Celeron, Duron, or Samuel II

[Arcadian]

I wanted to spawn a discussion of next year's low end champ. Is it going to be AMD's Duron, the Intel Celeron, or VIA's Samuel II chip. Those seem to be the three choices, unless you want to include the mobile arena (Transmetta, in other words).

Intel's Celeron will supposedly increase its front side bus to 100MHz, which we know from overclocking to be a significant improvement. AMD will release Morgan, but this won't be until February at the least. VIA has been pushing the Cyrix III, which some say underperforms the Celeron, but VIA's new chip, the Samuel II, will supposedly give Celeron and Duron a run for their money.

I want anyone to provide any info on these upcoming products. Keep your info technical, though. "Processor X will kick a$$" is not acceptible by itself. Rumors are OK, as long as you can back them up. Let's get a technical speculative discussion going.

[This message has been edited by Arcadian (edited October 04, 2000).]

[Ymaster]

Originally posted by Arcadian:
I wanted to spawn a discussion of next year's low end champ. Is it going to be AMD's Duron, the Intel Celeron, or VIA's Samuel II chip. Those seem to be the three choices, unless you want to include the mobile arena (Transmetta, in other words).

Intel's Celeron will supposedly increase its front side bus to 100MHz, which we know from overclocking to be a significant improvement. AMD will release Morgan, but this won't be until February at the least. VIA has been pushing the Cyrix III, which some say underperforms the Celeron, but VIA's new chip, the Samuel II, will supposedly give Celeron and Duron a run for their money.

I want anyone to provide any info on these upcoming products. Keep your info technical, though. "Processor X will kick a$$" is not acceptible by itself. Rumors are OK, as long as you can back them up. Let's get a technical speculative discussion going.

[This message has been edited by Arcadian (edited October 04, 2000).]

Duron 100%

The locked multipyers of the newist Celerons kind of messed them up. You can't push the fsb without going to far to fast.

The Via chip with cyrix is not going any place. I spit on cyrix! Curesed to play games on one a long time ago.

[Arcadian]

So Ymaster, do you think AMD's attempt to lock the multiplier on its processors would weigh your decision differently?

While it's true that Celerons have had locked multipliers for some time, the jump to 100MHz front side bus should make a large improvement to performance. Do you think that a 700/100MHz Celeron will overclock easier than a 700/100MHz Pentium III? With a smaller cache, it may be possible....

Also, keep in mind that VIA's Samuel II is not using the Cyrix core. VIA figured that there was some name recognition with the Cyrix name, so chances are they will call their new chip that. However, they are using a completely different core from a company called Centaur. If anybody has any information on the performance of this new chip, can they please repond?

If the Samuel II is that much better, Ymaster, what would your opinion be then?

[Ymaster]

Originally posted by Arcadian:
So Ymaster, do you think AMD's attempt to lock the multiplier on its processors would weigh your decision differently?

While it's true that Celerons have had locked multipliers for some time, the jump to 100MHz front side bus should make a large improvement to performance. Do you think that a 700/100MHz Celeron will overclock easier than a 700/100MHz Pentium III? With a smaller cache, it may be possible....

Also, keep in mind that VIA's Samuel II is not using the Cyrix core. VIA figured that there was some name recognition with the Cyrix name, so chances are they will call their new chip that. However, they are using a completely different core from a company called Centaur. If anybody has any information on the performance of this new chip, can they please repond?

If the Samuel II is that much better, Ymaster, what would your opinion be then?

Yes it would...
But Durons are normally set to 100 fsb from the start.
It would also make me think twice if the Celeron had a 100 fbs normally..like a 900 celeron at 100x9=900 would be sweet.
But knowing Intel it would be 13.5x66=900 rounded.. It also forces you to buy a new motherboard that has a 13.5 Multiplyer
. So this does not save money in my eyes...

