Setting the AMD crowd straight - Page 2

I have not had any personal experience with the stability issue you talk of but the other pionts are all valid, of course there is advantages to buying an intel CPU the same as there is for buying an AMD CPU, you have to look at the issues that will have the biggest effect on you and your system....most of my fast systems are AMD based but my servers that require that bit more are all on multi CPU which is required to be Intel......Everthing or CPU has it's place even the cyrix cpu's can do good in the world...

so everybody stop ya bitchin and accept we have a good unch of CPU's to choose from for whatever you want to tdo witht them...

M

one of the first UK T-Bird users....

Seems like a pretty simple solution. And honestly, Intel chips can be "blown" by overclocking as well. I agree that there is a certain problem with ordering a pre-built system with an AMD inside because of the HSF/delicate-die issue. The possibility that your processor could be destoryed en route is unreasonable. However, for those of us who build our own systems, AMD is an excellent choice (for now at least).

Charles

I'm just trying to up my post count and I can get away with doing it nonsensically here

We aknowledge them, we just disregard them after looking at the performance (especially on the P4) and the money (super especially on the P4).

Some very good points indeed. Although, AMD will have SMP suport in like 2 months. Thermal protection isn't a big deal either.

Actually I believe it was on the market. I remember looking around on sites (pricewatch? I forget...) and it was there. That's why everyone calls it a recall, cause that's what it was. That's why some people still have the 1.13GHz P3s. (not too many i can assure you.)

They're not good just because they work. My old Cyrix piece of crap works but it's not good. My old 286 (if I wouldn't have smashed it) works but it's not good. Hell it's only 16bit and 12MHz!!!!!!

They do. It's called the 1.2 and 1.3 GHz TBird. (soon to be Palomino).

Bollocks, the mobo chipsets maybe questionable, but the awards are real and refer to the CPU itself, which is superior to Intel P2 and P3. Its FPU is so powerful it can render images in Photoshop in half the time of an Intel at the same speed.
Please stop holding on to the past....ditch the i love the old Pentium Pro P6 core, its old and over priced. most apps are faster with Athlon.

"We aknowledge them, we just disregard them after looking at the performance (especially on the P4) and the money (super especially on the P4)."

Well, that's your loss. Intel makes a lot more than just a P4 chip. And I'm sure that future chips based on the present P4 are going to have a better FPU in them as well. It's only the current P4 that's such a weird animal.

"Actually I believe it was on the market. I remember looking around on sites (pricewatch? I forget...) and it was there. That's why everyone calls it a recall, cause that's what it was. That's why some people still have the 1.13GHz P3s. (not too many i can assure you.)"

Actually, being listed on Pricewatch doesn't mean that the hardware has hit the market yet. A LOT of people will sell a product before it's even available based on the fact that it SHOULD be available in a month or two. Which is what happened with the P3 1.13GHz. It was SUPPOSED to be available. Companies started allowing people to pre-order it. Then before it actually came to be time to sell, Intel said, "Sorry, we made a mistake.",and everyone who pre-ordered now had nothing to recieve. The fault there really lies with the companies who were selling the hardware before they actually had it in their hands. And anyone who actually did get a hold of a 1.13GHz chip got a pre-production engineering sample that wasn't meant for retail sale.
At least this is how I saw things happen. There may have been a rare case of an OEM somehow getting a hold of a shipment of these bad chips and selling them. I haven't heard of any such case though. To my knowledge, the retail version of the P3 1.13GHz never actually made it to market. And if I'm wrong, I'd love to hear about it.

"They're not good just because they work. My old Cyrix piece of crap works but it's not good. My old 286 (if I wouldn't have smashed it) works but it's not good. Hell it's only 16bit and 12MHz!!!!!!"

You completely misunderstood what I was saying. I didn't mean that it is a good chip by performance standards. I meant that it was a good chip in that it hasn't been recalled and if you own one, you at least know it works. It may not work as well as you would think it should, but it works. It isn't a damaged product.

"They do. It's called the 1.2 and 1.3 GHz TBird. (soon to be Palomino)."

Yes, but you completely missed my point about running optimized software. When a T-Bird running software optimized for AMD chips is compared to a P4 1.5GHz running software optimized for a P4 and SSE2, the P4 will run faster.

