Intel bribing THG? Is it possible?

[kinney]

I about buying off journalism sites.
That is ridiculous and I've never supported that notion.

I think the objection by the majority of the community was the apparant intel bias (as we took it) with the oc'd P4EE, which itself was barely in the hands of reviewers, let alone consumers. And the 3.2 Ghz we accepted, grudgingly.

Though I feel THG might have a intel bias nonetheless. That is ok in my mind as its a private organization.
Removing those results was great to satisfy the far majority of readers across the 'net though.

I defend you guys as much as possible because I'm a longtime reader and defend THG in any skirmishes.

The debate could never end, people could now say, "well do they send you a ton of extra hardware and then if the review comes out in their favor do they continue to pamper you guys with goodies?"
Or something to that degree. It wouldnt be direct payoff but essentially ends up that way.

Either way, its a private review site and your opinions are valid.
Though when something glaringly odvious comes out such as the inclusion of the 3.4/3.6ghz P4EEs to the A64 launch article, the community who supports you notices.
Its exceptional that you do respond accordingly.

-----
eden is my intel/ati superboy

 

[saintbabs]

No-one here needs or wants free hardware. They would like, at some point, to get paid more, have a the lifer that every working stiff wants.

Free hardware and goodies? This is a publication. 22 people work in the organization. Some are in college, out of college, married, have kids, are having kids, etc. etc. Free hardware and goodies?

Do we have an Intel bias? What the hell is an Intel bias? What do we care? We don't get scoops, more ad money, favorable nation status at Craig Barrett's barbecues, or even a pat on the back.

Being favorable to Intel gets us bubkas. Intel actually gives out fewer CPUs than AMD ever does to review sites, and for evaluation. Each company has different controls in place, and Intel's are not as site-friendly as AMD's.

So, there is a whole bunch of fallacious stuff in the innuendo that makes me cringe.

Now, here is the really funny part of things, the enthusiast community, which used to be driven by overclocking is now crying foul when we have overclocked CPUs in the mix?

Seems completely out of whack?

It seems to me that there is a group of people out there who want to stifle all debate, and to protect their own prejudices.

In a free market, democratic society, there are people who want to support protectionism, and argue for a socialist ideal when it comes to hardware purchases: cheap, powerful CPUs for all!

I can feel an article coming on about Leon Trotsky and the hardware enthusiast movement.

And, while I am in a pissy mood, what is it with some enthusiasts setting up review sites: instead of dreaming of being Michael Dell they want to be a Tom or Anand? No-one wants to go out on a limb and make it happen, they just want to comment on it?

This is where the whole enthusiast community thing wears me down. It doesn't seem right to have people who claim to be into hardware wanting to talk about it, but they don't seem to be able to do anything about making it happen for AMD. They can't seem to create the shift in the market that would make AMD go to the next level.

Where is the Michael Dell for AMD? Is he too busy trying to score a free mobo to review?

Omid Rahmat
GM & Publisher
Tom's Guides Publishing LLC
www.tomshardware.com

 

[kinney]

Well put.

Could you explain why the 3.4/3.6 were not colored seperately to indicate they were overclocked cpus as you have done in the past?
Example being the <A HREF="http://www.tomshardware.com/cpu/20030210/barton-14.html" target="_new">Barton's here</A> article.

I'm not attacking, I'm just curious if your stance is that it was a complete mishap or intentional in thinking that it would not matter?
The explanation they were even in the article was explained by THG with the A64 launch article update. But I think with the simple color coding there would have been much less flak. I would have been on the other side of the fence in that case.

Since you are posting on the subject, and this is related we would be interested in knowing.

BTW, the part about AMDs salvation being in the hands of its supporters is so true. I personally having an associate degree in LAN technology and being only a student now getting my bachelors, I feel the only way I can help is encourage users to purchase AMD products. But I also, believe in them.
No kinneytech in my near future.
Though being in college the bonus hardware would be great. :tongue:

edit- I've been reading your site since at least '99 IIRC and have bought much of my hardware based off your recommendations. Not all of us are noobs here to THG and in a way, expect alot out of you guys.
I'm surprised you decided to post on this topic. We appreciate your honesty and involvement with issues.