[nkeezer]

I'd have to say that Duron will probably be the low-end champ. While overclocking the Celeron to 100 FSB does make for a pretty significant performance boost, most benchmarks I've seen show that it still lags behind the Duron clock for clock. Furthermore, it also seems that the discrepancy between the two chips gets bigger as speeds increase. Possibly the result of the reduced L2 cache associativity of the Celerons? If that's the case, then things could start to look even worse for Celeron once it starts getting up into 850+ territory.

Another thing that may make a difference would be motherboard support. IF Intel bumps up the performance of the Celeron (by either going to 100/133 FSB and/or increasing the L@ associativity), then they might run the risk of cannibalizing PIII/4 sales (I'm sure they remember what happened with the PII). So I'd guess that they might take steps to ensure that the Celeron remains a low-end solution. I could definitely see Intel something along the lines of preventing a Celeron from working with DDR or Rambus. Especially considering the fact that the PIII is slated for a die shrink in the near future. We know the .18 Coppermine die is only good to about 1.1GHz. So the PIIIs have topped out, but Celeron still has a ways to go. Whole lotta stuff that could be going on over on Intel's side...it'll be interesting to see what happens.

As for the Samuel II: I don't know. Cyrix chips have been historically bad performers. It's tough to say if they'll manage to pull a complete about-face and pump out a good chip. My gut feeling is that it won't happen. The fact that Via is making it, though, could make it a contender. I'm sure that they've been developing the chip and it's accompanying chipsets hand in hand. So maybe Via will have some tricks up it's sleeve that wrings a little extra juice out of Sammy?

------------------
www.nweb-design.com <-- Send me a client, I'll send you a 5% finder's fee

"You have just destroyed one model XQJ-37 nuclear powered pansexual roto-plooker....and you're gonna have to pay for it." -- Frank Zappa

[LiquidGoop]

Seems like thos forum has a population of 3, nkeezer, ymaster, and arcadian. Seems like people want to stay away from the "Highly Technical" PC forum, but to return to the topic at hand, I agree with nkeezer, celeron even at 100FSB is slower clock for clock than durons. And VIA chip isn't even out yet, and you can't completely judge a product on the companys history. However, I feel that it will not be enough to compete with the duron.
What socket is the Samuel 2 using anyways?

[Arcadian]

Originally posted by LiquidGoop:
Seems like thos forum has a population of 3, nkeezer, ymaster, and arcadian. Seems like people want to stay away from the "Highly Technical" PC forum, but to return to the topic at hand, I agree with nkeezer, celeron even at 100FSB is slower clock for clock than durons. And VIA chip isn't even out yet, and you can't completely judge a product on the companys history. However, I feel that it will not be enough to compete with the duron.
What socket is the Samuel 2 using anyways?

Well, this topic was just created today, so I hope more people decide to start technical discussions here as they find out it exists. I mean, you just posted here .

BTW, Samuel II will use socket370. Cyrix has always used the same socket as Intel CPUs, and I think VIA plans to continue that.

[Rick_James9]

Originally posted by LiquidGoop:
Seems like thos forum has a population of 3, nkeezer, ymaster, and arcadian. Seems like people want to stay away from the "Highly Technical" PC forum, but to return to the topic at hand, I agree with nkeezer, celeron even at 100FSB is slower clock for clock than durons. And VIA chip isn't even out yet, and you can't completely judge a product on the companys history. However, I feel that it will not be enough to compete with the duron.
What socket is the Samuel 2 using anyways?

I was going to stay out of this one. But what the hell! The low end champ will be the Duron. Why?

1. The FPU is the key. AMD finally got the "uppperhand". And this IS the key to CPU power(RISC anyone?). This why years ago you had to buy a math co-processor.

2. The motherboards will have faster a FSB. I'm sorry its a little late for 100mhz for the Celeron.

3. VIA? Well, they may own Cyrix but I don't think they're up to making CPUs even close to AMD or Intel standards. They are still trying bring some respect to their chipsets BTW, IBM you did very well getting rid of Cyrix; Smart move.

4. Finally, AMD has proven the Duron is a excellent cpu and can sell it very cheap.



------------------
Tech GOD

[Arcadian]

Originally posted by Rick_James9:
I was going to stay out of this one. But what the hell! The low end champ will be the Duron. Why?

