Core 2 Duo Released

[PhogHawk]

What do all of you think about the newly released Intel Core 2 Duo Processor? The benchmarks look astounding, but I don't really like Intel a bunch. Should my new system build be with a Core 2 Duo E6400 or an Athlon 64 X2 4200+?

[BlacKwidoW!]

I agree they "sound" awesome but I'm a AMD fan myself. Speaking of processors the new AMD for the laptop with the TURION sounds pretty awesome too.

[PhogHawk]

I'm an AMD fan too and I'll probably get one. But these new Conroe processors are ridiculously beastly.

[Nokarot]

Intel Centrino Duo, is that what you're talking about?

If so, I got one in my laptop.. Its fairly beastly. I'm not one of the guys who runs thousands of benchmark checks, but I haven't had a problem with it. 1.86GHz Dual Core, running the newest games just fine.

[Rip]

No nok, he's talking about the overreaching core 2 architecture that most review places were under NDA about their specific test specs until late yesterday night. Since then a lot of sites have put up their info, and its very impressive. For most applications, it appears that the $500, and in some cases the $360 version beats the FX-62 hands down, all at lower temps and less power usage. (the FX-62 is $1000 remember). Not to mention they appear to be readily overclockable by raising the FSB (multiplyers are locked except on the extreme (2.93ghz) edition, which also sells for the $1000 mark but gives AMD a rather nasty beating. Having said this, when it comes to gaming, [H] demonstrated that most games currently out are GPU bound, that is the graphics card is the real bottleneck. So unless you've got two nice 7900s in SLI or 1900XTs in a crossfire setup, you won't see a huge difference between top end processors. In the long run however as much faster GPUs come out that will eventually change. I'm very impressed by the core 2 stuff, and merom looks impressive for the laptops. I will probably wait until Q1 2007 when the quad core version of the core 2 architecture is out, but we'll see what AMD has to offer or if there are other circumstances. I can't wait to build a new desktop.

[asphix]

Unless AMD hits some major price cuts (they're claiming there will be.. however the "how much" is yet to be disclosed), you'd be insaine to purchase an AMD chip over the C2D as performance/price C2D flat out wins. If you choose AMD due to brand loyalty (not the fanboy kind, but the "I really like AMD and have used their products for a long time thus I feel comfortable purchasing their products" kind) then thats certainly fine. But if you're looking for the best bang/buck and best performer C2D is the logical choice.

Personally, I'm all set for C2D but wont be purchasing one for a long while (probably not till December or so). I got myself an Asus P5W-DH-Deluxe and an 805D along with a Zalman CNP900 CPU cooler to hold me over in the meantime (processor will cost you just under $100 right now). I have the baby OC'ed to 3.76GHZ which is prime95 stable for 8+ hours (only tested for 8 while I slept one night). I've been able to run benchmarks up to 3.9 ghz but errored in prime after 3 or so hours.

But yes, all preferences aside, if you're looking for the logical choice it would be C2D.. now whether or not you can find one, and at the MSRP none the less, is another matter entirely. I saw some x6800's on Newegg.com couple days ago selling for $1,300 or so which is a good 300 over MSRP. THe fact that newegg was doing that doesnt bode well for those looking to be an early adopter at what should be normal retail pricing.

I'm still trying to get over the 4.1ghz overclocks and sub 10 second 1MB SuperPI's some guys are pulling over at extremesystems. The other thing to keep in mind is Kentsfield quad core which is due out in Q1 07 which the enthusiast level motherboards being released along with Conroe should support (965p chipset and i'm fairly sure some 975x). AMD doesnt have anything substantial due out until K8L, and while its said to expect that late 2007, its not currently on any of AMD's roadmaps and rumor has it that it may not be ready until early 2008.

Bonus with AMD is that if you get an AM2 based platform, it will be DDR3 compatable as long as you get an AMD chip that supports DDR3 when they are released (future proofing is always a nice incentive).

