Intel, AMD, And Reseller Success: A System Builder Weighs In

[Sharft6]

I've never heard someone claim that intel has a better heatsink mechanism than amd before

 

[Lans]

I strongly disagree with the implication that AMD's CPUs are less "reliable" (I used AMD CPUs for years and never had any issues) than Intel's but otherwise I think Jon Bach's assessment of the current market is correct. However the EU ruling is for time period of October 2002 to December 2007 (NOTE: "Core 2 brand was introduced on July 27, 2006", Wikipedia) and not the current market... Today, in the consumer space, Intel is just kicking AMD's butt even with the Phenom II out although less so than before... As an AMD fan and more importantly as someone who do not wish to see the return of $1000 for an entry level CPU, I am greatly saddened by the current situation. At least things have been trending up for AMD.

There seems to be some truth to heatsink retention but I never ordered a PC. Although personally I never had any issues the times I have shipped my own white box back when I was in college (Athlon Thunderbird and XP machines)... I guess it could been luck and/or AMD systems need higher care in shipping & handling department. Still, I think the killer for AMD is system builders footing their own bills for CPUs broken during shipping. AMD being a much smaller company, I can completely blame them either...

I agree Intel is doing great today but I still think the EU ruling was right. As noted much earlier, the time frame we are talking about is, for the most part, before Core 2 was introduced and, as someone who assemble his own PCs, AMD was hands down the better choice then yet their market share did not grow. EU's ruling reveals that it is because Intel paid outlets to not use or delay AMD products. Rebates are legal but I feel sorry for Nvidia and personally I think is much more a gray area when you have only limited competitors and one owns 80+ % of the market.

Anyways, if you are talking about heatsink retention, I can't see why you would go for Pentium 4s (again pre-Core 2 days)... Then Core 2 came out...

 

[cokenbeer]

I honestly believe that AMD after being blindsided by C2D took way too long to respond. It wasn't until P2 that we see a processor capable of changing their image.

When people building 5000 dollar systems are using Intel, and people building 1500 dollar systems are using Intel, it's only natural for people building budget systems to WANT Intel. It has to start with a positive company image and build from there. I think the the Phenom II is a great start, and we now see people who are spending 1500 dollars going PII because it frees up some cash for better components elsewhere. AMD has a great platform at the moment, hopefully their next line of chips are the beast we've been waiting for. If we don't see pressure from one side, then the other side stops becoming innovative and starts pushing mediocre new features. Core i7 is pretty awesome, but it's nowhere near the bombshell that C2D was.

 

[blackened144]

[citation][nom]ohim[/nom]And i`m still a little bit amazed how this guy never looked back at AMD since K6 line processor ... well on what planet did he lived when Intel got only Pentium 4 and AMD was mopping the floor with them with Athlons 64 ? that period was still bad and AMD didn`t sell at all against those P4 ? how the hell can you make an article based on this guys words ...[/citation]
It makes absolutely no difference if the AMD CPUs were wiping the floor with Intel. He said he customers did not want AMD CPUs. So, using your logic, he should have only been trying to sell CPUs that his customers do not want?

 

[pogsnet]

How can you tell that AMD are not reliable and not performing?

Intel E8400 or Q8200 approx $169 newegg
is pawned with

AMD Phenom II X3 720 $145 newegg

http://www.pakgamers.com/forums/hardware-zone/28025-phenom-ii-x3-720-vs-core2-q9300.html

 

[knutjb]

This article highlights why Intel has the only high end processor. They pressed AMD out of the mainstream market, through illegal means, which is where the profits are made, starving AMD of vital capital. 1.4B well spent. He sells high end which, at this time, means Intel, duh. This was a superfluous article.

 

[arkhon]

Another major reason OEMs and system builders go for Intel is that not many people have even heard of AMD. I have never ever seen an AMD commercial on TV. Most of my classmates and teachers don't know who they are, but have heard of Intel. Many of them seem to think that Intel is the only company that makes CPUs. Intel just has much higher brand recognition among the general population.

