Spitfire_x86

Splendid
Jun 26, 2002
7,248
0
25,780
<A HREF="http://www.theinquirer.net/?article=10125" target="_new"> Read @ the Inquirer</A>

Something called "Thorton" exists? x86 secrets has removed the roadmap from their website

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new"> My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new"> My Rig</A></b>
 

slvr_phoenix

Splendid
Dec 31, 2007
6,223
1
25,780
Something called "Thorton" exists?
Quoted from the ever-so-intelligent folks at <i>the inquirer</i>. "<font color=red>And the mysterious "Thorton" core, which AMD has publicly denied, has gone and made its appearance all over again - a hybrid Thoroughbred and Barton crossed.</font color=red>"

Why that's bloody genius, inquirer! Cross a Thoroughbred with a Barton! What do you get? An Athlon XP with a 384KB L2 cache??? What's the point of that again?

The point, should the folks at <i>the inquirer</i> have bothered to do <i>any</i> research whatsoever and given even a half-hearted explanation of Thorton, is that Thorton will be the new core for the Duron. It will replace the 256KB L2 AXPs as in the future all new AXPs manufactured will have 512KB L2. The 256KB L2 AXPs will no longer be produced. This will supposedly be happening when AMD moves to 0.09 micron for the AXP and the A64. Thorton may even be stuck at 0.13 micron.

So Thorton is really just a re-branding of the 256KB L2 AXPs as Durons, which is a far cry from <i>the inquirer</i>'s "<font color=red>a hybrid Thoroughbred and Barton crossed</font color=red>".

Good grief the people at <i>the inquirer</i> really suck at reporting.

x86 secrets has removed the roadmap from their website
<A HREF="http://www.x86-secret.com/index.php#596" target="_new">Silly peoples. Did they really think that they could get away with posting that?</A>

"<i>Yeah, if you treat them like equals, it'll only encourage them to think they <b>ARE</b> your equals.</i>" - Thief from <A HREF="http://www.nuklearpower.com/daily.php?date=030603" target="_new">8-Bit Theater</A>
 

sabbath1

Distinguished
Apr 4, 2003
460
0
18,780
is that Thorton will be the new core for the Duron.

Oh, that's true? Sounds cool, a new el-very-cheapo but fast offering from AMD. The Celeron has indeed been a pale comparison. If only AMD were as competitive in the high-end right now, they'd be much better off...than what they are now ;)


My system: Intel Pentium 4 2.8, 800 FSB / TwinMOS 1Gb DDR400 / ASUS P4P800 / Sapphire Radeon 9800 Pro / Antec True Power 550W / Western Digital Raptor / Hercules G.T XP /
Samsung DVD / Lite-On CDRW
 

slvr_phoenix

Splendid
Dec 31, 2007
6,223
1
25,780
Oh, that's true? Sounds cool, a new el-very-cheapo but fast offering from AMD.
Well if the A64 becomes AMD's high-end desktop PC model, that'll squish the AXP in the middle end at best. But because they're not technically the same market (one being pure 32-bit and one being x86-64) that will cause some overlapping. So AMD will likely keep the AXP around for some time to come to appease the people who don't want 64-bit. The AXP will be on it's 400MHz FSB with it's 512KB of L2 (and possibly get even better when the 0.09micron shrink comes ... but not likely as AMD will want to push the superiority of the A64) and thus AMD will be left with no real 'low end' processor anymore. To sell the AXPs as 'low end' will really hurt them because it'll lower the prices too much.

The solution is to make a new Duron, which is where Thorton comes in. Take the AXP, possibly restrict it to a 333MHz FSB and 256KB of L2, and call it a Duron. Keep their speeds low and use as cheap of a manufacturing process as you can and sell <i>them</i> as the low-end where AMD makes almost no profit. Now the AXPs can stay as high end and even if AMD loses money on a few of the lowest-rated Thorton Durons they'll still more tham make up that difference with the higher AXP prices that most people will be buying.

I think Thorton is really more of a business strategy than a CPU. ;) But that's how I understand it will work, and it's a good plan. AMD needs to bring back the Duron, but in such a way that it doesn't convince most people to use that <i>instead</i> of the AXP.

The Celeron has indeed been a pale comparison.
I agree. I think Intel has gone overboard in making sure that the P4 Celerons suck. They really could boost the Celeron performance significantly, or else lower the price for the Celerons to the same as around the AXP1700.

If only AMD were as competitive in the high-end right now, they'd be much better off...than what they are now ;)
AMD's real problem right now is that their high-end is so intermingled with their low-end that the low-end is dragging down their possible profits from the high-end. The end result is this really weird situation where the high-end is overpriced for what it is worth and the low-end is underpriced for what it is worth, causing AMD to lose out on a lot of high-end sales and lose a lot of high-to-middle sales out to their low-end products where they make much less money on.

What AMD really needs to do to put themselves back on track financially is to find a way to restructure their product line where their low-end chips aren't good replacements for their middle-end chips and thus their middle-end chips can cost more so that their high-end chips can be lowered in price slightly. Then AMD will be making a <i>lot</i> more money.

Take a look at Intel's P4 Celerons for example. Yes, they're overclockable. However even when you OC them you get trashed by a real P4. And even the P4Cs are very much worth buying over the P4Bs. So Intel can set prices where they make a good profit from their whole range of CPUs except for the Celerons and set it up so that most people won't want the Celerons. It's a very good profit-making layout even if Intel loses most of their low-end sales to AMD because of it.

For AMD to be better off they really have to learn to offer a low-end product that won't screw up their midrange sales and will allow them to keep good profit-making prices on their midrange products. Having a good high-end product is nice, but it usually has a much smaller impact on their profit than how their low-end is set up. Somewhere along the way AMD completely lost sight of this and put themselves into the very bad position that they're in right now. They're almost completely mired into selling midrange products at low-end prices. It's good for us consumers, but it's really bad for their finances. And really bad for their finances means much less competition at the high end and fewer advances in technology, which in the long run is bad for them and us both.

"<i>Yeah, if you treat them like equals, it'll only encourage them to think they <b>ARE</b> your equals.</i>" - Thief from <A HREF="http://www.nuklearpower.com/daily.php?date=030603" target="_new">8-Bit Theater</A>
 

Spitfire_x86

Splendid
Jun 26, 2002
7,248
0
25,780
You're correct, I was wondering- how it's possible to make a processor combining T-bred and Barton

If AMD decides to put SSE2 on Thorton, it would be very good. But this not going to happen, most possibly.

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new"> My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new"> My Rig</A></b>
 

reever2

Distinguished
Apr 13, 2003
231
0
18,680
What I dont get is the Thorton has been said to have the same die size of the barton, but still have 256kb of L2 cache, and still have lower clock speeds than their equivalently PR'd tbred models?
 

slb132

Distinguished
Mar 21, 2003
116
0
18,680
What I think they mean by a hybrid Thoroughbred and Barton crossed is only in name. Thor (Thoroughbred) ton (Barton) = Thorton. Just a combination of both names Thoroughbred and Barton into Thorton.

DFI i815EP, Intel Pentium III (Tualatin) 1.42GHz, ATI Radeon 9100, SoundBlaster Audigy 2 Platinum, Samsung 512MB SDR SDRAM, Maxtor 60GB 7200RPM 2MB Cache
 

ChipDeath

Splendid
May 16, 2002
4,307
0
22,790
simple. Manufacture it using the same process, but disable half the L2 cache. Less chips get rejected for dodgy cache, as 50% of it is out of the equation, and besides - isn't this part of the way the P4 Celerys are made?

---
:smile: :tongue: :smile:
 

TRENDING THREADS