IDC, isn't it also true that as we move further along each node becomes more expensive to develop? If this is the case I would expect Intel's lead to eventually diminish.
Intel's lead is getting larger, not smaller. Simply because they have more money than everyone else.
IDC, isn't it also true that as we move further along each node becomes more expensive to develop? If this is the case I would expect Intel's lead to eventually diminish.
It is true, development costs rise with each successive node, as do the production costs of the node once it goes into production.
But this doesn't ensure that Intel's lead will diminish, just means their own rate of node cadence stands to slow down. But it will slow down at a slower rate than that of their competitors.
So the gap will continue to just get larger.
Makes you wonder what the CPU industry is going to look like in 10, 15 years...
But this doesn't ensure that Intel's lead will diminish, just means their own rate of node cadence stands to slow down. But it will slow down at a slower rate than that of their competitors.
So the gap will continue to just get larger.
GlobalFoundries disagrees.
As you can see, GF expects risk production to start on 22 SHP exactly 2 years after it started on 32 SHP.
If we compare that to Otellini's recent comment about intel "beginning production by the end of the year" and their slower than normal showcasing of 22nm (this month compared to their normal February timescale), we can see that while intel has almost certainly slipped on 22nm, GF should close the gap to just over a year assuming nothing goes wrong.
Process technology is interesting. But it is also very dangerous.
When you rely on process technology for your advantage, you rely on a something that is a.) far more difficult and b.) far more expensive, with each generation.
Look at today's 32nm Xeon vs. 45nm Opteron. About the same performance. About the same power. AMD is lower in price. As a customer, what is the 32nm benefit?
AMD's advantage is in design. ARMs advantage will be in design. Intel will continue to push process, but their challenge will be that each corresponsingly smaller node will be more expensive and harder to control.
If I had to bet on long-term strategies I would rather not bet on the thing that becomes more difficult each year.
If I had to bet on long-term strategies I would rather not bet on the thing that becomes more difficult each year.
Why is it then that GloFo is attempting to pull-in 22nm SHP for AMD to close the gap to Intel's 22nm timeline?
You mean like attempting to make a career out of selling server gear for a company that steadily loses server market share?
You mean like attempting to make a career out of selling server gear for a company that steadily loses server market share?
Why is it then that GloFo is attempting to pull-in 22nm SHP for AMD to close the gap to Intel's 22nm timeline?
JF's point was that AMD is capable of matching intels performance and power draw on a larger process without HKMG. It shouldn't be difficult to figure out what could happen to intel if those FinFET's cause serious yield issues at 22, leaving the current Xeons vs BD. That 30% share AMD had a while back would be under threat pretty fast.You mean like attempting to make a career out of selling server gear for a company that steadily loses server market share?
You mean like attempting to make a career out of selling server gear for a company that steadily loses server market share?
Process technology is interesting. But it is also very dangerous.
When you rely on process technology for your advantage, you rely on a something that is a.) far more difficult and b.) far more expensive, with each generation.
Look at today's 32nm Xeon vs. 45nm Opteron. About the same performance. About the same power. AMD is lower in price. As a customer, what is the 32nm benefit?
AMD's advantage is in design. ARMs advantage will be in design. Intel will continue to push process, but their challenge will be that each corresponsingly smaller node will be more expensive and harder to control.
If I had to bet on long-term strategies I would rather not bet on the thing that becomes more difficult each year.
You mean like attempting to make a career out of selling server gear for a company that steadily loses server market share?
That's just brutal lol.
But you don't have the same performance profile. You only have roughly comparable throughput if the customer uses enough threads to push every core. Otherwise, the Xeon is going to be faster. The fewer the cores needed to hit a certain theoretical throughput, the better that CPU is going to be in actual use.Look at today's 32nm Xeon vs. 45nm Opteron. About the same performance. About the same power. AMD is lower in price. As a customer, what is the 32nm benefit?
I wouldn't bet on the strategy that requires twice the cores and three times the die size to match the theoretical throughput of my competitors products and a strategy that my competitor could easily adopt and leave me with no other response... than maybe to slow down my cores further and double up their count again?If I had to bet on long-term strategies I would rather not bet on the thing that becomes more difficult each year.