Sign in with
Sign up | Sign in
Your question

Canterwood owners and buyers have been screwed!

Last response: in CPUs
Share
May 21, 2003 8:15:46 PM

According to THG's latest comparison of chipsets, Springdale TIES the Canterwood, making this piece of crap at the moment a real joke. Any bloke who pays more than 50$ more for Canterwood for this supposed joke of a performance boost, needs a serious treatment. The figures of the benchmarks, show that the P4P800 is on par if not 0.1% (and I'm being GENEROUS HERE!) less performing.

I don't understand this, is Canterwood seriously unoptimized and has no real chipset driver that is appropriate?

In any case, AMD has absolutely lost, there is little reason now to buy any, when the 2.4 800 WITH HT is a deal-blower. They are seriously powerful chips, the 2.8 was even outrunning the old 3.06GHZ sometimes! That and a cheap Springdale will really kick some serious butt.

And I'm just sad for Canterwood owners and future buyers (DON'T!), since they just got royally screwed by paying so much more, when you get the same features.

--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue:
May 21, 2003 8:55:46 PM

Anandtech.com paints a different picture (<A HREF="http://www.anandtech.com/chipsets/showdoc.html?i=1823&p..." target="_new">see here</A>) showing about a 4.5% boost for Canterwood over Springdale but your point is taken. Canterwood doesn't seem worth the extra cost.

<b>56K, slow and steady does not win the race on internet!</b>
May 21, 2003 9:43:16 PM

The only difference between Canterwood and Springdale is PAT, which produces lower latency memory accesses. Seeing as a jump from FSB533 to FSB800 brought about a ~10% increase in performance (possibly less in some applications, more in others), I'd say a ~5% boost from lower memory latency alone is to be expected. Is it worth the money? That depends on you. Is paying $600 for a 3.0C worth it?

"We are Microsoft, resistance is futile." - Bill Gates, 2015.
Related resources
May 21, 2003 9:59:28 PM

Not worth it to me! Canterwood was way overhyped while Springdale seemed to come in somewhat under the radar screen.

Does this all mean RDRAM solutions are not dead?

<b>56K, slow and steady does not win the race on internet!</b>
May 21, 2003 10:00:58 PM

Ok, I'm not saying Springdale is out of the picture, but considering that my perf upgrade is already small by going from the XP3000 to P4 2.8 800, then I really want to squeeze the extra perf out of my new setup. Sure I could go for the 3.0 P4 combined with Springdale, but that'd still cost more, and give more perf of course, but I can't afford it right now. I'd much rather have a 3.2 P4 or even 3.4 Prescott when I get the money and lust to upgrade again.

I don't believe Canterwood is unoptimized though. I just think you won't get any perf increase at all in benches that don't rely on memory latency and/or bandwidth.
But, like I said in my reply to you at that other thread, I don't wanna risk to lose 10% in new games coming out when I honestly can get the C-wood for only $40 more. That's the cost of 3 CD's here in Sweden. (And I buy lots of those)

The 2.4 is really nice, but the 2.6 and 2.8 is well-priced too. Unless you want absolute top perf, the 3.0 is...well, not so well-priced though.


My system: AMD Athlon XP 3000+ CPU / TwinMOS 1Gb DDR400 / Soltek 75FRN-RL /
Sapphire Radeon 9800 Pro / Antec True Power 550W / Maxtor 80Gb ATA-133 / Hercules GTXP SC /
Samsung DVD / Lite-On CDRW
May 21, 2003 10:03:26 PM

Warning, I read a lot that I don't understand, but this may apply anyway...

I read a piece on the new intel chips that said the full benefits of hyperthreading and the 800FSB wouldn't be realized until the next chip that will have a 1MB cache. It's supposedly scheduled for Q4.

Is it possible that the new board is a little ahead of the chip?
May 21, 2003 10:05:42 PM

My bet is that the P4's pipelines are starting to run full. Prescott will have longer pipelines, and thus you will probably get more from the 800 FSB, that's my bet. But not my science.

