Make Home Page
Add to Favorites
XbitLabs is using Twitter
Windows Live Alerts
1.
Is there really that big of a market for quad core mobile chips? Considering the extra power they would consume, and the nature of mobile computing, it seems like a relatively small market to me, and that's putting it mildly.
With regards to transitioning to the P8, it is hard to imagine Intel being in a worse situation when those CPUs come out. I have to believe they will delay the P8s for a while to allow the transition to take place, because who will buy a P7 once the P8 is out? They have a difficult time competing with the mediocre Athlon, coming from a knock-off brand company. How will they compete against a vastly superior processor like the P8, from the leading company (themselves) when it becomes available? How will they continue to sell the wretched P7s they are making while the transition is still in progress? Why would anyone buy these chips when the P8 is out? Actually, why would anyone buy them even now with the Athlon 64 out? But, it will get even worse.
Has anyone else noticed that Intel "odd" numbered processors suck big time, whereas the even numbered ones are excellent? 8086 kind of sucked, but it was a big improvement over the 8085, actually, an enormous one. So, call it mediocre. 80186 sucked so bad it was barely used except in embedded. 80286 was shockingly fast, and opened up multitasking operating systems and massive amounts of memory. I still remember using an PC AT and being stunned by the performance. The 386 was a miserable processor, and could not even perform as well as a 286 on 16-bit code clock normalized. It added 32-bit processing, but was so slow it really didn't matter all that much, except to run multiple real mode applications. The 486 ran like a raped ape too, being roughly twice as fast as the 386 clock normalized. It also allowed multiplying the internal bus by the memory bus, because it had a L1 cache (a whopping 8K on the original 486). The Pentium kind of sucked too, particularly since one could get a 486 running at 100 MHz at the same time Intel was selling their buggy, very hot 60 and 66 MHz Pentiums. The Pentium wasn't that bad though, but still was disappointing. The P6 core is legendary and put an end to the RISC era for workstations, essentially. It needs no further comment. The P7 is also legendary, at sucking. It's a fascinatingly bad design that I still find difficult to believe. How did Intel let this piece of crap see the light of day? A huge mystery I'll never understand. Now we have the P8, which I haven't seen yet, but by all indications seems utterly fantastic, well, within relative terms. The days of the 286 type performance increases are over, and even the 486 type improvement will probably never happen again. But, considering the crap we've been getting for the past few years from Intel and AMD, this is a huge leap forward. I knew it would be better than the Athlon 64, but I had no idea it would that much better. The bad part is, the P9 is probably going to be another Intel dog. Weird how that pattern is so hard for them to deviate from.
With regards to transitioning to the P8, it is hard to imagine Intel being in a worse situation when those CPUs come out. I have to believe they will delay the P8s for a while to allow the transition to take place, because who will buy a P7 once the P8 is out? They have a difficult time competing with the mediocre Athlon, coming from a knock-off brand company. How will they compete against a vastly superior processor like the P8, from the leading company (themselves) when it becomes available? How will they continue to sell the wretched P7s they are making while the transition is still in progress? Why would anyone buy these chips when the P8 is out? Actually, why would anyone buy them even now with the Athlon 64 out? But, it will get even worse.
Has anyone else noticed that Intel "odd" numbered processors suck big time, whereas the even numbered ones are excellent? 8086 kind of sucked, but it was a big improvement over the 8085, actually, an enormous one. So, call it mediocre. 80186 sucked so bad it was barely used except in embedded. 80286 was shockingly fast, and opened up multitasking operating systems and massive amounts of memory. I still remember using an PC AT and being stunned by the performance. The 386 was a miserable processor, and could not even perform as well as a 286 on 16-bit code clock normalized. It added 32-bit processing, but was so slow it really didn't matter all that much, except to run multiple real mode applications. The 486 ran like a raped ape too, being roughly twice as fast as the 386 clock normalized. It also allowed multiplying the internal bus by the memory bus, because it had a L1 cache (a whopping 8K on the original 486). The Pentium kind of sucked too, particularly since one could get a 486 running at 100 MHz at the same time Intel was selling their buggy, very hot 60 and 66 MHz Pentiums. The Pentium wasn't that bad though, but still was disappointing. The P6 core is legendary and put an end to the RISC era for workstations, essentially. It needs no further comment. The P7 is also legendary, at sucking. It's a fascinatingly bad design that I still find difficult to believe. How did Intel let this piece of crap see the light of day? A huge mystery I'll never understand. Now we have the P8, which I haven't seen yet, but by all indications seems utterly fantastic, well, within relative terms. The days of the 286 type performance increases are over, and even the 486 type improvement will probably never happen again. But, considering the crap we've been getting for the past few years from Intel and AMD, this is a huge leap forward. I knew it would be better than the Athlon 64, but I had no idea it would that much better. The bad part is, the P9 is probably going to be another Intel dog. Weird how that pattern is so hard for them to deviate from.
[Posted by: TA152H 
| Date: 04/14/06 04:08:51 PM]
| Date: 04/14/06 04:08:51 PM]
