Call me Crazy.. but i have a feeling AMD is up 2..

I tend to agree. It has been quite awhile since the 1.4 was released I think they're going to ambush Intel again, I say that you should look to see them increase the speeds of their processors a LOT in the near future... within the next year there will definitely be well over 2.0 ghx AMD processor, I guarantee it.

Duron 700
Windows ME
GeForce 2 MX
15 GB HD
128 MB RAM

The 1.4ghz tbird overclocks to 1.6 easily, I bet they release the palomino at 1.7, that would equal a 2.2ghz p4.

~Matisaro~
"Friends don't let friends buy Pentiums"
~Tbird1.3@1.55~

It took them 8 months going from 1.2 to 1.4. The only reason they still have a good performance against P4 (went from 1.3 to 2.0) is the much improved chipsets witch are comming out (KT266A is very fast).
Im sure they have been stockpileing pallys the last 8 months.

It didnt take them 8 months, they had 1.2s overclocking to 1.5's when the axia stepping came out, they WAITED to release the higher speeds to keep even with the p4 hoping to win on the price comparison. The 1.4's are overclocking WELL, and the athlon mp's are also hitting 1.5ghz.

~Matisaro~
"Friends don't let friends buy Pentiums"
~Tbird1.3@1.55~

Yup!

Sounds like you just got hit in the head by a HAMMER!!

Ouch!!

Bet it hurts!

(pun intended)

Mark-

When all else fails, throw your computer out the window!!!

Northwood... P4 with FPU fixes and 512kb L2 Cache. Coming soon and intel showing 3.5ghz northwoods.

Nice Nvidia and ATi users get a Cookie.... :smile: Yummy :smile:

It is true that AMD has something up their sleeve- the sledgehammer (and later clawhammer- how much do u guys think these will cost?). However, I am worried by the fact that the P4 will have a better FPU at 3.5 Ghz and therefore will almost match the Palamino clock for clock. Is it possible AMD may not be able to avert this?

When I rule the world, Apple will only mean the fruit.

I wouldn't be too worried, actually that would be good news as long as AMD can counter and remains profitable. Intel needs to turn up the heat. Funny thing is that AMD could have been the one to turn up the heat if they released the Polimino desktop 2-3 months ago at 1.5ghz plus. They still can but when the Northwood really hits, the lines I think will be clearly drawn between the top performer and the second best in the cpu world. The P4 was just a stop gap tolling with AMD making them feel over confident, Intel I see will hit and hit hard. Hopefully AMD has projected this and are pushing up their 64/32 bit desktop chip which Intel does not have an answer to. Interesting times are indeed ahead, lets hope AMD can produce otherwise it will get pretty ugly for them.

<P ID="edit"><FONT SIZE=-1><EM>Edited by noko on 09/03/01 05:46 PM.</EM></FONT></P>

Ahwoogah ! AMD haven't got it in them, I bet they are running around like a bunch of headless chickens.

<font color=purple> **I eat confusion and spit it out slow**</font color=purple>

>The 1.4ghz tbird overclocks to 1.6 easily, I bet they release the palomino at 1.7, that would equal a 2.2ghz p4.

only in MSword.

"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"

Stop contradicting yourself, you're making yourself look like an idiot.

--
It's Princess Leia, the yodel of my life. Give me my sweater back or I'll play the guitar.

Not so oh great amdmeltdoofus. Take another look at the benchmarks, ALL fpu intensive apps make the p4 look like [-peep-].

~Matisaro~
"Friends don't let friends buy Pentiums"
~Tbird1.3@1.55~

I'm not worried about about 3.5 Ghz Northwood. Intel releases on Moore's law predictability. Their processor speeds (at least in MHz) double every 18 months, making the release of the 3.5 GHz due about the first quarter 2002. It has been almost 18 months since the P3 1000, so they're a little ahead of schedule in MHz, though not performance, on the current P4. AMD, though, has until october to release right on schedule.
Intel really isn't ahead of schedule, though, since I guarantee that the P4 2.0 is NOT 100% faster than the P3 1000, even if the clock speed is. When they release the 2.0 GHz Northwood, I think they will be fulfilling their Moore's law obligation.
It is exciting though, just think how fast computers will be 5 years from now!

