last time they were the underdog? We got the athlon. The cpu speed wars went into a frenzy. For the last several years (5 or so) Intel has been sucking balls. Their chips have not been performance competitive. Clock speeds in both camps have stagnated. AMD chip prices have went way up compared to how they used to be. This is good news, AMD will go into overdrive developing their next-gen chips. Amd chips will become dirt cheap again. We'll see a new performance war. This is something i've been waiting for anxiously for a few years. I am very excited. Another thing is that the new intel chips take much less power than the old ones . (thank god)

load-balancing pandas

I not quite comfortable with the way in which Intel/AMD set out their roadmaps to gradually increase the number of cores. It seems a bit fast for me, something done more for generating sales than any substanial performance improvement. By the time any substantial portion of software products are developed from the ground-up to be optimized for dual cores, the number of cores would likely have moved on. It won't be fun to have to upgrade your processor every two years just to get those extra cores that will be woefully underused by what would most likely be a market full of un-optimized software. Load balancing would really only be the way software makers would bother to optimize their code for anythimg more than 4 cores. My understanding is that load-balancing threads or whatever on cpus is not an exact science, and can quickly kill performance when not done right. Can it be justified to 'upgrade' to a newer more-cores processor every two years if you don't see a corresponding increase in UT2009/2012 framerates, or better perfmance in MS Excel...?

Great news for the low end systems

Quote from the article: "Intel, then, has moved the goalposts as far as consumer-level CPUs are concerned. Its low-end Core 2 Duo parts are more than a match for anything that has come before." Thus, Intel raises the capabilities on the low end systems. This is great. But besides gaming, are there anything needing such performance boost?

Re:Great news for the low end systems

Quote from the article: "Intel, then, has moved the goalposts as far as consumer-level CPUs are concerned. Its low-end Core 2 Duo parts are more than a match for anything that has come before." Thus, Intel raises the capabilities on the low end systems. This is great. But besides gaming, are there anything needing such performance boost?

• Google earth
• Windows Vista
• Sun Java
• Flash (Hell my athlon 64 chokes on some some flash content)
• Adobe Photoshop
• Adobe premiere
• h264 decoders/encoders
• XGL compiz

... I'm sure you can come up with plenty of other examples.

Energy efficiency

Everything else aside, that was the one thing that interested me the most about the review - the fact that the new conroes are allegedly going to be consuming about half as much power as current desktop chips. Why is this important? Well, if such gains can be made on the desktop, I'm _really_ looking forward to the laptop chips. Maybe the 7hrs claimed battery life by laptop manufacturers will actually be accurate in the near future.

Re:Energy efficiency

Maybe the 7hrs claimed battery life by laptop manufacturers will actually be accurate in the near future.

I don't think it has been innacurate until now.

I own a Fujitsu Amilo V2000 laptop (in the UK) which uses the original Intel Centrino chipset. I work mostly at home, but am on the move once or twice a week. Several times early in its' life (first few months while the battery is fresh) I had come home in the evening from an onsite job, then got up in the morning and switch the laptop on and started work only to have the battery warning (10%) give me a nudge around 4pm (from a 9am start). My work is web development, so while it's not too intensive I'm running email, web radio, text editors etc. constantly. Admittedly it was running on a wired network, and using the built in wireless chip results in a loss of an hour or two from that figure...

I was completely amazed the first time it happened - forgetting to plug it in I assumed it would die a couple of hours later but it lasted almost the entire workday. (Other notes about that model : the battery itself died after 6 months, how annoying... and the screen is a bit glarey but overall I was very happy with the laptop.)

Noticable Difference?

Even though the benchmarks show that the intel conroe beats the amd fx, the real question still remains, the value. Would you honestly notice a difference in fps when both processors were running relatively close in frames per second. Maybe the conroe can get 20 more fps per second, but is that worth the extra money. Amd is notorious for being less expensive then Intel. Either way you could run the top of the line games, its just a question of which allows you to get more bang for your buck. If Amd sets a signifigantly competitive price, then it really doesn't matter how well the processors perform, people will choose whichever one provides the most performance per dollar value. While the Conroe beats the Fx in the performance battle, it still has not yet won the war. Let the price battle begin.