-----
eden is my intel/ati superboy<P ID="edit"><FONT SIZE=-1><EM>Edited by kinney on 09/25/03 05:55 PM.</EM></FONT></P>

 

[saintbabs]

You get processors a few days before the article has to go live. You have to run hundreds of tests over a week-end, a week-end you don't get paid extra for. You have to get everything into data charts and ready and up for a ridiculous embargo because, everyone wants you to post it at 12.01 pm EST, and not a minute later.

[-peep-] happens.

I can assure you that no-one in the editorial department cares that much about Intel or AMD that they would want to make a case for either one other than through tests and benchmarks and analyses of those benchmarks.


[-peep-] happens.

And on a final note: there isn't anyone at either Intel or AMD marketing who is clever enough to deliver a conspiracy. Both companies are run by committee, and they just don't have it in them. If they did, I'd find it very exciting and it would be a lot more fun dealing with them, but it is usually a bunch of boring white papers, pleading, press releases, to and fro, and just plain gossip as each company tries to position itself with editors.

In the real world, all anyone wants is to get the job done, collect their pay packet and get on with their lives. Intel used to be much more aggressive in Andy Grove's day, but not anymore. AMD has issues with their marketing that requires management addressing the structure of the company.

I wish it was more like a good ol' daytime soap, but sometimes it doesn't get much better than a day with greasy dishes.

Omid Rahmat
GM & Publisher
Tom's Guides Publishing LLC
www.tomshardware.com

 

[Crashman]

I hope you're arguing in favor of what I actually said, instead of simply arguing against the topic title. I used reverse psychology to suck in the protestors.

My point was that Intel has too sketchy a history with THG to win favors. So that if any article says an Intel processor won any comparison, there can be little doubt that it did.

Your point about corporate scandal is even stronger proof to debunk this thread’s title, which was the whole point of the thread from its creation.

By the way, I’m unemployed, you should hire me!


<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>

 

[spud]

Hey I am employed and you could hire me to. Dude I can buy a nice bat and go smacking fanboys up the head. Trust me less problems for you in the end.

-Jeremy

<A HREF="http://service.futuremark.com/compare?2k1=7013108" target="_new">Busting Sh@t Up!!!</A>
<A HREF="http://service.futuremark.com/compare?2k3=1311896" target="_new">Busting More Sh@t Up!!!</A>

 

[MajinChewY]

You people dont realise that most of the sites where the FX-51 owned the p4ee were the sites that used standard windows xp to run the 32 bit software used in benchmarks. Tom used 64bit windows xp, their is an overhead when running 32 bit software on amd-64 in winxp64, this degrades performance. This was also not made clear on tom's site. Their was absolutly no reason that the 64 bit OS was needed either, as he chose to not include the 64 bit programs AMD sent him in his benchmarks. He also chose not to use linux, because basicly its not a mainstream OS, and not "choosing" to use linux means he had the option of picking the OS, and he chose WinXP-64 over WinXP, knowing this would give the pentiums an advantage. God forbid giving AMD an advantage by using 64 bit software though.

 

[NMDante]

Did you read the review, at all???

Look at the Test Setup page, under SOFTWARE. Where does it say, WINXP 64? Nowhere?!? Geez.

Here's link directly to the test setup: <A HREF="http://www17.tomshardware.com/cpu/20030923/athlon_64-20.html" target="_new">LINKIE</A>

Read the <i>whole</i> review before you start accusing...



How many watts does it take to get the center of CPU core?

 

[infiltrator]

Well, there's another story, the 2 versions for XP are different. MS have only relesed their BETA 2 for AMD, and Intel one has been around 4 a bit longer. Both these Chips impliment 64bit in different ways, so Good Luck for the Benchmark people in representing a fair rating system 😀

 

[kinney]

You get processors a few days before the article has to go live. You have to run hundreds of tests over a week-end, a week-end you don't get paid extra for. You have to get everything into data charts and ready and up for a ridiculous embargo because, everyone wants you to post it at 12.01 pm EST, and not a minute later.

It only leaves me wondering... if time is one of the main issues, someone mustve had some extra time or they wouldnt have stuck two (3.4ghz and the 3.6ghz) extra complete results for every benchmark.