[WearyGamer]

i would go for the AMD 64 bit thats what i have click on my xfire and loook at my rig

[Rip]

asphix, you forgot 4x4, although in all honisty, I don't expect it to make a big splash considering the power usage requirements and actual performance. Fun stuffs.

[asphix]

asphix, you forgot 4x4, although in all honisty, I don't expect it to make a big splash considering the power usage requirements and actual performance. Fun stuffs.


Yeah, 4x4 may be good but I see it as marketing hype more than anything. Supposidly rumor has it that initial tests with some 4x4 based solutions have yeilded an 80% performance increase in some applications (saw this on the Inq I believe).

However, 4x4 is more likely there to compete with Kentsfield (Intels Quad core based on the Core 2 architecture) due out in Q1 2007. The 4x4 platform has a bunch of pitfalls as well. A speculative illustration of this point would be:

Intel
1x quad core cpu @ 2.6 ghz costing ~ $500

AMD
2x dual core cpus @ 3.0ghz costing ~ $350 ($700 for both)

I throw in a 3.0ghz AMD offering as that is what frequency K8 would generally be comparable to a 2.6ghz core 2 offering. Since Intels quad core would be manufactured on a single die (and chip), it would therefore be cheaper/easier to produce not to mention intel is on 65nm where as AMD is still on 90. THis price difference becomes significant when you consider having to switch up two processors every time you upgrade (i'm assuming for optimal performance you will want to marry the stats of the two processors in the system) not to mention the cost of the motherboard etc etc etc. Basically 4x4 has been presented in concept.. and it sounds really great. It hasnt been stated how it will be marketed or positioned toward consumers. Thus I am unsure what sort of future it holds in all practicality. However, SLI took off so who knows...

Of course, I reiterate that these numbers are pulled directly out of my rear based on speculation and my insticts and admit they're probably pretty well off the mark.

A really interesting read that helps explain where I developed this perspective and gives a good simple insight into just how AMD will be hurting in the coming years, is this posting: http://www.overclockers.com/tips00996/

Another really great read that gives great insight into what goes into producing a line of processors can be seen at the link below. Its unrelated to AMD/Intel (as it pertains to the CELL processor and the rumors of horrible yeilds that were floating around last week) but helps give some perspective.

http://episteme.arstechnica.com/eve/forums/a/tpc/f/77909774/m/361004600831

[PMS Wedge]

This maybe should be filed under site suggestions, but I'll make the comment here.

I've been thinking that maybe it would be a good idea to have a section on hardware - troubleshooting, or building a rig, etc.

I wouldn't mind contributing some advice. I do have a bias towards Intel, though, seeing as how I work there. I wouldn't be able to answer everything (due to NDA type stuff), but I can put in 2 cents on basic questions.

Also, enjoy all the Core 2 Duo/Core 2 Extreme reviews. :) They've been keeping me super busy working on all this stuff! And for those who have AMD machines... enjoy the price cuts. :) Competition = good!

[Rip]

I've been reading ars pretty regularly, they've got some good stuff.
Speaking of core2, looks like intel has decided to play hardball, quad cores comming out Q4 of this year!
AMD said the FX-62 should be taking a price chop of about 55%, which will bring the benchmarks for price performance much closer to essentially flat. Good for the consumer. :)
And while i'm a hardware guy personally, I just don't know if we have enough demand here for a hardware forum. I do like hearing about the stuff though, esp from people in industry.

[asphix]

/agreed

I was pretty happy to hear that quad core has been pushed up, as I decided to hold off on a core 2 duo purchase till DX10 cards are out. Now I have the option of going with a kentsfield :)

[PMS Wedge]

I've been reading ars pretty regularly, they've got some good stuff.
Speaking of core2, looks like intel has decided to play hardball, quad cores comming out Q4 of this year!
AMD said the FX-62 should be taking a price chop of about 55%, which will bring the benchmarks for price performance much closer to essentially flat. Good for the consumer. :)
And while i'm a hardware guy personally, I just don't know if we have enough demand here for a hardware forum. I do like hearing about the stuff though, esp from people in industry.