Aklein

Just me and my computer livin' on the dark side of the moon.

Rcf... showing is the operative word. Intel is likely at least a year away from releasing a 3.5ghz P4. They had to use some serious cooling on the sample chip they showed and THAT system didn't exactly work well far as I read.

So, that's no concern right now.. especially when a 1.7ghz Palomino will equal a 2.4/2.5 P4 on a KT266a based motherboard.

Mark-

When all else fails, throw your computer out the window!!!

Hey Everyone,
Regardless of what AMD does or doesn't do, I for one, hope AMD does come out with a Blind-Sided Intel Sack! You can say all you want about AMD being the Best or Intel Always Rules.........But the truth is a Competitive AMD makes both companies produce a better product for less! I love the way we have a wide array of hareware upgrading options for our PC's. Life would suck without AMD..... we all might as well get a Mac!
Peace Out..........tile

What do you mean, SEE THE BIG PICTURE, you mean I need a New Monitor too!

Nothing beats a well planned sneak attack.


<font color=green>I do whatever my Rice Krispies tell me!</font color=green>

Actually, Moore's law has to be modified every so often. The time frame for processor speeds to double is getting shorter.

When I rule the world, Apple will only mean the fruit.

Right, but still, Intel came out with the 1000MHz in April 2000, its now september 2001, 17 months later, and the performance of the P4 2000 is not double the P3 1000, though the soon to be released northwood will probalby be twice as fast.

Aklein

Just me and my computer livin' on the dark side of the moon.

Guys,
Your assumptions are based on one incorrect hypothesis: that the Athlon's performance rises at the same rate per 100 MHz jump that the P4's does, this is not true. The P4 performance jumps more every 100 MHz than an Athlon's. So, to say an Athlon 1.5 will be equivalent to a 2.2 P4 is a little over optimistic. The thermals of the Athlon mean they are having trouble scaling, this is why we have the intermiable wait on the 1.533 GHz piece. the Athlon is reaching the end of it's sweet spot with regard to ramping, and even the Palomino is just a final stretch, to try and bridge the gap to their next gen product. The other thing which I find amusing is when AMD were first to 1GHz , clockspeed apparently was a lot more important to them than it is now. The P4 is designed to scale, and scaling it certainly is. Overclocking does not tell the true story about what life's left in a processor.

Tejas you are correct, they do NOT gain the same per 100mhz, the tbird gains MORE, the 1.2 took on the 1.5, the 1.3 bested the 1.7 and the 1.4 bested the 2(with newer chipset admittedly). Athlon mhz are worth more than p4 mhz, and the scaling of the benchmarks shows just that.

~Matisaro~
"Friends don't let friends buy Pentiums"
~Tbird1.3@1.55~

Matisaro

Exactly

--call it what you wish, with this machine I can make mercury flow in 3 directions at once--

Yes, AMD is going to have to come up with a new core to overtake the 3.5GHz P4! Even switching to .13 micron wil only get them as far as maybe 2.5GHz (hey, maybe 2.8GHz if they're lucky, but then Intel will have the 4GHz out!). Time for a new core AMD.

Back to you Tom...

Funny how you all Intel lovers seid that when the p3 1.13. Flooped

What? You mean the P3 didn't need a new core to get to 1.13? That's news to me! I hate Intel as a company, they abuse their customers. But until recently, what other choice did I have for a good chipset? Maybe I should cancel my Palamino upgrade and wait 5 years for the price of the 4GHz P4 to drop, seeing how I love Intel so much? Or maybe I should skip that and wait 10 years for IA64 to work it's way down to the desktop? Who do you think your talking to, Journeyman?

Back to you Tom...

Actually, Tejas, as others have mentioned, Athlon seems to scale better per 100mhz than does the P4.

You are correct in your assessment of thermals. Palomino is designed to run 20% or so cooler, which will hopefully allow it to scale 20% higher (to around 1.7ghz)

As for clockspeed results etc...