But I accept your answer nonetheless.

-----
eden is my intel/ati superboy

 

[Ncogneto]

Being favorable to Intel gets us bubkas. Intel actually gives out fewer CPUs than AMD ever does to review sites, and for evaluation. Each company has different controls in place, and Intel's are not as site-friendly as AMD's.

I think the emphasis should be given to the marketing dollar and the money each company may pay a website for advertiseing on your website. Are you trying to tell us that reviews with a less than glowing report of Intel (or AMD for that matter) will not negatively effect ad income?

Now, here is the really funny part of things, the enthusiast community, which used to be driven by overclocking is now crying foul when we have overclocked CPUs in the mix?

The foul occurs when you are supposed to be doing a review of released products. If you wish to show the overclocking potential..great...do a seperate review. Or at least have the courtesy of overclocking the competitors CPU as well.


It's not what they tell you, its what they don't tell you!

 

[Ncogneto]

And on a final note: there isn't anyone at either Intel or AMD marketing who is clever enough to deliver a conspiracy.

Maybe, maybe not. The Intel/Rambus fiasco certainly made more than a few people raise thier eyebrows. Then there was the odd investment by intel in a certain website (zdnet if I recall).

[-peep-] happens.

This is why you have editors.

Difficult as it may be if you want to give the viewer the perception of non-bias the best way would be to set up a site modeled after a publication such as consumer reports. Accept no advertising from companies whose products you frequently review.


It's not what they tell you, its what they don't tell you!

 

[slvr_phoenix]

This is why you have editors.

This may be the first place that I really agree with you in this thread. The one thing that I've always knocked THG on is their editing quality. There seem to be more typos, chart coloring mistakes, mistakes in the sorting of the results, etc. at THG than at any other place. It's kind of sad since (IMO) THG is better than 99% of the other review places when it comes to everything else.

But then again, with common sense enabled it's really not a problem anyway. It's just frustrating is all.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>

 

[slvr_phoenix]

man you sure make a big deal out of ECC "not being designed for desktops" or something. Now, first please tell me; according to you, what exactly is the difference between a workstation, and a (high end) desktop ?

The motherboard, the RAM, and the processor. A high end desktop will use a desktop board, standard RAM (often at the highest possible speed and timings), and a desktop processor. A workstation will use a workstation/server board, registered/ECC RAM and/or slower (more stable) RAM ... and usually a lot more RAM at that, and will often use a workstation/server CPU and/or a dualie configuration instead of a single CPU. Which is exactly why P4EE is desktop and A64FX isn't. Even if you don't make it <i>just</i> about the CPU there is still the entire platform that is <i>required</i> for the CPU as well.

Guess what, workstation users tend to appreciate ECC.

I never said that they didn't. I just said that A64FX is server/workstation, not desktop. And as you yourself stated, <i>desktop</i> users don't give a squat about registered/ECC.

If I run a huge simulation or render job on my Ultrasparc that takes hours, if not days, I am damn happy a bit flip in my 6 GB ram doesnt crash my puter or corrupt the result. I don't mind a few percent less performance (otherwise, I wouldnt be using SUN in the first place I guess). Trust me, I am REALLY trying to convince my employer to get me BOXX dual opteron machine, as the software I use is being ported to AMD64 under linux. If that doesnt work, trust me, a single CPU FX would make a great workstation for me too, as long as it takes (enough) ECC Ram.

I don't disagree in the slightest. That's my entire point. The Opteron (or if you just rename the exact same CPU then the A64FX) is a <i>wonderful</i> server/workstation CPU. However because of it's required platform it is <i>not</i> a desktop solution. Yet. Once AMD comes out with Socket939 then it will be. <i>Then</i> it'll finally make sense and all will be right in the world.

Now, gamers is another thing. I agree that ECC/registered ram isnt a requirement for them. tough luck. Either you pay up for the more expensive RAM, or you don't. Same thing applied to the P4 initially, either you where willing to pay a HUGE price premium for couple megabytes of RDRAM, or you werent, and you picked something else. Its a cost consideration,nothing more, nothing less. Big deal.

