Not only is the the first billion transistory chip but it should also bring the dying breed back into action and retake the server throne, or so it would seem. Surprisingly with all those transistors inside the chip still manages to consume 20% less power yet offers over 1.5x the performance, and 2.5x the bandwidth of the current Itaniums, sounds good in my book. http://www.newsfactor.com/story.xhtml?story_id=131000030T9I
I'd prefer my cluster of Opterons, Itaniums are good but you can't say they're good until you see the price/performance ratio.
Good point but the current gen is no slouch and a couple of these dual-core chips should give some pretty good results, but yeah dual-core opterons are quite awesome. We'll see.
http://www.buy.com/retail/product.a...92913&Type=PE&Category=Comp&Gad=0&dcaid=17379 I'm sure we could buy a few more Opterons for the price than that. Remember RAMBUS? It was high performance and very expensive, not suitable for the market so it failed.
Holy mothers, I had no idea, I can hardly believe it's that bloody expensive, yeah in that case I can't imagine the chip going anywhere, even in big deal firms that can afford it with AMD as a viable alternative. I hope that price is way wrong for everyones sake.
While the price may be part of it, the Itainium isn't in the same CPU class as what you're familiar with. It's not an x86 chip, and to make it work, you need a different coding. The problem is, most people don't want to do this (and justifiably so). The x86 architecture may be the popular one, but that doesn't mean it's the best. Realistically, we'd need to see the industry flip over to something else. AMD's x86-64 technology is a good move, but it's not an entirely new architecture. Until x86 hit's it's proverbial brick wall, I don't see another architecture being seriously considered for widespread use. The Itainium may be pricey, but if it were used en masse, the price would drop due to supply and demand (unless Intel ran into snags with high production numbers).
Yes, I'm afraid Intel may have gone too far into left field on this one, assuming they'd simply move the market in a new direction with or without consumer cooperation. What they didn't wager on was the huge success of the Opteron in the existing x86 market, which provides awesome performance for the money. Nobody wants to code for multiple platforms if they don't have to, so the x86-64 crossover was a killer move for AMD -- move to next-gen, maintain backwards compatibility. Intel provided too little too late with their x86-64's, and the Itanium has been aptly dubbed "Itanic". Sucks to have your monopoly threatened, I guess. -AT
I know x86 isn't the best arhictecture, and its surprising the period of time its managed to survive, but to be a successful you need to be able to meet the demands of the current market, while progressing into new ones. Its not easy to suddenly drop the current hardware/software programming architecture for something completely new, even if it is better in some ways. The problem with Itanic(oops) is that there is almost no niche in the market for them. I can't see how they can change the current architecture quickly and smoothly as possible myself.