 

[apache_lives]

Back in the Pentium D/AMD X2 era we still sold 80+% Intel based rigs simply because of the Intel Pentium name - no one cared about efficency, performance etc - they simply wanted Intel

Its going to take a lot to change that sadly, but now that they have there own chipset/platform its going to help bit by bit - AMD has always been f***** in the a** because of there platform - first VIA and there garbage, now nvidia (especially in laptops - HP etc) - AMD chipsets for AMD processors will make that change for the better.

[citation][nom]strikker1[/nom]Boy,this guy sure knows his AMD stuff, being all up to date with the K6 line of processors. Oh, and the rebates are sure nice that intel gives, just ask nVidia how great they are for the ION platform where just the atom cpu cost $45 BUT if you buy the 3 chipset bundle it only cost $25. The rebates were sure great when, i think it was for the first xbox, MS was going to use Athlons but at the very last minute went with Celerons, which were far, far inferior. I'm sure intel didn't do anything "illegal" there either. This stuff doesn't even mention all the reports of intel "bullying" customers into going with there products. Where there's smoke, there's fire and you can barely see a foot in front of your face with all the smoke from intels "practices" in how they push there product.I have never heard of processors coming out of their sockets before. I would rather risk the processor coming out than have the motherboard bend and crack. I'd say more than 50% of the reviews for heatsinks that I have read all have problems when installing heatsinks on intel motherboards with those pushpins, a good deal saying the motherboard has so much pressure that it bends. I can't recall problems with heatsinks for AMD since they started using sockets 939/754 forward.Its just like the guy said, all of it has nothing to do with performance. I'm sure he didn't push AMD even when there lowest performance 3800+ X2 was wiping the floor with Pentium dual core EE $1000 processors. Until C2D, AMD had been ahead since the Thunderbird cache came out on the first K7s. Anyone remember how long a run that was? If it wasn't for AMD, intel would still be selling pentium III's and they would probably cost you about $3000 for just the processor.[/citation]

Load of crap about the Intel stock HSF designs - there fine, its the amateurs having trouble with them - my shop front builds thousands of PC's using socket 775 stock HSF's and no a single one of our builds have ever had issues - atleast not untill the customer or an apparent "tech" tried to do something with it, and if you actually read the box it comes in it actually says "to be installed by a professional" etc - it says what it means

While there is nothing wrong with the AMD retention design, Intels design is more efficent for flow of air and size/volume of hsf etc - better design.

 

[ohim]

[citation][nom]blackened144[/nom]It makes absolutely no difference if the AMD CPUs were wiping the floor with Intel. He said he customers did not want AMD CPUs. So, using your logic, he should have only been trying to sell CPUs that his customers do not want?[/citation]
I don`t belive there is any company that will have customers that want only and only intel, maybe they should look at their priceing and from what i`ve seen on their site they look more like an Intel owned company ... i quote from one of the pudget technicians

"Daniel Brown (Customer Support - Lead Technician) Says:
Intel has stuck with the 'metal bracket locking the CPU down' type of CPU mounting for LGA 1366. This means it is practically impossible to have a CPU damaged during shipping. This is a big advantage over AMD."

can you even belive telling this bullshit to customers ?

"Daniel Brown (Customer Support - Lead Technician) Says:
I like the Intel LGA 775 (and now the new 1366) platforms because of the more-secure CPU mounting when compared with the current AMD CPU sockets. The AMD AM2 and AM2+ sockets still use copper pins that stick out of the CPU, and a lever that applies pressure to those pins to hold the CPU in. This is mechanically inferior to Intel's method which has a metal plate that is locked down over the top of the CPU to hold it in. This difference results in far less shipping damage to Intel CPUs relative to their AMD counterparts."

"Jon Bach (President) Says:
I am amazed with these processors. It isn't very often that a performance leap like Conroe happens, and even though that was years ago, AMD still hasn't caught up. Seems like a no brainer to me...sorry AMD!" seems even their president is a nobrainer in not understanding that the most cash comes from low and and mainstream not top of the line products.

