Dothan and Turion Compared and Explained
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
-
- *Lifetime Patron*
- Posts: 1809
- Joined: Sat Apr 24, 2004 1:45 am
- Location: At Home
I’ve just built a desktop system in the UK and considered a Turion; but they are hard to source and cost more than my budget. I did though find a Sempron mobile 2800 (S754 25W) for the same price as a desktop Sempron 3000. If you’re having trouble sourcing a Turion MT and don’t need x64, then that’s an option worth considering. I doubt the Sempron mobiles will get x64 for sometime, as Intel have no competing product for ages.
Am I correct in saying that AMD optimised the core in the Turion range for low power and that they aren’t just hand picked A64s?
Am I correct in saying that AMD optimised the core in the Turion range for low power and that they aren’t just hand picked A64s?
-
- Posts: 58
- Joined: Tue Mar 22, 2005 11:54 pm
I recently got (after my Winchester somehow stopped working) a cheap and cheerful Sempron E6 (i.e. 64bit Sempron) 2800 and while it doesn't support Cool and quiet, it doens't really need it either:
Running at stock 1.4V (Asus K8V SE doesnt seem to want to undervolt in the BIOS), it gets no hotter than 50° under cpuburnK7 cooled by a Thermaltake SonicTower (that's right, that supposedly crappy cooler, yes) AND ONE SINGLE 120mm case fan (Fanless Etasis PSU) at 7V in a dirt cheap Antec SLK3700 (next thing I'm gonna replace, I want a THICK STAINLESS case . Admittedly, my Radeon 9000 is cool, but still, cooling the CPU passively is entirely possible. It wouldn't take too much to go to a completely fanless system I guess but I rather prefer having my Harddisks sitting in SOME airflow
So I don't really see the point in getting a Turion for desktop use, anymore
Running at stock 1.4V (Asus K8V SE doesnt seem to want to undervolt in the BIOS), it gets no hotter than 50° under cpuburnK7 cooled by a Thermaltake SonicTower (that's right, that supposedly crappy cooler, yes) AND ONE SINGLE 120mm case fan (Fanless Etasis PSU) at 7V in a dirt cheap Antec SLK3700 (next thing I'm gonna replace, I want a THICK STAINLESS case . Admittedly, my Radeon 9000 is cool, but still, cooling the CPU passively is entirely possible. It wouldn't take too much to go to a completely fanless system I guess but I rather prefer having my Harddisks sitting in SOME airflow
So I don't really see the point in getting a Turion for desktop use, anymore
-
- Posts: 1608
- Joined: Tue Jan 04, 2005 4:02 pm
- Location: United States
From what I understand, the CPU actually has a "microcode" programmed into it that the motherboard interprets to set voltage, multiplier, etc. However, if the microcode for a particular CPU has not been added to the BIOS, it won't know what to do. But generally this isn't a problem, as most BIOS have a failsafe designed into them which sets the multiplier and voltage to lowest CnQ settings.~El~Jefe~ wrote:doesnt the chip itself automatically tell the board what to set itself at?frostedflakes wrote:I have an ASUS K8V-X on the way for my Turion. It uses Via K8T800.
Turion should work in any board with a rev. E compatible Sempron BIOS, though. The Turion cores and rev. E Sempron cores aren't identical, but they're similar enough that the board should still run fine with a Turion. Worse case scenario you'd probably have to manually set voltage, multiplier, etc. in the BIOS.
yay or nay?
I have an Abit kv8 pro. if i were to get a new board, i then wouldnt bother with turion. I would just get a venice core. so.... i wonder if it will work?
the board isnt undervoltable in bios.
But if your board doesn't even allow undervolting, I doubt it would support Turion. Are you using the most recent BIOS?
smilingcrow: According to AMD, Turion has been optimized at the hardware level for lower power consumption, and it is not just a San Diego core running at a lower voltage.
-
- Posts: 58
- Joined: Tue Mar 22, 2005 11:54 pm
I dont really know. It was working in the evening and when I came back in the morning, nothing was on anymore. The ASUS A8V didnt show anything on screen but used the speech post stuff to tell my about a CPU error.vertigo wrote:What happened to the Winchester? Were you doing anything which might have caused it to break?
The dealer refused to take it back (that's what you get for buying from the cheapest one I guess, won't do that again) and seeing that a venice would have cost me more than the the new socket754 board plus CPU I went the Sempron64 route. It's plenty fast for me anyway.