But I agree that when running today's software without optimizations for either chip, the T-Birds will out perform the P4s. I don't even dare to suggest otherwise there, because we all know that the P4 has only half of it's intended FPU, and even if it had a full FPU, AMD's FPUs are better than Intel's FPUs.

Don't anyone get me wrong. AMD can kick some butt. They make high performance chips with great FPUs. And the double-pumped bus was an interesting idea to put into a chip.

BUT, there ARE cases where Intel products meet the demands of a customer that AMD products cannot. AMD is working on fixing this, but they haven't fixed all of these cases yet. So there are consumers who may have a very good reason for going with Intel over AMD. But for the general home use the AMD system is the way to go.

- Sanity is purely based on point-of-view.

It is you my friend that lives in ignorance. You are now even contradicting yourself. In one post you flame AMD for leaving something out of there chips ( thermal protection) still you go on to agree that you wish intell would have not crippled the fpu of there p4. So, you have admitted that intell has also left something out of there chip so by your own definition they (Intell) are as bad as AMD.

A little bit of knowledge is a dangerous thing!

Hey griz, how do you do the in reply to thing?

- "That's no moon, it's Rambus!"

It's unfortunate that people like yourself have been brainwashed into believing that AMD can't do, especially in the corporate desktop space, the same job as Intel. How can you claim that Intel is stable with all the engineering fiascos they've had over the past year? Fiascos that were caused by the competitive market brought about by the AMD Athlon's presence.

AMD chips are no more a hobbiest or home CPU than Intel chips are. I run a network with 65 workstations, 50 of them running Athlons. While 50 desktops are a farcry from the 3000 in your company, my experience with 50 Athlons has been nothing but positive. Many of these machines are running 24/7, in ofices that very warm in the winter and downright hot in the summer. I spend no more time fixing or patching Athlon boxes than I did when most of my desktops were Intel based.

With an open mind, you would probably achieve similar results......

ever heard of the term "company expense". who needs to pay for stuff when your company does it for u.


--SR

I'm going to bring this thread to life.

Solution to AMD problems:
1)spend $15+ on a good fan/cooler
2)download new drivers
---total time spent----
5 minutes

Solution to Pentium problems
1)lower price by 40%
2)fix the crap called the P-IV
---total time spent---
1 year probably before they figure it out....

<<<Solution to Pentium problems
1)lower price by 40%>>>
yeah that would make intel solutions much cheeper considering the expense of a good amd board.

<<<2)fix the crap called the P-IV>>>
how do you figure??? a company designing a chip to push the software industry to create some innovative [-peep-] is bad. you are defenitly a anal receptive cum chugging amd zealot.

Solution to AMD problems:
1)spend $15+ on a good fan/cooler (make sure it will also fit a socket 370) ($15)
2)download new drivers (5mins)
3)swear at the computer (5hrs)
4)buy a cordless drill ($150)
5)buy a 1/4" cobalt drill bit ($3)
6)drill a hole in the VIA northbridge (1min)
7)drill a hole in the VIA sorthbridge (1min)
8)put your TBird chip on the shelf (1min)
9)remove your m/b and place it in the trash (15 mins)
10)spend $300 on a PIII and a 815ep m/b ($300)
11)install new m/b (15mins)
12)install PIII and HS/fan (15mins)

---total $$$$ spent----
$468

---total time spent----
5hrs 53minutes

*Notes:
- you may already own a cordless drill in which case the total would only be $318
- be very careful not to let any metalic fragments drop on your expansion cards while drilling.
- this is only a guide, actual results may differ

But seriously now...
I'm not really anti-AMD, but in my opinion the TBird systems need a little more time to mature before I'd buy another one. Maybe they've solved all the "issues" since I bought my AMD system, but it was HELLISH at first. For those of us who bought the earlier revision boards and had to use the earlier drivers it was painful. If it was a product that could of killed you there would have been thousands of deaths.

Anyway, I hope AMD continues to make great products because it only helps push the technology higher and the prices lower, but until I stop hearing the horror stories it's Intel for me.