When Athlon reached 1ghz, P3's were nearly as powerful clock for clock. Since that time, Athlons have increased performance clock for clock. Meanwhile P4 has come out and it is around 25% less powerful than the P3 clock for clock at least! The performance difference is even worse when comparing it to Athlon.

So what has Intel done?

They've created a chip that can ramp to much higher higher clocks, but the performance is shoddy at best....it's one of the lowest performing PC chips Intel has produced in years on a clock for clock basis. THIS is why clockspeed is less important to AMD. If you can't see that then you haven't been paying attention to the many many system benchmarks on numerous web sites showing just this.

In a nutshell:

When chip performance was nearly on par (within 5% or so clock for clock) clockspeed was a good indicator of the respective performance of the chips. Now that Intel has changed the rules with it's much poorer performing P4 chip, that isn't a good indicator of performance anymore.

Mark-

When all else fails, throw your computer out the window!!!

Calm down, Crash. He's talking about the original 1.13, not the Tully. And I don't think he was trying to say you didn't know what you're talking about.



<font color=blue>Quarter pounder inside</font color=blue>
<font color=red>Change the Sig of the Week!!!</font color=red>

Yes, well, read the post he was responding to. AMD WILL need a new core to make it to these higher speeds, which is what I originally said. And Intel DID need a new core revision to hit 1.13 reliably, which is what HE said.
Now about the Palamino, it sounds like it will be a performance bargain. And these newer chipsets are looking better all the time. So someone can now (and with increasing ease later) build an AMD system with a decent motherboard (finally). But that's not going to be al that important when Intel is making P4's that are ONLY 20% SLOWER clock for clock, but with a 100% faster clock rate! I'm looking toward the future, and AMD needs to get a new core to the market capable of matching Intel's clockrate with 6 months, or they will not be able to compete in the performance market. Tom even stated that 6 moths ago (and made a lot of enemies with the statement). And we know that Tom, after getting SCREWED by Intel, pulls no punches concerning them. And with no DIRECT competition from AMD, the market is starting to look as bleak as it was 5 years ago.
If AMD had any tricks up their sleeve, wouldn't they let us know? We knew about the original Athlon about a year in advance! And the Hammer is not a viable desktop processor, considering that it will perform on par with their other desktop processors in 32-bit applications. There won't be 64-bit apps coming to the market for the desktop FAST ENOUGH to make the transition viable for the masses. Of course they could come up with a dual 32-bit path on a single processor, as opposed to a 32-bit and 64-bit path on the Hammer. Whatever they plan on doing, they need to do it quickly, and I doubt it will happen.

Back to you Tom...

That's what I've been thinking. Both AMD and Intel usually spell stuff out a ways in advance.


Absolutely. Although for now, I'd still rather have a Tbird over a P4. Northwood could change that, but it isn't looking likely.



<font color=blue>Quarter pounder inside</font color=blue>
<font color=red>Change the Sig of the Week!!!</font color=red>

Well, if someone was just GIVING me a choice of either a FREE 2GHz Willy system with 1GB RDRAM or a FREE 1.4GHz Athlon with 1GB DDR, all other parts being the same, I'd take the crappy Willy. It still beets the 1.4GHz T-Bird, (barely) and will have a higher resale value.

Back to you Tom...

if the willy was on a DDR333 SiS645 Socket 478 board i would take over the tbird.

Nice Nvidia and ATi users get a Cookie.... :smile: Yummy :smile:

I'd take the P4 with RDRAM just because, when I sell it in a year, it will still be worth more.

Back to you Tom...

I'd agree, but only to resell it and either wait to see how Northwood + P4x266 look when they're out, or buy a nice Tbird setup and keep the extra.




<font color=blue>Quarter pounder inside</font color=blue>
<font color=red>Change the Sig of the Week!!!</font color=red>

I'd take it and sell it now for a discounted price and get enough money to buy another dual mp setup.

What is the difference between <font color=red>pink</font color=red> and <font color=purple>purple</font color=purple>? The <b>GRIP</b>!

i would wait for the 3.5ghz northwood. 512kb of L2 cache running at 3500mhz.