"Jon Bach (President) Says:
Intel's 45nm chips are cheaper, faster, and cooler running. A no-brainer if you want a dual-core CPU."

http://www.pugetsystems.com/cat_info.php?cat=n&show=CPU

 

[apache_lives]

oh and i agree too that AMD processors are no more or less reliable then Intel processors atleast at stock speed, cant say the same for there platforms (havnt worked with enough modern ATi/AMD chipsets to comment, they do seem ok tho)

 

[jonbach]

Thanks for the comments, everyone! I knew there would be some strong opinions out there. Just a few points of clarification:

1) I never say that AMD isn't reliable. I say that we focus on reliable computers. This means that they have to survive shipping -- I'm getting at the HSF retention issues I mention later. Both Intel and AMD CPUs are amazingly reliable -- in 2008 we measured a 0.6% failure rate on Intel CPUs and a 3.8% failure rate on AMD CPUs. However, 3/4 of the AMD failures were really shipping damage. They would have a ~0.97% failure rate if we take shipping out of the equation.

2) With Intel and AMD stock heatsinks, we have never seen a problem with retention. The problem comes when you use large heatsinks, which many system builders do not do. Back in the day just before Core 2 Duo came out, we used a lot of Thermalright XP120 heatsinks, and they were a big problem on AMD CPUs. More recently, even lighter heatsinks like the Arctic Cooling Freezer 64 have led to the 3.8% failure rate I cite above. Both of those heatsinks are good examples because we used them both for Intel and AMD CPUs, but only had retention problems on the AMD side.

Keep those comments coming!

Jon Bach
President - Puget Systems
http://www.pugetsystems.com

 

[arkhon]

AMD processors are less reliable? That's news to me and my Athlon XP 2600+, still going strong, without a problem in seven years, despite me taking apart my computer to dust it out and reassembling it without any prior knowledge of how to do so. It's also been overclocked to roughly 2.5 GHz from 2.1 GHz, and I don't know much about overclocking, either. I'm honestly amazed how well it's held up against my inexperience.

 

[curnel_D]

He's obviously not talking about CPU failure that happens over time. He's talking about CPU failure that happens during or shortly after. From my experience, if CPU's are going to go bad within ten years, it will be in the first 10 weeks after being placed in a system.

And AMD procc unseating is a huge problem with shipping. He's not saying that the heatsinks just fall right off in shipping, he's saying that because of the pin design, which is extremly outdated at this point, AMD processors become unseated during rough handling. And any loss of connection means a computer that wont boot, and possible damage to the CPU and motherboard.

Every reason he listed about not wanting to use AMD CPU's is very valid. At this point in the game, the only reason an enthusiest computer manufaturer has a reason to build AMD powered computers is because of customer demand. And that's exactly the point...
There is no customer demand.

 

[arkhon]

[citation][nom]curnel_D[/nom]He's obviously not talking about CPU failure that happens over time. He's talking about CPU failure that happens during or shortly after. From my experience, if CPU's are going to go bad within ten years, it will be in the first 10 weeks after being placed in a system.And AMD procc unseating is a huge problem with shipping. He's not saying that the heatsinks just fall right off in shipping, he's saying that because of the pin design, which is extremly outdated at this point, AMD processors become unseated during rough handling. And any loss of connection means a computer that wont boot, and possible damage to the CPU and motherboard. Every reason he listed about not wanting to use AMD CPU's is very valid. At this point in the game, the only reason an enthusiest computer manufaturer has a reason to build AMD powered computers is because of customer demand. And that's exactly the point... There is no customer demand.[/citation]

OK. But wouldn't the heatsink hold the CPU down anyway?

 

[curnel_D]

[citation][nom]arkhon[/nom]OK. But wouldn't the heatsink hold the CPU down anyway?[/citation]
Nope, not always. The reason is because despite it being torqued down, in harsh shipping conditions, it still has some give. And despite how much or how little and what quality thermal paste you use, it will usually take the processor with it in that 'give'. When this happens, the processor becomes unseated in the socket. And because of the levered zero force insertion design, unless that lever is lifted so the pins can fall back into place, the CPU will stay unseated. While it's unseated because the pins cant fall back into place, the heatsink is still putting considerable stress on the CPU pins which could lead to damage. Then combine that with the DOA appearance, and you have a pretty high AMD cpu failure rate.