My system: AMD Athlon XP 3000+ CPU / TwinMOS 1Gb DDR400 / Soltek 75FRN-RL /
Sapphire Radeon 9800 Pro / Antec True Power 550W / Maxtor 80Gb ATA-133 / Hercules GTXP SC /
Samsung DVD / Lite-On CDRW
May 21, 2003 10:16:12 PM

Strains on the memory subsystem has nothing to do with pipeline depth. It has to do with clockspeed and cacheline size. Once the P4 reaches 4 GHz (if it ever gets there), we'll see FSB800 become more fruitful compared to FSB533

"We are Microsoft, resistance is futile." - Bill Gates, 2015.
May 21, 2003 10:19:50 PM

but that was exactly what I meant.. ;) 

My system: AMD Athlon XP 3000+ CPU / TwinMOS 1Gb DDR400 / Soltek 75FRN-RL /
Sapphire Radeon 9800 Pro / Antec True Power 550W / Maxtor 80Gb ATA-133 / Hercules GTXP SC /
Samsung DVD / Lite-On CDRW
May 22, 2003 2:41:03 AM

Chances are if you're considering Springdale/Canterwood, you're also going to buy a "C" variant Pentium4. I don't see the logic in throwing out ~5 percent of performance, replacing Abit IC7 with Abit IS7 for example in which case the price difference is about $40, when you've already spent at least about $200 for the processor. Everyone's is entitled to an opinion though.
May 22, 2003 3:38:57 AM

[H]ardOCP also paints a bit of a different story.

Why is it that THG almost never is in line with other reviewers lately?


<font color=green>The Netherlands is where you go when you're too good for heaven.</font color=green> :tongue:
May 22, 2003 5:04:11 AM

Actually i have seen that too

--------
The only thing that i truly know...

is that i know nothing at all.
May 22, 2003 5:45:43 PM

Now this one review is VERY strange...can someone explain why the Springdale is OUTPERFORMING Canterwood in this review?

http://www.hexus.net/review.php?review=556&page=1


My system: AMD Athlon XP 3000+ CPU / TwinMOS 1Gb DDR400 / Soltek 75FRN-RL /
Sapphire Radeon 9800 Pro / Antec True Power 550W / Maxtor 80Gb ATA-133 / Hercules GTXP SC /
Samsung DVD / Lite-On CDRW
May 22, 2003 6:30:01 PM

i think anyone buying new technology is a retard. It's you're own fault. Never buy new technology. wait 6 months. Espeically motherboards! Canterwood is NEW technology so of course it's going to be buggy and not perform well at first. Thats with all new techs! Give it 6 months and it will be a dominating platform with prescott.

I don't even think Intel expected anyone to fork up 250 dollars for a motherboard! Keep doing that and eventually all motherboards will be 250. so stop paying 250 for a motherboard! thats insane!

I can see paying for the gigabyte ultra with it's onboard SCSI 320 for 300 dollars and thats as far as anyone should go!



"Bread makes me poop!" - Special Ed

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
May 22, 2003 6:56:29 PM

This maybe true.... but only for now... we have yet to have a full understanding all of the performance benefits that lie in the future....

There is no Spoon....
May 22, 2003 7:10:15 PM

I'm sure both boards will see another 5% or more improvement as new BIOS releases come out.
May 22, 2003 7:50:43 PM

Again, and I've repeated it many times, according to THG's benchmarks, YOU ARE NOT GETTING <b>5%</b> MORE, YOU ARE GETTING <b>0.06%</b> ON AVERAGE.
Only one benchmark had a suitable 5%.
If you're such a dope to love zipping stuff and you need 5 seconds less, then by all means, zip on.

This is based on THG, now if all other sites used Asus P4P800 as well and got much less, and the setups were similar, then consider all I said, to be a self-humiliation, to which I'll concede having been uninformed.

Otherwise, no one IMO is being right in their mind if they really wish to shell out 40$ more for 0.06%.