This is the point where you're totally wrong. It's not just about more money for RAM that's more stable. It's about an entire platform that will <i>never</i> let the user upgrade with a desktop CPU. It's about an entire platform that will <i>never</i> let the user trick out their system with higher-performing (but less-stable) RAM. It's not <i>just</i> the money. It's the money <i>and</i> the forced platform. If Socket939 chips were compatible with Socket940 mobos it'd be an entirely different story. If Socket940 mobos had the option of running normal RAM just as well as registered/ECC RAM then it'd be an entirely different story.

Some gamers <i>do</i> choose to run a workstation PC instead of a desktop PC. For them A64-FX is great. However the point still remains that this is a workstation, not a desktop. It's not about gamers and non-gamers. It's about markets: Desktop, Worstation, Server.

BTW, if will be solved in a few months when socket 939 is launched.

I completely agree. <i>Then</i> a Socket939 A64FX <i>will</i> be a desktop. (At least according to available information.) Right now though it isn't. It's just a renamed Opteron, and <i>that</i> isn't desktop. It's still a great CPU, but it's not the same market as a P4 or the standard A64.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>

 

[slvr_phoenix]

You missed the point. AMD doesnt have all the time and resources in the world.
If you'd reread my quote from the thread above you might understand WHY they did it that way.
It was actually a crafty move.

Oh I completely understand <i>why</i> AMD did it. And you're right, it <i>was</i> a very cunning move by AMD. Somehow AMD marketing has managed to convince everyone that a renamed Opteron running on an Opteron motherboard and using workstation/server class RAM belongs in a desktop market. Somehow AMD has managed to convince countless people to disable their common sense. All with just a simple name change.

Imagine what Intel could do if they sold Itaniums under the Celeron name. It'd revolutionize the desktop industry!

Find me where you pulled this definition of a desktop PC using exclusively nonECC ram and ANY one kind of socket or slot format.

Find me even two major OEMs that sell a PC with registered/ECC RAM and an uses a Xeon, Itanium, AthlonMP, or Opteron processor as a <i>desktop</i> box.

The FX appears to have become a "desktop" solution to me! As well as every reviewer and person in the world as of Sept. 23rd.

Amazing, isn't it? Had Intel done the <i>exact</i> same thing the world would have cried foul. Hell, look at how many people already cry foul over the P4EE and <i>that</i> at least runs on a desktop motherboard with ordinary RAM just like every other P4. It's truly fascinating just how far people are willing to bend their common sense just because AMD is the underdog.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>

 

[slvr_phoenix]

Fact of the matter is workstation class computers and desktop class computers distinctions are getting less and less defined. I care not to argue this point with you as it will ultimaly come down to a matter of opinion and personal intrepetation. If AMD (or Intel with the P4EE for that matter) want to market there respective CPU's as desktops processors, who am I (or you) to argue.

If it were <i>just</i> the processor involved I wouldn't even argue. If one could put any/all of the reviewed A64FXs into a socket939 mobo with ultra-low-latency unregistered DDR400 RAM right this moment I wouldn't have a single complaint about this.

Personally, yes I think so. No need to start a flame war here, until a signiicant period of time passes this will be pure speculation on either of our parts. Bear in mind at least three RAM manufactoreres released registered memory to co-ordinate with the launch of the A64. Am I saying it will come down to the price point of non-ECC Ram, no.

And that's the beauty of debate. We can agree to disagree. I'm perfectly fine with that.

Actually I am not quite sure what you mean by this. Are you saying you don't know anybody (your company withits servers included) that uses Registered ECC memory? I really find this hard to believe. I know my Fc-Raid card uses it for its cache memory.

I'm saying that we don't sell it to any of our customers. In theory we might have shipped a couple of SCSI adaptors with ECC because we use a lot of SCSI technology, but we've never shipped registered/ECC as main memory. There's never been a need to. There's no demand for it. Registered/ECC RAM is an extremely niche product.

