|
New site? Maybe some day.
|
I'm thinking of upgrading the RTTP server. can someone tell me the speed difference between these two?
#1
Conroe 3040 Dual Core - 1.86 GHz
Dell PowerEdge 840
1066 MHz FSB
1 GB DDR2 RAM
250GB SATA HDD
#2
Intel \ 2.4 GHz \ Pentium 4
Generic \ 1024 MB \ DDR 266
Unknown \ Onboard \ IDE
Seagate \ 80GB:IDE:7200RPM Barracuda \ ST380011a
Western Digital \ 80GB:IDE:7200RPM \ WD800JB |
|
it's DDR 266.. so I'm guessing a 266MHZ bus.
the reason I'm worndering is cause I got an upgrade offer on my RTTP server. the new server is #1, the old (current) server is #2.. I would need to add 2 more banner ads per month to cover the extra cost. |
|
if it is 266 vs 1066 fsb, that is gonna be a huge diff. |
|
ZenErik said:
Conroe is the name for the newer dual core processors.
|
|
I'm looking for some benchmarks here folks |
|
The conroe is faster than the pentium 4, in the sense that it will be processing two streams of data independently where the pentium 4 would only do a single. However, with smaller chunks of data, the multitasking will be less evident and the 2.4 ghz will be quicker than 1.86. You'll get faster bus speeds from the DDR2 memory, but it doesn't list how much. It could be anywhere from 533 to 1066. Finally, for data access you yourself have said that the ATA drive will be much faster than IDE. |
|
right now with the RTTP server this is the issue:
VIRT RES SHR S %CPU %MEM TIME+ COMMAND
566m 224m 2868 S 37.2% 22.3% 9453:54 mysqld
the SQL server is the pig. |
|
and when someone's doing a search on the body of the discussion board
VIRT RES SHR S %CPU %MEM TIME+ COMMAND
566m 224m 2868 S 68.2 22.3 9454:37 mysqld
|
|
Raw horsepower comparison:
http://www.tomshardware.com/2007/03/26/the_gigahertz_battle/page13.html
I sez:
Significantly faster because you're running LAMP, so having two cores is a big win. One can crank along parsing PHP while another processes your SQL queries. The raw comparison above is with both at 2.4 GHz, I interpret that as getting you about a 30% performance boost in a single-threaded app, but probably more like 60% in this case. SQL stuff is particularly memory-bandwidth sensitive, so it might be even more than that given the disparity in memory speed. |
|
Where the fuck is Alex on this?? |
|
why dont u just add a "nice" thing to the search script |
|
it makes sense. if u just made it so that when there was a search it changed the priority of mysqld. and then changed it back when it was done. |
|
"hey, I'm writing a paper for school. can you nice whenever I type the letter a?" |
|
ya know you could just run a second instanace of mysqld just for searches. |
|
and technically you could everytime you type the letter a. |
|
you don't know how mysqld works do you... that would require an extra 600MB of RAM too.
I wouldn't just have to keep 2 different mysqld's, but everything I did would take twice as long cause I would have to update 2 different databases and if I didn't do it that way, then they would still be accessing the same exact memory on the disk so it would be the same bottleneck, but with 2x the RAM used and 2x the calls done. |
|
ok my bad i thought 2 servers could update one database. but its not really supported i guess. |
|
two servers CAN update one database.
but since there is an issue with the speed of accessing that one single database, two people grabbing for something is even slower than 1 person grabbing at something. |
|
oh yea its one file... der |
|
ordered. but only with 1GB RAM. 2GB is like $375 or more more |
[default homepage]
|
[print][ | 2:38:16am Apr 25,2024 load time 0.02543 secs/12 queries] | [search] | [refresh page] |
|