--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue:
May 22, 2003 11:15:24 PM

Yes, You are right, that Canterwood owners are screwed. But I don't think Intel ever said the Canterwood is a mainstream Chipset. It was always a workstation chipset. So the people who bought the Canterwood to play around with it are sure to be screwed since they could have waited another month to get the much cheaper Springdale.

KG

"Artificial intelligence is no match for natural stupidity." - Sarah Chambers
May 22, 2003 11:18:56 PM

Quote:
This is based on THG, now if all other sites used Asus P4P800 as well and got much less, and the setups were similar, then consider all I said, to be a self-humiliation, to which I'll concede having been uninformed.

Dude, you don't have to go that far. :)  Humiliated over a springdale/canterwood debate? I think there are a lot more serious matters going on in the world than this. ;) 

The percentage difference depends on which reviews/articles you look at. From what i've seen, Springdale definetly is recommendable. Both good chipsets. I'm always open to change and considering everything you've said thus far. I would've waited for more reviews and articles before trying to start a <i>springdale revolution</i>. But, it's great that this is being openly discussed.;)
May 23, 2003 1:29:09 AM

But if it was a Workstation chipset, WHAT would differentiate it, assuming current performance is NOT noticeable at all?
It boasts all that Springdale does, and costs more, yet its performance according to THG is not there.

OF COURSE, you COULD choose to have the "It's gonna perform better eventually" way, but who here really wants performance LATER than now?
Are we to buy into the nVidia crap and go get the FX cards, in hope the Detonator FX would come out sooner or later, and during that time we just spent on inflated price crap with no performance? Of course that is an example of back then, not now anymore.

I guess what I am saying is, if the current Canterwood doesn't perform any higher than 3% on average, then shelling out so much money, when even the MSI Neo self-overclocks to give you an easy 7% boost across the board for less, is pointless.
Unless Intel does promise a significant performance update revision driver.


--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue:
May 23, 2003 3:06:20 AM

I find this very interesting. When Canterwood and Springdale where being reviewed Canterwood was the better Board Springdale had less. Canterwood offered the best for the buck. Now all of a sudden Springdale is released an its as good as Canterwood. So how does the Cheaper Chip Springdale better than the top chip Canterwood?
May 23, 2003 6:05:00 AM

Its all good. You spent a few bucks to get it first. Is it an amazing board or what. Do you want your old 533 board back. Have you already tweaked it past what toms is showing. So you put a few more sheckles in Inel's pocket.
Now they can better battle the infidel Amd. Good job!!
a b à CPUs
May 23, 2003 7:01:22 AM

d00d, 4.5% is HUGE when you consider the price difference is less than $20!

<font color=blue>Watts mean squat if you don't have quality!</font color=blue>
May 23, 2003 9:50:18 AM

i865PE rocks!!! This is the only thing I could say after all reviews I've read. Based on current situation, those who were sooooooo impatient and rush for i875P can only say, "Hey, my Canterwood beat your Springdale in price! How's that!" And in fact, I found that Asus P4P800 is actually better feature than P4C800, so i875P buyers have been screwed by no one else other than themselves.

It is morally wrong to allow suckers to keep their money
<A HREF="http://www.anandtech.com/mysystemrig.html?id=22996" target="_new">My System Rig</A>
May 23, 2003 11:22:36 AM

Yes, if you want to have such in rare situations. And where do you see 20$ difference, ON AVERAGE around the web?

--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue:
May 23, 2003 2:35:12 PM

Well ive just read the hexus review of the asus springdale board and ive had my mind changed completely. I was set to buy a P4C, but the price was way too high. Here comes springdale and outperforms the P4C or atleast matches it and costs £50 ($75) less here in the uk! What a surprise.
May 23, 2003 5:17:17 PM

I think the difference between comparable Springdale and Canterwood motherboards of the same brand is closer to $50.

Anandtech.com compared the IC7-G to the IS7-G. I'd more likely be looking at something like IC7 or IC7-E (although your mention of timing issues would have me worried) vs IS7, so either a $30 difference or a $50 difference. However this is all beside the point.