So, in larger RAM configurations it becomes increasingly possible to have enough of a significant amount of Error occuring in RAM to start to become a concern. With the new 64 bit processors being able to address much more RAM than thee 32 bit counterparts this becomes a concern

It's amazing how you can chew out marketing lies. The amount of RAM itself is in no way related to the number of errors. It's the integrity of the data paths, the number of DIMMs (since the further the slot is, the more error prone the signal gets) and to a minor extent the infrequency of access to a specific location of RAM that generate errors. The actual amount of RAM is inconsequential.

Further Registered/ECC RAM isn't purchased because of the amount of RAM. It's purchased because the impact of <i>any</i> error is unacceptable. Which is why it's meaningless to desktop systems which can be rebooted with only a grumble or curse, but is vital to servers where dollars and customer's confidences are lost for any amount of downtime, and it's vital for workstations where any lost hours of crunching data means lost productivity.

Stick to the facts. I'd like to say that marketing tricks don't work here, but considering the number of people defending a Socket940 A64FX system as a desktop system, obviously marketing tricks <i>do</i> work here ... on some.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>

 

[slvr_phoenix]

Omid, you like a good rant. What's your take on this? Has AMD marketing managed to convince you that an Opteron by any other name, even if it's on the exact same platform, is somehow no longer a workstation/server CPU? Is the Socket940 A64FX with it's multiple hypertransport layers for multiple-CPU support and registerd RAM <i>requirement</i> just another desktop while the Socket940 Opteron on the exact same mobo with the exact same RAM isn't?

I mean honestly, this isn't about the 64-bit war. This isn't about AMD vs. Intel. It's about the consumer and what a company should and shouldn't be allowed to do to them.

Yes? No? Can we complain about Intel releasing the P4EE for desktops and in the same breath praise AMD for releasing the Socket940 A64FX for 'desktops'? What do you think? You don't have to represent THG officially. I just want to hear what <i>you</i> think.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>

 

[sjonnie]

doesn't require ECC RAM to run

Lets clear this error up, the AMD64 FX-51 doesn't require ECC memory either. It requires registered (or buffered) RAM which may or may not have ECC which you may or may not be able to disable if it does. There is a very good reason for having a register when you have large amounts of RAM (1Gb or more) or if you have >2 DIMMS. The register directly consumes 1 clock cycle but at the same time improve data handling. In the real world I can guarantee you will not see a difference in performance (only in some poxy meaningless benchmark - oh no, only 360FPS in Quake III and not 370, god damm my performance sucks! etc. etc.).

Incidentally, Tom was wrong to say registered DDR333 and 400 is not available.

Currently there is no memory for DDR400 (registered version) or DDR333 on the market.

I have registered DDR333 from Transcend (also been available from better known Mushkin and Kingston for at least 6 months). Transcend also sell DDR400 registered RAM.

<A HREF="http://www.anandtech.com/myanandtech.html?member=114979" target="_new">My PCs</A> 😎

 

[Ncogneto]

It's amazing how you can chew out marketing lies. The amount of RAM itself is in no way related to the number of errors.

LMAO...if you say so. You need to do some homework. While the amount of RAM is only one aspect of the ratio it does play a factor. Easily put, the more bits you have, the more probability of one of those bits changing its state you have. It is simple mathematics. Just like the more drives you have in a single raid 0 array, the better your chances are for a failure. Certainly things like the quality of the drives also play an important factor, but that does not discredit the size as playing a role.

Further Registered/ECC RAM isn't purchased because of the amount of RAM. It's purchased because the impact of any error is unacceptable. Which is why it's meaningless to desktop systems which can be rebooted with only a grumble or curse, but is vital to servers where dollars and customer's confidences are lost for any amount of downtime, and it's vital for workstations where any lost hours of crunching data means lost productivity.

HUH? Well you heard it here first folks. Having your at home desktop computer/workstation computer crash is perfectly tolerable as it only will require a simple reboot.



It's not what they tell you, its what they don't tell you!