 

[curnel_D]

[citation][nom]arkhon[/nom]OK. But wouldn't the heatsink hold the CPU down anyway?[/citation]
And not sure why this guy was -1'd for asking a serious question. *grumbles and clicks the thumbs up*

 

[zolddude-tkc]

Jon Bach,

The only reason you have an issue with shipping is simply that you fail to box the units correctly.
Take the blaim and stop trying to pass the buck.

 

[arkhon]

[citation][nom]curnel_D[/nom]Nope, not always. The reason is because despite it being torqued down, in harsh shipping conditions, it still has some give. And despite how much or how little and what quality thermal paste you use, it will usually take the processor with it in that 'give'. When this happens, the processor becomes unseated in the socket. And because of the levered zero force insertion design, unless that lever is lifted so the pins can fall back into place, the CPU will stay unseated. While it's unseated because the pins cant fall back into place, the heatsink is still putting considerable stress on the CPU pins which could lead to damage. Then combine that with the DOA appearance, and you have a pretty high AMD cpu failure rate.[/citation]

Okay, I see. I don't have much experience actually building my own systems. The only experience I have is with my old Athlon XP computer, and that heatsink put up a serious fight when I tried to remove it and replace the thermal compound, so I have a hard time imagining it coming off or even loose.

 

[curnel_D]

[citation][nom]arkhon[/nom]Okay, I see. I don't have much experience actually building my own systems. The only experience I have is with my old Athlon XP computer, and that heatsink put up a serious fight when I tried to remove it and replace the thermal compound, so I have a hard time imagining it coming off or even loose.[/citation]
Just wait till the first time you see a fedex guy kicking a machine across a parking lot because his hands are full, and you'll understand right away. lol

 

[unlicensedhitman]

I noticed there are more custom cpu heatsinks and fans for Intel than AMD. AMD needs more custom cpu heatsinks and fans

 

[unlicensedhitman]

I live in San Jose, CA and Microcenter is selling Core i7 920 2.6ghz for $199. I didn't buy one because Intel X58 mobos are still pricey compared to AMD mobos. I managed to get a deal on newegg for a Phenom II X4 2.8ghz for $165 + a $100 SLI mobo.

 

[ohim]

[citation][nom]TheCapulet[/nom]What's your point? Every single bit said there is completely true. It really just sounds like to me you've got an AMD procc, wanted an intel procc, and plan on taking it out on everyone around you.[/citation]
Actualy i have AMD CPUs since 5x86 , and never got tempted to buy an intel even if i afford it, i realy enjoy my PII 940 , and i never give to agressive marketing stuff, and wonder how those computers are handeled that during shipment the CPU gets dislocated from there since i belive the shocks would have to be quite substantial to do that ... meaning other parts of the computer will have to suffer too not only the CPU. And with that harsh negative comments ofc you scare away potential customers of that brand and encouraging the monopoly of intel. Tbh i would like to see the look on your face when AMD(hope it never happens) will go bankrupt and Intel will sell you a shitty CPU for the price of the car you are driveing.

http://lh5.ggpht.com/hungryforads/RzGJFEfLDKI/AAAAAAAAAI8/ybMpwVGg1H8/s800/1989_tandy_pc_ad.jpg

 

[curnel_D]

[citation][nom]ohim[/nom]Actualy i have AMD CPUs since 5x86 , and never got tempted to buy an intel even if i afford it, i realy enjoy my PII 940 , and i never give to agressive marketing stuff, and wonder how those computers are handeled that during shipment the CPU gets dislocated from there since i belive the shocks would have to be quite substantial to do that ... meaning other parts of the computer will have to suffer too not only the CPU. And with that harsh negative comments ofc you scare away potential customers of that brand and encouraging the monopoly of intel. Tbh i would like to see the look on your face when AMD(hope it never happens) will go bankrupt and Intel will sell you a shitty CPU for the price of the car you are driveing.http://lh5.ggpht.com/hungryforads/ [...] _pc_ad.jpg[/citation]
Ok, so what you're telling me is that you're entire harsh opinion is based of preconceived notions about a processor line that you've never used. And furthermore, you assure others that what more experienced system builders say isn't true because you cant wrap your mind around it?