- JW

If these prices are really worth it to you, then go burn money:

P4 1.3 -
$489.95 +15.40 shipping
(cheapest one w/ a 20 day warranty otherwise add about $80)

T-Bird 1.2 w/ 266 FSB eta 2/15 -
$294 + $9 shipping

Both need a heatsink/fan/cooler...

It seems to me pretty obvious that price/performance gives the AMD super high marks vs the P4. So unless you have a pile of greenstuff waiting to be eaten or burned or whatever you do with your money, I'd buy the AMD...at least right now.

Wow, this thread lives!
JGarrett, not sure what benchmark you were using but...
AMD memory bandwidth is half that of a P4, known factoid
P4 moves 3.2Gb/s on the bus.

Fugger Wrote:

Wow, this thread lives!
JGarrett, not sure what benchmark you were using but...
AMD memory bandwidth is half that of a P4, known factoid
P4 moves 3.2Gb/s on the bus.

Here is a cut and paste of exactly what I said. It seems correct to me. Can you tell me where I mis-spoke?

3) In every benchmark I have run, my system smokes the 1.5GHz Pentium 4 except in the memory bandwidth benchmark. In all CPU tests, it killed everything else.



John Garrett
System Admin - www.elementk.com
Editor - Exploring Windows NT Professional

"Well first of all I can spell the word dumb."

Congrats. You're just so special. Like you've NEVER made ANY typos or gramatical errors in ANY of your posts. If you have to use something so petty to argue about, obviously you don't have anything important to say.

"Oh yes I know I forgot to mention the I815 that intell had to pull out of there hat after the MTH issue.....Still could not beat the bx....but was that by design? Or was it that good ol' intell did not want a SDRAM chipset making there RAMBUS counterpart look even more pathetic?"

And why is it everyone who thinks about this fails to realise two important things:

1) The 440BX was NOT specced at a 133MHz FSB. Yes, it can run at it, but it wasn't designed to. And thus it won't overclock as far as the i815 will.

2) The 440BX running at a 133MHz FSB is using an overclocked AGP bus. The i815 running at a 133MHz FSB is using a the standard AGP bus speed. So right there is the explanation of the performance difference. It's only because the AGP bus is overclocked that the 440BX outperforms the i815, you nitwit. It has nothing to do with Intel trying to make their SDRAM solution look bad.

"However to try to force there consumers to use the product when it offers no resonable advantage (at least at the time of release) at more than three times the price and to get paybacks from Rambus for doing so is more like extortion"

The P4 is the ONLY Intel chip that requires RDRAM. And how do you even know that there is no reasonable advantage? Have you run a P4 with DDR SDRAM? I can see several reasonable advantages of RDRAM over SDRAM for the P4.

First, there's the fact that RDRAM does have a bandwidth advantage. Second, there is the fact that RDRAM has a higher potential for bandwidth growth than DDR SDRAM does. Yes, eventually QDR SDRAM will come out. Third, when QDR SDRAM does come out, don't you think that you'll have to use a completely different motherboard to use it, just like you have to do when upgrading from single-rate SDRAM to DDR SDRAM? With RDRAM you don't need a new motherboard. You just put the new memory in and go.

There are several advantages that RDRAM has over SDRAM; as painfully annoying as it is to admit. And there are of course several advantages that SDRAM has over RDRAM. It's the job of the consumer to decide which they need.

And what, pray tell, could I do if I wanted an RDRAM Athlon combo? There isn't anything that I could do. So before anyone condemns Intel for only supporting one memory type with their P4, maybe they should think about the fact that there is only ONE memory type that AMD supports. At least you can use either SDRAM or RDRAM with a P3.

"Wrong again my friend did you sleep thru the whole motherboard recall? "

Did you even attempt to read what I said? I quite clearly specified OR INSTANTLY RECALLED. Which is EXACTLY what Intel did with their MTH crap. If anyone is using one of those screwy things today, it's their fault because Intel HAS offered to replace and/or reimburse.

Do we EVER get replacements or reimbursements for those crappy early revision VIA motherboards? Ha!

"Oh boy! On one specific application the p4 has proven with optimized software it can outperform the athlon!!! Hurray! and at three times the cost too! Dam I should buy a dozen of them."