I should have qualified my statement. Canterwood is not worth it to me because 1> I can't afford the top 800 Mhz processors (not even the bottom one.) 2>There is no apparent gain for 533 Mhz processors on Canterwood.

Questions> How does overclocking factor in? Since Canterwood needs 800 Mhz to show gains over Springdale how does this bode? Are the 800 Mhz CPUs as overclockable as their predecessors?

The answer would sway my opinion one way or the other.

Looking at it from another point of view however (not everyone is as cheap as me), if you are going to buy a $425 processor and maybe $200+ for a motherboard and perhaps $200 for memory then $20-$50 more (<b>EDIT-></b> for Canterwood over Sprindale) for a 4.5% performance boost would be a bargain.




<b>56K, slow and steady does not win the race on internet!</b><P ID="edit"><FONT SIZE=-1><EM>Edited by phsstpok on 05/23/03 01:50 PM.</EM></FONT></P>
May 23, 2003 9:35:50 PM

The new 800fsb P4 cpu's overclock just as well if not better than the 533 versions. Its well known and been proven that the 2.4c will reach anywhere between 3.2 and 3.6ghz! The FSB of both canterwood boards and springdale are stable upto about 275mhz so a good choice would be to run a 2.4c on 260-270mhz at 3.1ghz+ or if you want to run in sync with 466 memory then a 2.8c on 225-235mhz should reach 3.2-3.3ghz. JFYI im getting a 2.8c and 512mb of pc3700 to run on the asus springdale board.
a b à CPUs
May 23, 2003 10:45:20 PM

The IC7 cost lass than $30 more than the IS7. Look at Newegg for example. The chipset makes up $20 of that difference.

<font color=blue>Watts mean squat if you don't have quality!</font color=blue>
May 24, 2003 3:22:47 AM

Canterwood is better for OC. Springdale good board for the non OC'er
May 24, 2003 3:55:49 AM

Congrats on your success with Springdale but now you have renewed my interest in Canterwood.

If Canterwood has a small advantage at 200 Mhz over Sprindale does the advantage improve at even higher speeds?



<b>56K, slow and steady does not win the race on internet!</b>
May 24, 2003 8:54:02 PM

What in the world are you talking about?
Springdale and Canterwood BOTH need to run at 200MHZ to get the whole benefit, and at such FSB speeds, both are tied according to many reviews. Why would that renew your desire for it?

--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue:
May 24, 2003 11:15:26 PM

I already gave a link showing the IC7-G outperforming IS7-G at 200 Mhz, by 4.5%, but equal in performance at 133 Mhz. I left the question open on whether things improve with overclocking. Here's one review that leans in that direction.

<A HREF="http://www.ocaddiction.com/reviews/mobos/abit_is7/" target="_new">http://www.ocaddiction.com/reviews/mobos/abit_is7/&lt;/A>

Since, the 200 Mhz P4s are still good overclockers I would be interested in the platform that performs the best when overclocking, even if the margin is small. If not today then in the future.

Granted, we need to see more overclocking comparisons with a broader range of production mobos.


<b>56K, slow and steady does not win the race on internet!</b>
May 25, 2003 4:22:45 AM

Wasn't the Canterwood the one who was prone to a lot of picky timings?

--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue:
May 25, 2003 5:58:19 AM

No. Anand mentioned in his Springdale review that the IC7 was quite picky with memory while the Canterwood board they tested worked fine with the Corsair PC3200 memory. Remember, Springdale chipsets are really just Canterwood chipsets that failed to make the grade.

"We are Microsoft, resistance is futile." - Bill Gates, 2015.
May 25, 2003 1:59:46 PM

Yes however PAT seemed to really add strain on memory timings.

Anyways, I could be wrong, or maybe I am mixing up my history lately.

--
If I could see the Matrix, I'd tell you I am only seeing 0s inside your head! :tongue:
May 25, 2003 11:12:58 PM

PAT has very little to do with the memory itself. The memory controller is aggressively timed. The memory itself still has the same cas/ras/page latency.

"We are Microsoft, resistance is futile." - Bill Gates, 2015.
!