At the time of failure, the Athlon was running at 1.225V which in theory shouldn't matter much but talking to an electro engineer friend of mine he said under certain circumstances it might still have fried the CPU running it outside of spec. It should not, but it's not entirely to be ruled out. It could have been a bad CPU, simply, after using over 20 AMD CPUs without any trouble whatsoever over the past decade I'm bound to encounter a bad one at some point I guess.
Interesting read about the Turion,
I have a friend with a 754 motherboard that wants to upgrade next spring to a dual-core--with the dual core Turions plug right in with a BIOS update? Will the dual-core Turions be available in a Socket 939 configuration? My son's Athlon XP Mobile 2400 is a cool running chip even overclocked (2.1GHz 210x10 at stock voltage) A Turion dual core would be a no brainer for home theater PC use, it will have the horsepower without the heatpower to go with it.
As far as laptops go, finally! A laptop that has decent video performance and many options to play with. The 25 watter would be great if coupled with X700 onboard graphics, a gig of DDR and a large battery. I don't own a laptop yet, the video performance sucks on a Dothan due to the crappy onboard graphics.
Is a dual-core Turion S939 in my future? Not sure but it would make my HTPC bark and do so quietly while running cool.
I have a friend with a 754 motherboard that wants to upgrade next spring to a dual-core--with the dual core Turions plug right in with a BIOS update? Will the dual-core Turions be available in a Socket 939 configuration? My son's Athlon XP Mobile 2400 is a cool running chip even overclocked (2.1GHz 210x10 at stock voltage) A Turion dual core would be a no brainer for home theater PC use, it will have the horsepower without the heatpower to go with it.
As far as laptops go, finally! A laptop that has decent video performance and many options to play with. The 25 watter would be great if coupled with X700 onboard graphics, a gig of DDR and a large battery. I don't own a laptop yet, the video performance sucks on a Dothan due to the crappy onboard graphics.
Is a dual-core Turion S939 in my future? Not sure but it would make my HTPC bark and do so quietly while running cool.
-
- Posts: 1608
- Joined: Tue Jan 04, 2005 4:02 pm
- Location: United States
-
- Posts: 1608
- Joined: Tue Jan 04, 2005 4:02 pm
- Location: United States
-
- Posts: 1386
- Joined: Fri Jun 18, 2004 6:53 pm
Going back to the actual review. Without venturing into the desktop area that includes overclocking/undervolting/cooling/measuring how desktop memory and such affect performance but looking at the review purely from the mobile point I'd say this review isn't bad. On average Turion performs on par with P-M and while the CPU is only somewhat cheaper than P-M, Turion laptops themselves are considerably cheaper (or centrino laptops are much more expensive) due to the "centrino" factor. So Turion might be a very good and inexpensive alternative to centrino.
If you noticed I said might and that's because I think that one of the most important sections of the review - the batterylife is a total rubbish. Mainly because it doesn't explain how both laptops were tested. My banias laptop as most of the other centrino laptops come with power management software that lets you control the CPU frequency clock, hard drive shutdown time, automatic screen brightness depending on the battery life and such. The review says nothing about how the tests were done. I for example can forcibly set my laptop to operate at minimum frequency to prolong the battery life (for example if I'm just typing or low on battery), or I can forcibly set it to maximum to get the maximum performance out of it, or I can set it to vaiable frequency that will vary depending on the load. It is perfectly possible that centrino laptop was set to variable frequency and turion to lowest which might explain sudden increase in battery life. It's also possible that power saving implementations in centrino and turion CPUs are vastly different, for example if set to variable frequency centrino laptops idle at about 1Ghz instead of 600Mhz minimum, it's possible that turion goes all the way down which might explain to its minimum frequency which might explain 20 minute advantage in battery life. However if that is the case than it doesn't mean that turion is automatically better because you can manually set centrino to operate at 600Mhz at which point it might take over the turion. It is also possible that on one laptop the hard drives were set to shutdown after five minutes and on the other one after half an hour.
The point is they did not do the battery life test properly. If they wanted to do it properly they had to do at least three sets of tests, one battery life at maximum frequency, second one at minumum frequency and third one with adjustable frequency to compare how well power saving features work on both processors. They would also have to set all other settings like hard drive shutdown the same. As it is right now though, the battery life section of the review is absolutely useless.