 

[Ncogneto]

<font color=red>If the probability of an error is one in a hundred quadrillion, and if the memory system is running at 10 MHz (100 nanosecond), and if you have 125 Megabytes of RAM (1 billion bits), then you would expect on average to see one single-bit error every ten seconds and one double-bit error every thousand quadrillion seconds (somewhat more than the age of the universe). That is why ECC memory is worth using, and why it is designed to detect but not correct double-bit errors. </font color=red>

Taken from this link http://oak.cats.ohiou.edu/~piccard/mis300/eccram.htm

Now raise the amount of RAM and you can see that yes unlike what you wish to believe SIZE DOES MATTER.

edit: another link for your homework pertaining to the notion you present us that amount of RAM has no effect http://www.atpusa.com/pdf/FAQfinal.pdf

Pay particular note to where they say amount of RAM and RAM error scales linearly.

It's not what they tell you, its what they don't tell you!<P ID="edit"><FONT SIZE=-1><EM>Edited by ncogneto on 09/26/03 04:26 PM.</EM></FONT></P>

 

[kinney]

I suppose its came down to matter of opinion.

I feel the difference between a desktop versus workstation computer is how you use it.
Not so much the hardware.
Because then you could draw innumerable lines and make arguements such as yours up.
Which yours is valid, because its opinion and cannot be proved one way or the other.

You attack AMD for doing this horrid relabeling of the Opteron for the FX (for now at least).
Yet the 2 reasons you said that makes the P4EE not a Xeon MP just doesnt hold enough water.

I dont think I have any problem calling my Athlon XP system a workstation because it is my workstation.
You are confusing marketing with hardware. They have an image to present with each line. They want the Opteron to appear more professional so its taken seriously, not some kid with $80 buying one so he can play Counter Strike.

I'm sure theres been crossdevelopment between all processors, P4/Celeron, Athlon/Duron, P3/P4, Itanium/Xeon, Xeon/P4 (examples, not necessarily true).. so to draw hardware lines on server class/desktop just doesn't make sense to me furthermore.

But again it comes down to opinion..
its just some ppl's stink.
I would be happy to start a poll on this topic.

Though I'm afraid people who are reading this thread will come down to people who prefer intel will choose your idea and those who prefer AMD, mine.
Its odvious we both have our own sides divided on the issue.

I think though in the end more would agree with my "definition" of a workstation from simply common sense, that its marketing and how you use it.
Go vote on my poll.

-----
eden is my intel/ati superboy

 

[slvr_phoenix]

Easily put, the more bits you have, the more probability of one of those bits changing its state you have.

Close. Easily put the more bits that are accessed the more probability that one of those bits changing its state. It's about heavy memory usage against light memory usage, not more memory against less memory.

Just like the more drives you have in a single raid 0 array, the better your chances are for a failure.

Funny how your example is based on the number of devices, not the amount of storage on each device. Which I agree with. If you have more sticks of RAM it increases your chances of an error. That does not necessarily coincide with more RAM however. I could get a 1GB DIMM or I could get four 256MB DIMMs. Same amount of storage space, very different number of devices.

but that does not discredit the size as playing a role.

Size itself does not play a role. Size just happens to often be related to the other factors that <i>do</i> play a role.

HUH? Well you heard it here first folks. Having your at home desktop computer/workstation computer crash is perfectly tolerable as it only will require a simple reboot.

I never said that it was enjoyable, just that it isn't vital. It <i>is</i> tolerable. If it wasn't then people wouldn't even be running desktop PCs in the first place. People would always use RAID1 or RAID5 arrays. People would always have redundant power supplies and plug their PC into a UPS with AVR. Why don't people spend money on these kinds of things? Because it <i>is</i> tolerable. It's not enjoyable, but it <i>is</i> tolerable. And if you heard it here first than you must be living in a closet.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>

 

[slvr_phoenix]

Yet the 2 reasons you said that makes the P4EE not a Xeon MP just doesnt hold enough water.

Why? Beause the P4EE has a <i>desktop</i> pin configuration, runs in a <i>desktop</i> motherboard, uses <i>desktop</i> RAM, and <i>won't</i> fit into a Xeon motherboard? If that doesn't hold enough water to differentiate it from a Xeon then what does?

You are confusing marketing with hardware. They have an image to present with each line. They want the Opteron to appear more professional so its taken seriously, not some kid with $80 buying one so he can play Counter Strike.

<i>I</i> am confusing marketing with hardware? I seem to be the only one that knows where the line between desktop and workstation hardware is because I refuse to let <i>marketing</i> decide for me. Do you even realize how flawed your statement was?

