Sign in with
Sign up | Sign in
Your question

Intel makes Ant Hill an Opteron competitor

Last response: in CPUs
Share
April 27, 2003 2:49:01 AM

<A HREF="http://www.theinquirer.net/?article=9156" target="_new"> Read @ the Inquirer</A>

----------------
<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>
April 27, 2003 3:14:34 AM

Nothing in that article points me to any facts about Anthill or what it is, or that it is competitive with Opteron
April 27, 2003 8:39:07 AM

From what i know it just some Xcale in X core as it very small.

[-peep-] french
April 28, 2003 2:14:50 PM

Sheesh. Talk about 'the inquirer' making a mountain out of a mole (or ant) hill. Their own statement best sums up their article:

"<font color=red>The only thing not in doubt is that the whole thing smells fishy. µ</font color=red>"

<font color=blue><pre>I'm proud to be an American,
who served my country in the US Air Force,
to protect the rights of my fellow Americans,
to hold protests against others like me.</pre><p></font color=blue>
May 4, 2003 4:37:38 PM

Still seems likely in my mind that Intel will release their own x86-64 (or similar) CPU.

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 4, 2003 11:59:54 PM

after a good read around the net i have figure this is just a joke.

[-peep-] french
May 5, 2003 1:56:43 AM

What's interesting to ask, is what else can you do that can make your 64-bit X86 different than AMD's, really?
Either way you extend register size, so what else can you do to programmability and keep it x86 64-bit?!

Kind of weird if you ask me, on Intel's part.

--
This post is brought to you by Eden, on a Via Eden, in the garden of Eden. :smile:
May 5, 2003 11:05:15 AM

x86-64 is a really good idea and if AMD earns a lot of money because of Opteron, then I believe Intel will follow suit and release a similar product. (It's not gonna be identical, that's for sure.)

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 5, 2003 11:34:28 AM

I am repeating myself when saying this, but again, what possible extension can they develop for x86?
AMD's x86-64 is a logical evolution, not a revolutionary architecture.

--
This post is brought to you by Eden, on a Via Eden, in the garden of Eden. :smile:
May 5, 2003 11:42:18 AM

I don't know. I'm just saying the two CPU's probably won't be identical. The x86-64 parts of Intels and AMDs CPUs would probably be 'very' similar. But what I meant was that Intel probably will change something else, that doesn't have to do with x86-64 at all, since like you say, 64-bit is 64-bit and nothing more.


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 5, 2003 1:36:27 PM

For compatibility reasons, they would have to use the x86-64 extension. Their implementation is bound to be different but the ISA extension would be the same. Otherwise it's pretty pointless. Why not take advantage of all the developer support AMD has managed to muster up?

"We are Microsoft, resistance is futile." - Bill Gates, 2015.
May 5, 2003 3:20:09 PM

Sure, but I still believe for prideful reasons, that Intel will make the eventual Yamhill/Anthill/etc. a somewhat different product than what the Opteron/Athlon 64 is.

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 5, 2003 4:24:06 PM

They may market it differently and again, the implementation would be different of course, but making the ISA intentionally different (and hence, incompatible) with x86-64 would just be technically stupid. But then again, it's not like Intel would never do something technically stupid (*cough* Willamette Celeron *cough*).

"We are Microsoft, resistance is futile." - Bill Gates, 2015.
May 5, 2003 5:30:27 PM

Wouldn´t it be possible that Intel´s Yamhill implementation, if such a thing exists, has its own different specs, but is compatible with AMD´s x86-64? That way, it can run x86-64 written code and run its own code - which x86-64 can´t. Just one small thought of mine...

Then again, like Eden said, there´s little to do differently, I suppose...<P ID="edit"><FONT SIZE=-1><EM>Edited by Mephistopheles on 05/05/03 01:33 PM.</EM></FONT></P>
May 5, 2003 8:13:58 PM

What in the world could make another x86 processor different in its 64-bit ISA?

--
This post is brought to you by Eden, on a Via Eden, in the garden of Eden. :smile:
May 5, 2003 10:30:02 PM

x86-64 64-bit add:
00110101 00000001 00000002

Yamhill 64-bit add:
01101010 00000001 00000002

Both instructions do the same thing (there is no logic to it, you specific the sequence that counts as a 64-bit add instruction and the hardware decoder recognizes that sequence because it was pre-programmed into it), they're different instructions. It's like speaking another language. You're still saying the same thing, just in a different way.
If Intel were to make an x86 extension and make it x86-64 compatible but add more, it would have to add BOTH x86-64 instructions AND its own extra set of extensions. AND to be compatible with x86-64, it would have to have an *exclusive* x86-64 mode. Unless they wants to make an extension to x86-64. On top of that, they have to add whatever instructions they want. And get developer support. It's a lot of extra work, a lot of die space and unless their additions actually provide functionality, no sane engineer would agree to do it.

"We are Microsoft, resistance is futile." - Bill Gates, 2015.
May 6, 2003 2:44:19 AM

So what you're saying is, Yamhill is no longer x86 in itself, but something like an attempt to diverge into another architecture like IA64?

--
This post is brought to you by Eden, on a Via Eden, in the garden of Eden. :smile:
May 6, 2003 6:24:08 AM

No, what I'm saying is, the instruction sequence wouldn't neccessarily be the same. That's why it's an x86 *extension*. Think of it this way, x86 has a set of instructions, say add, sub, mul, div. You make an extension of the ISA, let's call this x86-SSE-SIMD-64-XP-SE or whatever marketing name they decide to call it. What is it? You add new instructions that can be used. Let's say padd, psub, pmul, pdiv. They work on new data types, packed 128-bit vectors of 32-bit integers. Hence, we have an x86 extension. If Intel were to make Yamhill and make it different than x86-64, then it would not use the same instructions as x86-64 (or at least, have instructions that x86-64 doesn't have). x86 an instruction set, x86-64, MMX, SSE/SSE2 are all extensions which add *more* instructions. Could an x86 extension potentially have instructions of the format of IA-64? Sure, but you'd have to switch to another mode (to tell the processor which format type to use). Itanium 2 currently has a similar model. You can switch to x86 mode or you can switch to IA-64 mode. Opteron requires you to switch to x86-64 mode as well, but when you're in x86-64 mode, you can still run legacy protected mode software. That's the difference between an extension (in which you have the x86 decoders active, but simply add more sequences that it recognizes) vs a separate ISA.

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