Guess we can just keep it in this forum, then. :)

I don't think the price chop for the FX-62 is that much - only like 20%. The X2 processors, though, are getting huge price cuts. Some more than 50%... especially the 5000+.

For those who read the HardOCP Core 2 Duo gaming review... Firingsquad wrote a response:

http://www.firingsquad.com/hardware/jakub_rant_real_world_benchmarking/default.asp

Kinda an interesting read - HardOCP kinda blasted all the other reviewers as "lying" to their readers, so I'm guessing more of these will pop up.

[asphix]

Interesting read, HardOCP posted a "reply" (http://enthusiast.hardocp.com/article.html?art=MTExOCwxLCxoZW50aHVzaWFzdA==) on their site.

Although, I must admit this whole concept isn't entirely a new one for me. I've been through all this a year ago when bit-tech.net made their switch from standard benchmarking to "real world" benchmarking. Its hard to see any of whats going on between FiringSquad and [H] and anything more than a juvenile argument.

Change is good, reworking how we think spells PROGRESS. I can understand how firingsquad might have felt insulted by [H] questioning how reviews are done today.. but to dedicate an article to defend themselves IMO is a little extreme, though possibly was necessary to help reiterate the necessity and validity of traditional benchmarks / reviews. Thats not to say [H] (again in my opinion)didn't launch a highly unneccessry retalitory response. When will we all just grow up?

I think both benchmarks are useful in this day and age. The standard benchmarking method gives a great overview on performance gains and relative performance in different scenarios. I has become pretty de facto these days for readers to check multiple sources and cross reference results in order to broaden that general overview across an array of hardware configurations.

Chances are, a reader who is at a level where they don't cross reference and personally analyze benchmark/testing results, and thusly base a decision off one review (might I add, theres nothing wrong with that, to each their own.. right? As long as your happy with your decision and you wake up in the morning feeling good.. thats all that really matters) woudlnt benefit by a more radical benchmarking methodology as their inability to read between the lines and apply the information to their experience/understanding of the subject to develop a personalized understanding of whats in front of them would end up confusing them. Whew- that was a long sentence :)

I really like the "real world" benchmarking that bit-tech has been doing for close to a year now, but I still like the traditional benchmarking styles as well. They are both very usefull, both take a lot of time and coordination and both help me when making a purchasing decision. Sometimes I just want to see numbers (provided the testing variables are supplied to lend perspective to the results) and skip all the in depth explanations and analytical verbiage. Sometimes I want to hear the testers opinion. Neither is wrong.. and I woudlnt want to see either dissapear.

So, again.. I dont blame HardOCP for expressing they believe the traditional method of benchmarking is showing its age and promoting their new way of doing things. Its a lot of work to think outside the box and try new things. I dont blame firingsquad for feeling a need to explain the practicality and usefulness of such classical/traditional ways of doing things. They're there for a reason... However, I believe both publications execution on such thoughts could have been pulled off in a slightly less hostile, more insightful manner.

I've also got a feeling we'll be seeing a LOT more of this stuff. Hardware journalizm is beginning to reach saturation. With so many products out there, and so many people who love them.. there are a lot of people who want to share their ideas on said hadware and be in a position where those ideas/opinions are revered by their peers (i'm a culprit of this one myself, just look at my posts in this thread). Due to the saturation, the next logical step is to diversify and make your way of expressing your opinion/ideas or testing results stand out.. with that you get people who embrace the change, people who resist the change and a whole slew of people inbetween.

Thanks for sharing the link :) And sorry for the book of a response. I love a good conversation/opportunity to throw my opninions / ideas out on the chopping block.

[Rip]

[H] had one point, and their review I believe improperly addressed that point. The fact of the matter is that current games are mostly GPU bound as opposed to CPU bound. So having a core2 ee as opposed to the $360 or $500 version of the chip won't make a difference now unless you invest in a crazy crossfire setup with respect to gaming. Its a valid point. However, that doesn't mean that tom's or any of the other benchmarks aren't valid, and while [H]'s gaming relativity point is important, looking at tom's or anand gives one a good indication of future performance when the GPU's move forward and you're not GPU bound. [H] shows real world performance for people without thousands to burn, while tom's shows future performance of what people who do have that $ can get.