You'd be surprised by the thousands of support calls dell/hp/Gateway gets from simple unseating issues with ram and processors. This was a huge problem with the socket 478 design, which also used a zero force insertion design. And now it's continued to be a problem with AMD's line.

And as far as aggressive marketing, if you truly think that AMD processors are faster clock for clock over Intel, you're every bit as unrealistically fanboish as you sound. AMD doesn't compete on a superior technological level at this point, which is why they've gone to competing in a market price level instead.

 

[strikker1]

[citation][nom]ohim[/nom]I don`t belive there is any company that will have customers that want only and only intel, maybe they should look at their priceing and from what i`ve seen on their site they look more like an Intel owned company ... i quote from one of the pudget technicians"Daniel Brown (Customer Support - Lead Technician) Says:Intel has stuck with the 'metal bracket locking the CPU down' type of CPU mounting for LGA 1366. This means it is practically impossible to have a CPU damaged during shipping. This is a big advantage over AMD."can you even belive telling this bullshit to customers ?"Daniel Brown (Customer Support - Lead Technician) Says:I like the Intel LGA 775 (and now the new 1366) platforms because of the more-secure CPU mounting when compared with the current AMD CPU sockets. The AMD AM2 and AM2+ sockets still use copper pins that stick out of the CPU, and a lever that applies pressure to those pins to hold the CPU in. This is mechanically inferior to Intel's method which has a metal plate that is locked down over the top of the CPU to hold it in. This difference results in far less shipping damage to Intel CPUs relative to their AMD counterparts.""Jon Bach (President) Says:I am amazed with these processors. It isn't very often that a performance leap like Conroe happens, and even though that was years ago, AMD still hasn't caught up. Seems like a no brainer to me...sorry AMD!" seems even their president is a nobrainer in not understanding that the most cash comes from low and and mainstream not top of the line products."Jon Bach (President) Says:Intel's 45nm chips are cheaper, faster, and cooler running. A no-brainer if you want a dual-core CPU."http://www.pugetsystems.com/cat_in [...] n&show=CPU[/citation]
[citation][nom]apache_lives[/nom]Back in the Pentium D/AMD X2 era we still sold 80+% Intel based rigs simply because of the Intel Pentium name - no one cared about efficency, performance etc - they simply wanted IntelIts going to take a lot to change that sadly, but now that they have there own chipset/platform its going to help bit by bit - AMD has always been f***** in the a** because of there platform - first VIA and there garbage, now nvidia (especially in laptops - HP etc) - AMD chipsets for AMD processors will make that change for the better.
Load of crap about the Intel stock HSF designs - there fine, its the amateurs having trouble with them - my shop front builds thousands of PC's using socket 775 stock HSF's and no a single one of our builds have ever had issues - atleast not untill the customer or an apparent "tech" tried to do something with it, and if you actually read the box it comes in it actually says "to be installed by a professional" etc - it says what it meansWhile there is nothing wrong with the AMD retention design, Intels design is more efficent for flow of air and size/volume of hsf etc - better design.[/citation]
I was refering to the article, where the guy said they used large heatsinks for high end systems, which wouldn't be stock heatsinks. The reviews I have read about the intel heatsinks issues weren't done by amatuers, but by multiple review sites where its there profession to test new hardware. I have never tried to install a pushpin heatsink for an intel processor, the only time I have ever dealt with a socket 775 was repairing someones Dell, and it had a large aftermarket heatsink that was held on with a braket that screwed into a plastic base that was attached to the motherboard...no pushpins. I really don't like how the only support the pushpins have is the PCB, no backplate/faceplate to take the pressure off the PCB. Having a HUGE copper aftermarket cooler in a vertical tower hanging off the side of a motherboard, not a good idea without a backplate, which the heavier ones include with the heatsink.

 

[cangelini]

FWIW, I personally have encountered issues with system builders using oversized heatsinks on Intel Core i7 platforms. The Cooler Master V10 does NOT like to be shipped pre-installed. It's far too heavy to bend every which-way...