And even <i>if</i> a Socket940 A64FX cost $80 there would <i>still</i> need to be an entire workstation platform to run it on, unlike any other <i>desktop</i> processor. And let's face the truth, it's certainly more expensive than $80. It has nothing to do with appearing professional. It has everything to do with not having a bad A64 launch. Beleive anything else if you like.

I feel the difference between a desktop versus workstation computer is how you use it.
Not so much the hardware.

If that opinion works for you, then that's nice. The rest of the world may not agree, but that's your choice.

I dont think I have any problem calling my Athlon XP system a workstation because it is my workstation.

You can call it anything you want, but that doesn't change what it <i>actually is</i>. Just like AMD can call it an A64FX all they want, but it's still just an Opteron. And until the Socket939 version comes out, that's all it ever will be.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>

 

[Kemche]

Sorry, Too thread is too big to read all the replies. Anyways, I agree with you Crashman.

I don't remember when but in previous review of one of the AMD's processor on THG they made a comment about how AMD wanted them to install one of the patch with some optimization. And everyone installed that patch (which wasn't available to consumers) but THG refused to install it. I can bet that's probably what's going on in these reviews. Everyone is following AMD's specification on how to setup the BIOS and what timings to use etc. And THG is doing what consumers would be doing.

To make another point. I don't know why people are arguing about P4EE is a paper launch. Who said P4EE is already launched. I don't think Intel ever said they are laundhing it. They did a demo at IDF and they said it will be available soon. They didn't pull the P3 1.13 trick when they released it and recalled it. They didn't even pull AMD's trick and launched the processor and it wasn't even available. I do think that Intel probably got afraid that Everyone will announce AMD as a winner and launched this processor.

Also, Fx51 the 940 pin version will not even be compatible with Fx51 939. So why isn't everyone mad at AMD. I used to see lots and lots of post not to buy P4 willy when it was released since Woody was just around the corner.

What I think is that AMD made a big mistake by integrating the Memory controller on the Chip. Now every new memory technology upgrade you will have to buy a new processor with the new memory. For example, You will soon start to see DDR II memory in the market. Intel will be supporting it sometimes next year. When AMD decides to support it you will end up buying the board, memory and the Processor. I don't know how that could be good for users.

And that's my rant...

KG

"Artificial intelligence is no match for natural stupidity." - Sarah Chambers

 

[Ncogneto]

Funny how your example is based on the number of devices, not the amount of storage on each device. Which I agree with. If you have more sticks of RAM it increases your chances of an error. That does not necessarily coincide with more RAM however. I could get a 1GB DIMM or I could get four 256MB DIMMs. Same amount of storage space, very different number of devices.

Now you are clouding the issue and backtracking from your original argument. And yes, more sticks of RAM is an issue as well. But still, that does not change the fact that larger RAM modules play a role as well. You have done nothing to discredit that and I have provided 2 solid links supporting my statement. And try to buy 1 gigabyte unbuffered DDR. Now how could AMD claim to support over 4 gigs of RAM if ECC wasnt used?

Size itself does not play a role. Size just happens to often be related to the other factors that do play a role.

Wrong again. However both factors do play a role. But if you are trying to say that two sticks of 512 meg unbufferred are more likely to produce errors than 1 stick of 1 gig this I can agree with.

I never said that it was enjoyable, just that it isn't vital. It is tolerable. If it wasn't then people wouldn't even be running desktop PCs in the first place. People would always use RAID1 or RAID5 arrays. People would always have redundant power supplies and plug their PC into a UPS with AVR. Why don't people spend money on these kinds of things? Because it is tolerable. It's not enjoyable, but it is tolerable. And if you heard it here first than you must be living in a closet.

Who are you to say what is tolerable and what isn't? Buy your definition then, anyone using there home PC ( which certainly must be categorized as a desktop)should never be working on something in which losing there data isn't tolerable. Your talking yourself into a corner here. Heaven forbid that anyone buy a computer that is capable of doing both work and play, this just isn't bearable. One must either have a desktop pc or a workstation/server. Certainly not one capable of both.


It's not what they tell you, its what they don't tell you!