Having said that, I happen to do some 3d rendering, video editing....etc on occasion, so for me its important to note the other benchmarks available where things are very much CPU bound. Now I just want to get some info on the kentsfield (sp?). Wedgewu, do you know if they will be shipping out press samples of that Q3 / Q4 timeframe? I know gotfrag got some of the conroe's to review, but then again I still had to wait like everyone else for the specs even though i'm staff....stilly NDA's. (eh well i guess they're necessary). What can I say, I <3 hardware.

[asphix]

ES are already floating around and accessable to the few who know the right people. Coolaler over at Extremesystems.org threw up a peek of one he got his mits on a few weeks ago.

He ran it on an MSI 975x board.. you can find the forum post with pics here (http://www.xtremesystems.org/forums/showthread.php?t=103982).


and just because its too cool not to share... check out (http://www.xtremesystems.org/forums/showthread.php?t=107353)his world record super pi.. sub 10 second with a Core 2 Duo running at 5.3ghz. And on a badaxe none the less! :-D

[Rip]

interesting. I'm a tad skeptic with how badly photoshopped the first image is, but interesting none the less.

[asphix]

How do you mean? He obviously burred out anything that pertains to a serial number on the chip for obvious reasons. Other than that I see no photoshopping involved.... and if you go into the thread a couple pages he runs benchmarks and has cpuz up with the registered 4 cores.

Of course some of the numbers may not match up as CPUz and other applications arent updated to support the chip (usually vcore and what not).

For what its worth, if it helps lend to the credibility, these are some of the same guys who had ES (engineering samples) of conroe months ago. Do a search through the forums and you can find the threads along with benchmarks and what not.. of course they were running on modded badaxes back then as the rev304 (conroe supported) werent out at the time.

I believe it.. and the steppings/revision in the CPUz indicate it is indeed an ES.

**edit**

its also CPUz verified (http://www.xtremesystems.org/forums/showpost.php?p=1528175&postcount=72)

[Rip]

the CPUz is what makes me believe it, but the actual chip on the top looks funky (in addition to the obvious blurring.

[PMS Wedge]

Wedgewu, do you know if they will be shipping out press samples of that Q3 / Q4 timeframe? I know gotfrag got some of the conroe's to review, but then again I still had to wait like everyone else for the specs even though i'm staff....stilly NDA's. (eh well i guess they're necessary). What can I say, I <3 hardware.

I actually don't know, but I believe that we've announced early 2007 for Kentsfield, so Q3/Q4 would be an appropriate timeframe - more likely Q4.

Can you believe that I sat on Core 2 Duo numbers for almost a full year? I knew how crazy good this CPU was back in last August. So incredibly difficult to contain that kind of excitement...

As for the HardOCP fiasco - from what I understand it's not so much the fact that Kyle was doing "real world" testing, but rather how insulting he was about other reviewers. I completely see where he is coming from in a testing environment, as timedemos often don't include any sort of AI/Physics, which are rather CPU-dependent. However, the tone that he used, and the fact that he broke NDA and published an hour earlier than everyone else has alot of the reviewers rather peeved.

[Rip]

The tone really did frustrate me in the [H] review, despite having valid points. (Although if i'm going to spend upwards of $500 on a processor, I want it to perform well into the future).
It doesn't surprise me the numbers have been floating around internally for that long, but I'm really glad Intel went in this direction (not that there was too much choice I suppose). Regardless of market share, AMD was starting to get complacent and smug. The competitive scrambling around has been quite good for consumers. (Although stress on engineers might be a different story ;) )

What I'm hearing on the kentsfield is that it will be out by the end of the year. For example here's an news thing from dailytech (http://www.dailytech.com/article.aspx?newsid=3421) saying 2006 release. And there's a bunch of others. Since its essentially two conroe's in one package we're not really talking about a huge technical change (AMD of course will say its not true quad core, but close enough for consumers). I wasn't planning on doing my next rig until mid-spring next year (around the end of the school year), but all this stuff has gotten really tempting. (And this WD Raptor X I won at E3 sitting on my desk is just calling for a PC to be in haha)

[Ender2492]

I have no clue how you guys afford all this. Some of those cards are $500 + and require good expensive motherboaords. How do you do it.