Its all about competition.

The FX-62 was released to be in mild competition with intels next offerings, that to me was obvios, it was not intended to be the final product you have to chose between. Intel however IMO have released the conroe to be in a pissing contest with AMD, but AMD will just piss further then Intel can reach with the conroe, putting them behind again, and putting hte stupid people who rush out to change to an intel system just because the conroe is statistically better.

Quiet System Technology

From TFA:

The 965 Express chipset family supports Intel Quiet System Technology, which intelligently manages processor and system fan-speeds in relation to core temperature, ensuring the fan(s) are spun just fast enough to keep the processor from throttling.

Sure, it sounds like a feature, but it also sounds like a way for them to engineer their CPUs to go bust prematurely, but not prematurely enough to be in warrenty. Now who would that benifit?...

Erroneous price/performance in headline

The headline states that "Even a $180 Intel CPU can beat an Athlon FX-62 in a number of tests" but if you read the article, the $182 Core 2 Duo E6300 (1.83 GHz) chip wasn't tested [hexus.net]. All of the performance data relates to the $224 Core 2 Duo E6400 and pricier chips. The results are impressive, but I think the "$180 chip beats Athlon FX-62" deception should be pointed out to anyone who didn't pick that detail up from RTFA.

Some more real-life benchmarks

Rick Brewster of Paint.NET fame tested [msdn.com] two Core 2 CPUs with his own benchmark.

Sadly....

There are Athlon-FX 64-bit processors that still beat this benchmark by 2-3 thousand points on CPU-Z. I'm no Intel nor AMD Fanboy (I'm a Cyrix fan from beginning to end, and if you can't understand why then compare Unreal Tournament under a P2+MMX 233 against my Cyrix MII-233MX processor with the same RAM (Type+Amount,) and video+sound card. (Hint: The Cyrix beat the ever-loving shit out of the PII by a blazing 25 FPS (I don't use large abstract numbers, I use real-life performance tests/observable results, like the rest of you overclock geeks should use.) So my question is (Compared with my personal RL observations against Intel and AMD's claims,) is who the hell is bothering to rate their processors by what REALLY counts? By this I mean MIPS (Millions of Instructions Per Second [performed]) as opposed to GHz? Actually, I'd like to see MIPC as in Millions of Instructions per Cycle the processor is capable of. I don't care how many times it can do the same thing over again - how many times can it do the same thing, using a more efficient algorithm (like the divide by 10 thing, instead of dividing by 10 we multiply by .1) over and over again, and how can people finally realize the more true and better optimizations for their processors so we can have a far more accurate measure of performance? Benchmarks, AFAIC, (As Far As I'm Concerned,) are just an e-penis waggling contest. Gimme something I consider to be real results if you want to market your processor to someone that has even a modicum of a clue (And I admittedly have a very low clue about processors, but I'm still not a n00b when it comes to them. I've played with processors since I owned an Intel 8088 Packard Bell that I ran Jill of the Jungle and a few BBS servers on, funnily enough I was only 7-8 at that time and had full command of DOS and most of the standard BBS door-game system. Yep, that's sad, and my father was around computers FAR more than I ever was at that age.)

Re:Sadly....

thats it...you're fucking banned from using parenthesis :P

now goto your room

Re:Sadly....

So my question is [...] is who the hell is bothering to rate their processors by what REALLY counts? By this I mean MIPS [...] as opposed to GHz?

This figure is meaningless in modern CPUs as processors perform a variable number of instructions per cycle depending on (a) what instruction (b) whether data is in cache (c) whether there are pipeline stalls to account for. The best you can hope for is the number of cycles of some standardised performance test that are executed in a specific time, which is what articles like this one provide.