Hi Paullie -
To respond to your comments Re: AMD and Intel .. I've got a bit of knowledge in that tech arena so I'll shed whatever light I can ..
In the near future 64-bit computing will be brought to the consumer level. It has been a reality for many a year for enterprise computing but it's not in the hands of the average business yet (I'm talking 64-bits system-wide bus width, not just 64-bit on the graphics bus, which has been around for a little while, even on game consoles).
Intel announced its 64-bit "Merced" solution quite some time ago, and after reading white papers and technical notes on it, my opinion on it is .. it sucks. I honestly can't believe that the engineers there made such a bonehead play out of what was obviously a great opportunity to be the company that brings the world up to 64-bits. AMD's "Sledgehammer" solution is superior for two big reasons:
1) better backward compatibility - existing 32-bit x86 instructions (IA-32) are a subset of the Sledgehammer's 64-bit instruction set. The Merced must switch modes and effectively emulate IA-32. So users running existing apps on a P4-1GHz upgrading to a Merced-1.5GHz probably won't see any speedup at all (according to Intel reps) because the emulation is slower clock-for-clock.
2) less software work - the Merced has a radical new instruction architecture which relies on the compiler being really really good at figuring out the parallelism in a program. please take this with a grain of salt as i only mention it anecdotally to further illustrate my opinion, but .. a friend of mine who knew people in the Microsoft compiler group said that even though they've been working with Intel developing the compiler for their chip for years, they are still totally clueless as to how they are supposed to meet the demands of the chip. AMD's new 64-bit architecture will require compiler work as well, but since it wasn't designed to rely on the compiler for certain optimizations, it won't be nearly as complex to develop.
This reasoning (combined with announcements about Athlon) had me pounding the table to buy AMD 18 months ago at $10 (split-adjusted) I feel like I have great ideas but terrible timing, because the stock dropped to $8 before it climbed to $15, where I sold (too soon, but hey I made 50% in 6 months which ain't bad) Why did I sell? Well, in order to succeed, AMD had to buck a fierce trend, which I didn't see them doing. AMD's tech was fantastic, but everything else about the company flashed warning lights - supply problems, lack of OEM support and OEM distribution deals, extremely aggressive pricing, and past inability to turn a profit even on strong sales. They had a lot of hurdles to overcome. But they succeeded far beyond my expectations, so kudos to them.
Intel is just so entrenched in the market that it's difficult for AMD to make a dent. But I'm bullish on them now because the 64-bit space is a new market. Yes, Intel can bring to bear its distribution, marketing, and fabrication prowess to sell the Merced, but the Athlon gained an awful lot of ground over the P3. Who knows how Intel will fare in a market they don't already dominate?
64-bit, folks, that's where it's at. And even though AMD might release the Sledgehammer after Intel releases its Merced .. If businesses are upgrading to 64-bit systems, I don't think they'll be lining up around the block to buy an overpriced system that gives them a negligible performance boost.
I had contemplated rotating from Intel to AMD over the past few weeks .. lost some good profits on Intel but at least I got AMD at a bargain.
Good luck to all ..
-G |
