Personal computing discussed
Moderators: renee, Flying Fox, Thresher
Krogoth wrote:Care to enlightenment me?
sparkman wrote:Yes, I'm building myself an AMD gaming rig, because:
1) Phenom 9850 X4 is only about 10% slower than Q9300 in most apps, and that gap is smaller in most games, typically 3%.
2) I like the idea of the all-AMD system and the potential reliability of a single vendor.
3) I hate monopolies. (Intel, not to mention Microsoft and basically every power and phone company in the world.)
flip-mode wrote:I'm sure they'll be along. But did SB700 fix AHCI?
http://techreport.com/articles.x/14261/10 wrote:Like the SB600 before it, the SB700 appears to have problems scaling performance as the number of simultaneous I/O requests increases.
sparkman wrote:flip-mode wrote:I'm sure they'll be along. But did SB700 fix AHCI?
Hmmm. No, apparently they have not fixed the SB600 southbridge problems in SB700.http://techreport.com/articles.x/14261/10 wrote:Like the SB600 before it, the SB700 appears to have problems scaling performance as the number of simultaneous I/O requests increases.
Reliability is a much bigger problem in my mind than the 2% to 10% performance gap with Intel. No matter how much I might like AMD, I won't settle for broken hardware if I don't have to. I may have to go with an Intel system, or wait for SB800.
lex-ington wrote:Do you NEED to have hot-swappable drives? If not, what's the problem again?
sparkman wrote:flip-mode wrote:I'm sure they'll be along. But did SB700 fix AHCI?
Hmmm. No, apparently they have not fixed the SB600 southbridge problems in SB700.http://techreport.com/articles.x/14261/10 wrote:Like the SB600 before it, the SB700 appears to have problems scaling performance as the number of simultaneous I/O requests increases.
Reliability is a much bigger problem in my mind than the 2% to 10% performance gap with Intel. No matter how much I might like AMD, I won't settle for broken hardware if I don't have to. I may have to go with an Intel system, or wait for SB800.
sparkman wrote:lex-ington wrote:Do you NEED to have hot-swappable drives? If not, what's the problem again?
No I don't care about hot-swapping and don't really even know what AHCI is.
My concern is that TR reviewed the SB700 as having installation problems with Windows Vista (critical problem in my opinion), and poor performance scaling on serial ATA (serious problem in my opinion). http://techreport.com/articles.x/14261/10
mattsteg wrote:Out of curiousity, where's the reliability concern? It's a performance deficiency.
mattsteg wrote:sparkman wrote:http://techreport.com/articles.x/14261/10
If you don't need ahci, feel free to not enable it. I was able to install vista on a 780g box in ahci mode without jumping through any hoops, btw.
lex-ington wrote:If AHCI is not an issue for you, then this is the board for you. Turn off AHCI, turn off the on-board video and go to town.
sparkman wrote:mattsteg wrote:sparkman wrote:http://techreport.com/articles.x/14261/10
If you don't need ahci, feel free to not enable it. I was able to install vista on a 780g box in ahci mode without jumping through any hoops, btw.
Why then did the TR article say installing Vista in AHCI mode would cause problems? They complained: "It's ridiculous to ask users to jump through hoops just to enable AHCI."
Not that I even want to use AHCI in Windows. I just assumed it was the default, and that I'd have to jump thru hoops to avoid AHCI to install Windows in simple IDE mode. But you are saying AHCI mode works fine?
lex-ington wrote:If AHCI is not an issue for you, then this is the board for you. Turn off AHCI, turn off the on-board video and go to town.
sparkman wrote:My concern is that TR reviewed the SB700 as having installation problems with Windows Vista (critical problem in my opinion), and poor performance scaling on serial ATA (serious problem in my opinion). http://techreport.com/articles.x/14261/10
deruberhanyok wrote:sparkman wrote:
But if you look at the subsequent pages, charts and graphs you'll see that in some cases the worst you're looking at is the difference between 3ms and 5ms. Yes, it's 66% slower than 3ms. But it's 5 milliseconds.