Well, I'm not sure how "reliable" people think <A HREF="http://www.siliconstrategies.com" target="_new">Silicon Strategies</A> is as a news source, but there's some rather interesting news in it.
First, I knew that Intel could outspend AMD in capital expenditures, but I didn't think it was almost 7 times as much. AMD is happy to announce that their capital expenditures budget is going up 20% to $850 million, but Intel, while <i>cutting</i> their's back 24%, is still up at $5.5 <i>billion</i>. Capex is fab production and tools, etc. R&D at AMD is going up to $700 million.
Also in the same article, the confidence of AMD is apparent. Talking about the Hammer line in 2003, Jerry Sanders flat out said that "'we will break away from Intel with massive performance increases.'" They seem rather confident in their performance abilities. He also went on to say about 2002 that "'the absolute frequency difference won't widen, although our (top-of-the-line) Athlon processors will still have the (equivalent) performance level as Intel's (top-of-the-line) Pentium 4.'" Also mentioned was that SOI was proceeding as planned and yields were in the high 90's for current processors. I like AMD's attitude right now. They are looking at the fact that their earnings went through the roof (expecting to be profitable by 2002Q2) and their market share has been growing as a sign that they are winning the battle right now. Article at <A HREF="http://www.siliconstrategies.com/story/OEG20020116S0075" target="_new">http://www.siliconstrategies.com/story/OEG20020116S0075</A>
Also in AMD news, this little tidbit:
In Intel news, they cut about $1 billion from their capex budget because they feel that by going to 300mm wafers, they can save a lot of money and invest less in capex for fab costs with that migration. Sounds good, and maybe they will be able to compete with AMD on a price basis now. Article at <A HREF="http://www.siliconstrategies.com/story/OEG20020116S0039" target="_new">http://www.siliconstrategies.com/story/OEG20020116S0039</A>
Last, DDR333 news. Things aren't looking so bright, and JEDEC is at odds of how to package the new DDR333. Some would like to keep using the current process, saying that it still works, although some fear that timing issues might develop in systems with a large number of sticks. DDR II looks to be using µBGA packaging (as opposed to the current TSOP packaging), and some in JEDEC would like to see manufactures move to µBGA with the DDR333, and possibly DDR400, so that the chipmakers are ready for DDR II. Also, the other issue for DDR333 is that without Intel having announced support (there is an enhanced i845-D with 333MHz support to go along with the 533MHz Northwood in the works, according to JEDEC papers) lack of OEM demand for DDR 333 might keep it at as a novelty for a while, something that only a small segment of the public and private sectors would use; ergo, prices would be high. Also, is Intel is slow to adopt DDR333 support, and submits it's specs for DDR II not long after, DDR333 might not last too long, as there will be a quick move to DDR II specs by all JEDEC companies. Article at: <A HREF="http://www.siliconstrategies.com/story/OEG20020117S0062" target="_new">http://www.siliconstrategies.com/story/OEG20020117S0062</A>
Well, that's what I've found, and I find it all very interesting. I find it sad that Intel pretty much holds the fate of DDR333 and DDR II in it's hands, but I think it would be best for them to offer the maximum amount of options to consumers; ergo, having announced support for all availible memory types, until RDRAM or DDR/II/III wins out.
Comment please, either in this thread, or in a new one. I want to know what people think. I have on other article, but it's for a different thread.
-SammyBoy
First, I knew that Intel could outspend AMD in capital expenditures, but I didn't think it was almost 7 times as much. AMD is happy to announce that their capital expenditures budget is going up 20% to $850 million, but Intel, while <i>cutting</i> their's back 24%, is still up at $5.5 <i>billion</i>. Capex is fab production and tools, etc. R&D at AMD is going up to $700 million.
Also in the same article, the confidence of AMD is apparent. Talking about the Hammer line in 2003, Jerry Sanders flat out said that "'we will break away from Intel with massive performance increases.'" They seem rather confident in their performance abilities. He also went on to say about 2002 that "'the absolute frequency difference won't widen, although our (top-of-the-line) Athlon processors will still have the (equivalent) performance level as Intel's (top-of-the-line) Pentium 4.'" Also mentioned was that SOI was proceeding as planned and yields were in the high 90's for current processors. I like AMD's attitude right now. They are looking at the fact that their earnings went through the roof (expecting to be profitable by 2002Q2) and their market share has been growing as a sign that they are winning the battle right now. Article at <A HREF="http://www.siliconstrategies.com/story/OEG20020116S0075" target="_new">http://www.siliconstrategies.com/story/OEG20020116S0075</A>
Also in AMD news, this little tidbit:
They seem to like the fact that the die size is smaller than Intel's, but the drawback I see is that with a smaller die, there is less surface area for die-heatsink contact, meaning that the heat won't transfer as well. Let's hope that the .13µ process reduces heat output that much, or that they put a small heat spreader on to. Article at <A HREF="http://www.siliconstrategies.com/story/OEG20020116S0064" target="_new">http://www.siliconstrategies.com/story/OEG20020116S0064</A>AMD also reported that it has started production of AMD Athlon processors on 130-nm (0.13-micron) technology in Fab 30 in Dresden, Germany, during the quarter.
Sanders said the Athlon XP processor die, produced on 180-nm (0.18-micron) technology, is 12% smaller than the recently introduced "Northwood" version of Intel Corp.'s Pentium 4 produced on 130-nm processes. "The AMD Athlon XP processor on 130-nm technology is just over half of the size -- 45% smaller -- of the Northwood die," Sanders said.
In Intel news, they cut about $1 billion from their capex budget because they feel that by going to 300mm wafers, they can save a lot of money and invest less in capex for fab costs with that migration. Sounds good, and maybe they will be able to compete with AMD on a price basis now. Article at <A HREF="http://www.siliconstrategies.com/story/OEG20020116S0039" target="_new">http://www.siliconstrategies.com/story/OEG20020116S0039</A>
Last, DDR333 news. Things aren't looking so bright, and JEDEC is at odds of how to package the new DDR333. Some would like to keep using the current process, saying that it still works, although some fear that timing issues might develop in systems with a large number of sticks. DDR II looks to be using µBGA packaging (as opposed to the current TSOP packaging), and some in JEDEC would like to see manufactures move to µBGA with the DDR333, and possibly DDR400, so that the chipmakers are ready for DDR II. Also, the other issue for DDR333 is that without Intel having announced support (there is an enhanced i845-D with 333MHz support to go along with the 533MHz Northwood in the works, according to JEDEC papers) lack of OEM demand for DDR 333 might keep it at as a novelty for a while, something that only a small segment of the public and private sectors would use; ergo, prices would be high. Also, is Intel is slow to adopt DDR333 support, and submits it's specs for DDR II not long after, DDR333 might not last too long, as there will be a quick move to DDR II specs by all JEDEC companies. Article at: <A HREF="http://www.siliconstrategies.com/story/OEG20020117S0062" target="_new">http://www.siliconstrategies.com/story/OEG20020117S0062</A>
Well, that's what I've found, and I find it all very interesting. I find it sad that Intel pretty much holds the fate of DDR333 and DDR II in it's hands, but I think it would be best for them to offer the maximum amount of options to consumers; ergo, having announced support for all availible memory types, until RDRAM or DDR/II/III wins out.
Comment please, either in this thread, or in a new one. I want to know what people think. I have on other article, but it's for a different thread.
-SammyBoy