Web journalism is here to stay - Roy Greenslade, Guardian Online
From what we hear, it is, or will be shortly after launch.
All the news passing our way says that the trigger was pulled on the September 10th launch at 2.0GHz when things looked quite bleak for AMD. Someone decided that something was better than nothing, most likely before they knew what the next stepping would do.
So, the launch speeds are going to be slow, 2.0, maybe a little more. The up side is that the chip is going to ramp very quickly, almost 50 per cent before the New Year. AMD won't make 3.0, but that won't take much longer. When we said high 2s, we weren't kidding, just push things out six months.
In any case, the stepping after the launch part turned out good enough to run with, but by the time that was known, there were enough of the previous stepping parts in the oven to force the launch. When volume ramps and the consumer parts hit, speeds will be back in the game.
Keep an eye out for low volume sample parts of the next stepping, that will point the way much more than launch clocks. If you see higher binned parts floating, well, those are likely from the next stepping, and will be more in line with what you will get in Q4.
By the end of the year, I would not expect AMD to be ahead, a six month delay has put it against the Penryn cores instead of Woodcrest, but it should have enough CPU power to be competitive. From that point, it become a matter of which can ramp faster, and which chip scales better.
The enthusiasts will win in the end, there will more than likely be XE/EE/FX/Black/Blue/Pink parts enough to satisfy all by the non-demominational year end festivities. It is about time this game got interesting, a complete Intel rout is far less interesting than a pitched battle. µ