If you noticed I said might and that's because I think that one of the most important sections of the review - the batterylife is a total rubbish. Mainly because it doesn't explain how both laptops were tested. My banias laptop as most of the other centrino laptops come with power management software that lets you control the CPU frequency clock, hard drive shutdown time, automatic screen brightness depending on the battery life and such. The review says nothing about how the tests were done. I for example can forcibly set my laptop to operate at minimum frequency to prolong the battery life (for example if I'm just typing or low on battery), or I can forcibly set it to maximum to get the maximum performance out of it, or I can set it to vaiable frequency that will vary depending on the load. It is perfectly possible that centrino laptop was set to variable frequency and turion to lowest which might explain sudden increase in battery life. It's also possible that power saving implementations in centrino and turion CPUs are vastly different, for example if set to variable frequency centrino laptops idle at about 1Ghz instead of 600Mhz minimum, it's possible that turion goes all the way down which might explain to its minimum frequency which might explain 20 minute advantage in battery life. However if that is the case than it doesn't mean that turion is automatically better because you can manually set centrino to operate at 600Mhz at which point it might take over the turion. It is also possible that on one laptop the hard drives were set to shutdown after five minutes and on the other one after half an hour.
The point is they did not do the battery life test properly. If they wanted to do it properly they had to do at least three sets of tests, one battery life at maximum frequency, second one at minumum frequency and third one with adjustable frequency to compare how well power saving features work on both processors. They would also have to set all other settings like hard drive shutdown the same. As it is right now though, the battery life section of the review is absolutely useless.
The overclocking of the Pentium M vga card might also affect it's powersaving performance. This might explain the poor idle performance.JazzJackRabbit wrote:If you noticed I said might and that's because I think that one of the most important sections of the review - the batterylife is a total rubbish.
I had the same thoughts and I completely agree with this.JazzJackRabbit wrote:Going back to the actual review. Without venturing into the desktop area that includes overclocking/undervolting/cooling/measuring how desktop memory and such affect performance but looking at the review purely from the mobile point I'd say this review isn't bad. On average Turion performs on par with P-M and while the CPU is only somewhat cheaper than P-M, Turion laptops themselves are considerably cheaper (or centrino laptops are much more expensive) due to the "centrino" factor. So Turion might be a very good and inexpensive alternative to centrino.
If you noticed I said might and that's because I think that one of the most important sections of the review - the batterylife is a total rubbish. Mainly because it doesn't explain how both laptops were tested. My banias laptop as most of the other centrino laptops come with power management software that lets you control the CPU frequency clock, hard drive shutdown time, automatic screen brightness depending on the battery life and such. The review says nothing about how the tests were done. I for example can forcibly set my laptop to operate at minimum frequency to prolong the battery life (for example if I'm just typing or low on battery), or I can forcibly set it to maximum to get the maximum performance out of it, or I can set it to vaiable frequency that will vary depending on the load. It is perfectly possible that centrino laptop was set to variable frequency and turion to lowest which might explain sudden increase in battery life. It's also possible that power saving implementations in centrino and turion CPUs are vastly different, for example if set to variable frequency centrino laptops idle at about 1Ghz instead of 600Mhz minimum, it's possible that turion goes all the way down which might explain to its minimum frequency which might explain 20 minute advantage in battery life. However if that is the case than it doesn't mean that turion is automatically better because you can manually set centrino to operate at 600Mhz at which point it might take over the turion. It is also possible that on one laptop the hard drives were set to shutdown after five minutes and on the other one after half an hour.
The point is they did not do the battery life test properly. If they wanted to do it properly they had to do at least three sets of tests, one battery life at maximum frequency, second one at minumum frequency and third one with adjustable frequency to compare how well power saving features work on both processors. They would also have to set all other settings like hard drive shutdown the same. As it is right now though, the battery life section of the review is absolutely useless.
I just ordered a Asrock DualSata board and it comes with a slot that will accept a future M2 upgrade card. As far as I know this is the first board with this feature but hopefully more boards will have this - its a good money saver.~El~Jefe~ wrote:any word on the availability of the m2 socket??
no one talks about them.
-
- Posts: 1608
- Joined: Tue Jan 04, 2005 4:02 pm
- Location: United States
-
- Posts: 1608
- Joined: Tue Jan 04, 2005 4:02 pm
- Location: United States