[Rip]

I have no clue how you guys afford all this. Some of those cards are $500 + and require good expensive motherboaords. How do you do it.
we don't ;-)
My desktop is an Athlon 2400+ and a 9600XT, and my lappy is a centrino 1.86. Sadly i'm not rockin the heavy hardware, and the Raptor X I won...you wouldn't see me buying it stock anytime soon.
Having said that, I still love talking about it and researching whats going on. I can't quite say i've got as much knowlege as say some of our members who work in industry, but its nice being able to learn stuff from them.

[Ender2492]

I can't even run second life because my graphics card though...

[PMS Wedge]

we don't ;-)
My desktop is an Athlon 2400+ and a 9600XT, and my lappy is a centrino 1.86. Sadly i'm not rockin the heavy hardware, and the Raptor X I won...you wouldn't see me buying it stock anytime soon.
Having said that, I still love talking about it and researching whats going on. I can't quite say i've got as much knowlege as say some of our members who work in industry, but its nice being able to learn stuff from them.

Which Raptor do you have? The 150GB one? Cuz that one is a BEAST. Holy cow. There's a version where the top is clear plastic, so you can watch the thing spin... sooo awesome.

And to Ender - I can't afford this stuff, personally. I just happen to work in a job in which they dump these things onto my lap. :)

[asphix]

I have no clue how you guys afford all this. Some of those cards are $500 + and require good expensive motherboaords. How do you do it.

I usually buy smart.. not top of the line.. though about once every 5 or so years I'll go nuts and drop $2-2.5k on a new box.

There are lots of options to build a budget rig that will knock your socks off. If you recycle parts such as your case, optical drives, power supply.. etc etc etc.. all you really have left to keep on a regular upgrade cycle is Motherboard, Ram, CPU, Video card.

Intel Pentium D 805 - ~ $100USD
Intel chipset motherboard with 3 - 12 phase power ~$150-250
Memory (1GB) $180
Nvidia 7900 GT $250

Overclock the processor to 3.4ghz on a stock air cooler (as easy as changing a single parameter in bios usually.. this chip is an excellent overclocker due to its high multiplier.. so you can usually get away with a cheaper motherboard/ram/psu combo and cooling solution).. and you have a really nice rig for a max price of $780. You could probably get away with more around $600 if you eye slickdeals.net for great deals or buy 2nd hand off of e-bay. Ebay is going to be on fire in the coming months with people cycling to these new products.. especially once G80 and ATi's DX10 equiv. products hit the market.

[H2OSilentCid]

Just got myself a pair of BFG 7900 GT OCs which they are nice. On the new Intel chip, I not a fan of Intel and this new stuff is just a small win for them. AMD will release the quad core as well thier new ati stuff soon. so I am not worried at the moment.

[asphix]

Some more Kentsfield (http://www.xtremesystems.org/forums/showthread.php?t=109054) impressions/benchies from another guy over at xtremesystems.

I'm unsure if I should pick up a C2D this fall or just hold out and get a Kentsfield.. I guess it'll depend on how useful 4 cores will be in relation to my average useage.

[PMS Wedge]

Just got myself a pair of BFG 7900 GT OCs which they are nice. On the new Intel chip, I not a fan of Intel and this new stuff is just a small win for them. AMD will release the quad core as well thier new ati stuff soon. so I am not worried at the moment.

Out of curiosity, why are you not a fan of Intel?

Or really, why are you a "fan" of a company? Why not just go with whomever is better?

[viskey]

i'd go for AMD 64... which i have on my comp :) but thats just me