I would think that any company who's future plans are to implement SSE2 optimizations into their code or to use software that has SSE2 optimizations, they would do that very thing.

Yes, from a home-user's perspective it seems rather silly. But the majority of PC sales are to companies, not to home users. You have to remember that you're not the only person in this world and that the universe doesn't revolve around just you.

"Oh yes by all means. If you do not install a heatsink and fan properly on your amd chip it will die. This makes it a questionable product. Hmmm, now lets see if I drive my car and don't put on my seatbelt and get in an accident I could die. By your logic then all cars are questionable too! If I plug in my cd-rom backwards it could die...dam this is a questionable product as well. So evertyhing that is not completely idiot proof is questionable...go figure!"

This only shows how little you cogitate before writing a response. Your comparisons clearly are moronic. If a car manufacturer were to leave out something as simple as a place for an oil filter, don't you think that'd be a serious problem? It keeps the engine from becoming seriously damaged. It's cheap to put in. It's easy to put in. It ensures extra protection.

Anyone who were to run a car without an oil filter will have to constantly ensure that their oil is clean so that the engine doesn't become unnecesarily damaged.

Where as anyone who has a car with an oil filter simply doesn't have to worry nearly as much about such problems.

The heat protection IN the CPU is no different than this. It's a standard level of security implemented to increase the lifespan of the chip. And anyone who leaves such a simple thing to implement out really appears of questionable ethics to me. Maybe such isn't the case to you. But to call Intel shady for working with Rambus and to then not call AMD shady for leaving out such a simple and useful component in their chip with no good reason, well, it makes you appear incredibly biased.

- Sanity is purely based on point-of-view.

So who buys new (overpriced) processors (such as the P4) when they first come out? Foolish people you say... like perhaps the Government and huge corporations with big buggets to blow. "If you don't spend your whole budget this year they'll just cut it back next year..." Go read a Dilbert cartoon book if you don't know what I'm saying... In fact I'm sure some companies don't buy AMD chips just because they are priced lower. It's always that way. Why would Intel cut back P4 prices when government departments are snapping them up like hotcakes? Intel is still pushing the PIIIs for the home market, as their price cut to bring them down to the TBird level shows. Eventually the governments will have finished "upgrading" all their PIII-933EBs to P4-1500s "so the civil servants can type out their Microsoft Word documents faster", and then Intel will cut the prices of the P4 chips so the home users (who are on budgets) can afford them.

In a way we shouldn't complain, the government gets to deal with the early buggy revisions, and by the time the price cuts come around there should be optimized software out. Then again it is our tax money they're wasting.

- JW

All I have to say is "dayum" this is a long thread, and full of more information than I find useful. All I can gather is that "AMD's" have problems, but are cheap and fast, and Intels are stable, but not worth the cost. So I've summed up this 10 page long thread in a few sentences?

"It's unfortunate that people like yourself have been brainwashed into believing that AMD can't do, especially in the corporate desktop space, the same job as Intel. "

Okay then show me a dual-CPU AMD server. And show me an AMD system that DOESN'T have ANY problems running Windows 2000.

"How can you claim that Intel is stable with all the engineering fiascos they've had over the past year?"

Show me a single Intel CPU and mobo system today that is still using any faulty technology. You can't, can you? Why not? Might it be because the P3 1.13GHz was recalled BEFORE it was sold? Might it be because the motherboards using the MTH were replaced, AT INTEL'S COST?

Oh, but lets look at an AMD system. Hmm ... there are AMD compatability problems with Windows 2000. There are a great many early revision VIA motherboards that have a huge number of issues with just about every type of hardware in existance.

Yes, buying the LATEST revision VIA motherboards (or only very recently avoiding VIA entirely) will solve MOST problems with AMD systems. But that still leaves a lot of people who already have faulty hardware that no company is EVER going to replace AT THEIR OWN COST.

When you go with Intel systems you have the assurance that it'll work or that if it doesn't work, it'll be replaced with no charge to you. In fact 99.9% of the time you'll never even need to update a driver. The only real reason to do it is for minor performance gains, not to fix bugs.