1. The FPU is the key. AMD finally got the "uppperhand". And this IS the key to CPU power(RISC anyone?). This why years ago you had to buy a math co-processor.

2. The motherboards will have faster a FSB. I'm sorry its a little late for 100mhz for the Celeron.

3. VIA? Well, they may own Cyrix but I don't think they're up to making CPUs even close to AMD or Intel standards. They are still trying bring some respect to their chipsets BTW, IBM you did very well getting rid of Cyrix; Smart move.

4. Finally, AMD has proven the Duron is a excellent cpu and can sell it very cheap.

1. I totally agree. FPU is very important in a lot of apps. Intel definately needs to catch up in this department. I have a feeling, though, that AMD won't hold the FPU crown forever, though

2. I'll buy that argument, too. Upgrading the Celeron to 100MHz seems a little quaint when you compare it to a Morgan on 266MHz front side bus. I wonder if Intel will do a Celeron version of the Pentium 4. With a 400MHz FSB, the Celeron would be much more competative. Intel could even clock it at 1GHz, which would be slow for a Pentium 4, but not for a Celeron. Of course, Intel won't have the resources to consider this option until at least March (coinciding with Morgan, perhaps???)

3. Well, VIA may not have much to compete with Intel or AMD, but whatever they sell means that either of the two current champs will lose market share. How much? Even 5% would hurt, and I shutter to think of 10% being within the realm of possibility.

4. Yes, I imagine AMD will sell as many as they can produce, but how much is that? Intel can probably produce at LEAST 5 times the amount of Celerons as AMD's Texas plant can spit out Durons. Remember Dresden is ramping up for faster Athlons, and I don't believe AMD will take up space for Durons up there.

Thanks for replying!

[LiquidGoop]

We keep talking about duron being "low-end", but in reality it is a relatively fast processor that can hold its own against the likes of P3 and the TBird. It's admirable that the duron is even compared to those chips, bein "low-end".
When I was building my new box, there was no contest between celeron and duron. The celeron only seems like a good choice if you have an existing BX board and decide to upgrade, not buying a whole new box.

[Arcadian]

I agree that Durons are not low end processors, but the fact remains that AMD is making them low cost, and thus positioning them as "low end". One of AMD's mistakes is that the Duron performs too well in comparison to the Athlon. I'm sure this will affect their Athlon sales, when people start wisening up to the performance delta between the two.

[nova]

I think it is possible that the duron could be made to perform even better than now. With the news about qdr sram and later on qdr sdram I think I have read it is possible to move the EV6 bus up to qdr from ddr. This could provide fairly large jump in performance.

------------------
Remember, its only a dream!

[slick]

I don't know LiquidGoop this forum is for the "Highly Technical" maybe that alone scares ppl off but hey i happen to like this forum.
Not getting into the numbers so much i would have to say the Duron will be the king of the lowend just because people like more bang for there $$$.

[Adisharr]

I am thinking that the Samuel will be a dog. VIA makes a good chipset (better when they take care of these graphics card incompatibilities) but I don't believe they have a lot a cpu experience.

If the P3 loses to a Duron in some tests, that leads me to believe the the Celeron is still going to lag behind. I doubt Intel will improve that chip beyond the increased bus speed.

Regarding the centaur core, if that's the same company they made low cost 486 upgrades - they are doomed IMHO

$ .02

------------------
- Mouse not found - click to continue..

[Arcadian]

Originally posted by Adisharr:
I am thinking that the Samuel will be a dog. VIA makes a good chipset (better when they take care of these graphics card incompatibilities) but I don't believe they have a lot a cpu experience.

Regarding the centaur core, if that's the same company they made low cost 486 upgrades - they are doomed IMHO

Well something that interested me concerning VIA's new chip is a proposed system that will cause in total $199. Even if it performs like crap, I think a lot of people will be interested in buying something that dirt cheap. What do you think?