Nice Nvidia and ATi users get a Cookie.... :smile: Yummy :smile:

Not with kt266a heh.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

HA! The only reason nobody has found problems with the KT266A is that it hasn't been around long enough! VIA <i>always</i> has bugs in thier chipsets, sometimes their a little hard to find, but when they turn up, their even harder to solve!

Back to you Tom...

I sell <i>retail</i>. That means that it takes about <i>Six Months</i> for the systems I sell to loose enough value that I no longer make money on them. So if I won the system, I'd keep it for 6 months.

Back to you Tom...

Bias, with no evidence crash, we will see in time, but even tiny bugs like via is famous for will still not diminish the raw asskicking power of the kt266a.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

Umm...yes it will. Hopefully it'll be Via's first problem-free chipset, but I doubt it. They are taking their time on it, it seems. We'll see.

But anyway, it will diminish from the raw power. I'd rather
have something a little slower but stable. Maybe that's just me.

Little bugs? Did you forget about the wonderful 686b, just like the one in my box? I hate it. The Northbridge (KT133a) is awesome, I get better memory scores than some people with DDR. But that 686b...



<font color=blue>Quarter pounder inside</font color=blue>
<font color=red>Change the Sig of the Week!!!</font color=red>

Well, I don't really know. The only advantages the new version has is better memory bandwidth and more stable USB, correct? So hopefully nVidia will offer even better memory bandwidth with their dual channel solution.

Back to you Tom...

That is the understatement of the year, the kt266a is giving some 25% more performance than sdram and 10-15% more than the amd760, it is well worth any tiny patched bugs it may or may not have.(especially if you are upgrading from a kt133a like me).

As for the nforce, it may indeed be better, and if so w00t, amd's chipset situation is looking MOST awesome right now.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

I still don't trust VIA, and would be willing to sacrifice ALL that performance for a "good" chipset, as in one that will work with all my cards!

Back to you Tom...

To each their own. I would trade 25% performance for an antique legacy card any day, but that is just my preference.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

I will probably be getting a KT266a mobo, its performance looks to be very promising. My KT133a is extremely stable, never had any problems with the 686b south bridge, no SB live card. So I really havn't experience anything negative with this VIA chipset. Each chipset should be rated on its own merrits. Intels 820 chipset sucked, that doesn't mean all Intel chipsets suck, just that one.

Uh, every VIA chipset sinc the MVP3 has had problems, none were compatable with ALL my devices, I have not tried the KT266A, but I wouldn't personally put any faith in any company with such a poor track record. Intel was pressured into the MTH, and the 820 was originally supposed to be RDRAM compatable only. Intel tried as hard as they could to make it work with SDRAM, but failed. And RDRAM equiped 820 motherboards had no problems. VIA can't even work the bugs out of their chipset using the parts it was DESIGNED for.

Back to you Tom...

Some of these cards are "specialty" items I use to perform "special" task in real time, if you understand. Look, there's no way VIA is going to hold even a 10% lead for very long. And lots of people were defending VIA even when SiS was on top. My feeling is that if the SiS735 isn't fast enough, wait for the next version. Or the next ALi chipset. Or even the nForce. Sy I had bought a motherboard two weeks ago, with the 735 chipset, you would have thought it was great! And I would have probably not ahd any compatability issues. The chipset bus happesn so fast that by the time KT266A issues are identified, there will already be something better to replace it!

Back to you Tom...

And that is fine, like I said it is a matter of personal choice. I however would always purchase the fastest chipset available at the time I spent my cash, and we have yet to see any kt266a instability. Lets see what via has brought forth this time.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

They created a new southbridge for the kt266a, the origional kt266 had the 686b right, now the kt266a has a 686c, perhaps they finally decided to clean up their act.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

Since IBM have just released their SOI chips is it possible that AMD have managed to sneak onto SOI sooner than anticipated? If so they could ramp the Palomino to 2Ghz while still on the 0.18u process?

<font color=blue> The Revolution starts here... as soon as I finish my coffee </font color=blue>