When you go with AMD systems you push your luck. It's all a roll of the die. Maybe you'll come out a winner. Maybe after you patch things you'll come out a winner. And maybe you'll never be a winner no matter how many patches you use and no matter how many times you contact tech support.

For a business, Intel is often a less expensive way to go. Yes, the computer costs more initially. But think of how much a company a highly-skilled software engineer. We're talking about ranges from $20 to $100 an HOUR. And in just a day spent trying to download and install the latest drivers, an AMD system can have cost the company more in lost man-hours plus system cost an an Intel system that is going to work without driver updates would have cost initially.

- Sanity is purely based on point-of-view.

I think anyone who claims that the Intel folk's biggest argument against AMD is that AMD chips run hot must not be paying attention.

Obviously a decent heat sink solves that problem.

I haven't even heard anyone make that argument in ages.

But what about arguments like a lack of thermal protection? (A sensor to shut off the CPU if it starts to overheat.) What about Windows 2000 incompatabilities? What about VIA motherboard incompatabilities with a LOT of hardware? What about AMD not offering multiple-CPU systems? What about AMD not yet implementing SSE2? What about all of the driver updates that you'll have to install with an AMD system just to get it to work when an Intel sytem will work without driver updates?

I could go on even longer, but I think I've gotten my point across already.

- Sanity is purely based on point-of-view.

You've still missed a lot.

What the point SHOULD be is that each consumer should purchase a system that meets their needs. As such, neither AMD nor Intel is a perfect solution to everyone's needs.

That SHOULD be the summation.

But it isn't.

People are still trying to argue that AMD meets everyone's needs and Intel is just overpriced and not worth that price. People like that really irk me because it's pretty obvious that such isn't the case. Yet they just can't admit that there might actually be perfectly GOOD reasons to use EITHER system.

So the ACTUAL summation of this thread is not how one should just research and then purchase the system that best meets their needs. It is instead an amazingly long debate to try and prove to these closed-minded blockheads that both systems have their advantages and that the world would not be better off to just use only one company's computers.

- Sanity is purely based on point-of-view.

AMD has done a pretty good job so far. Give them some time and all those issues that you have listed will be resolved.

"Is this really true? Or is it the VIA chipset specifically?"

There are both. The Matrox G400/Win2K/VIA mobo lockups are a VIA problem. But there are other misc. performance problems with AMD chips running under Windows2000.

"Do you feel that I came across in such a fashion?"

Nah. You're pretty reasonable and talk intelligently. So that's pretty cool. You just seem to not realise that there are people out there who have really annoying problems with AMD systems that they wouldn't have had if they had an Intel system.

Not that this in itself is all that bad. It just means you have a little to learn. And at least you're smart enough to.

I put that in specifying, "But for everyone who", to get a jab in at all those ignoramusi who keep arguing AMD is god and Intel sucks even when you point out the flaws in their arguments and they respond to your perfectly good arguments with some pretty stupid responses. Such as Ncogneto.

People like that just can't admit that there are perfectly good reasons for going with an Intel system. They think just because they're so special, everyone should do exactly as they do.

Me, I just say go with whichever system meets your needs best. For people who can troubleshoot a PC while blindfolded, go with an AMD. For people who need a dual-CPU system, need to use Windows 2000 heavily, or just don't know how to fix a PC, they should go with an Intel system.

Each have their advantages. It's just that simple.

- Sanity is purely based on point-of-view.

slvr_phoneix wrote:
But there are other misc. performance problems with AMD chips running under Windows2000.

Can you tell me specifically what you have heard? I would like to check my system to see if they apply in my case.


slvr_phoneix wrote:
You just seem to not realise that there are people out there who have really annoying problems with AMD systems that they wouldn't have had if they had an Intel system.
Not that this in itself is all that bad. It just means you have a little to learn. And at least you're smart enough to.

I do realize that other people out there have problems. I was just hoping to inspire some people with a nice success story. :-) I agree that if a person has no experience with computers and troubleshooting, they should purchase an Intel system. For those who enjoy working with computers and have a reasonable amount of knowledge, then I find that AMD systems have a nice p ratio. Cheers.

John Garrett
System Admin - www.elementk.com
Editor - Exploring Windows NT Professional