AMD have always denied that the PR number is a comparison to the P4 (though it does seem to work out pretty close on average).
They claim it's supposed to compare against the the 'Thunderbird' model Athlon (the one that topped out at 1.4 GHz). Not the Duron.
Most people will keep matching it against the P4 regardless, of course. Will this continue to hold true against the Prescott (allowing AMD to hike up their PR numbers by a goodly amount), or will they stick to their supposed guns?
Repeat after me - this is not analog. You don't have to squeeze out every last Hz of bandwidth. All you need is enough (< 1 MHz) to get a few ones and zeroes across with enough clarity to tell one from the other. Plastic Toslink fibre or coax will do that just fine.
If you want to spend hundreds on high-quality cables, spend it on speaker connects, or on your DACs or amp stage. Don't waste your cash on ridiculously priced digital equipment (and then boast about it) when a $5 piece of wire and a $20 CD-ROM drive will do the exact same thing.
Sure - except, can you imagine the cost of a flash card that can store 2 hours of movie at 8.9 GB/hr? Or perhaps you take home an external 1394 drive, plug it into your PC & stream from there? You'd have to leave a big enough deposit to pay for the Roku itself.
Maybe with future plastic memory developments, or simply HD-DVDs, but not this thing.
...since this box doesn't receive or output a digital picture signal. Analog connectors only, so you avoid any evil bits (in exchange for a generation loss in picture quality).
Optical, coaxial - whatever, it's all digital, boys & girls. The ones and zeroes - and therefore the sound quality - is the same.
The advantage of optical cable is it's immune to RF noise, but you'd have to live in an unusually noisy environment for it to be bad enough to corrupt a relatively low-speed signal like that. I used to run ordinary S/PDIF over 20m of cheap-ass audio cable (computer to receiver's DACs), and couldn't pick the difference between a CD played on the computer to one played on the local CD-player.
I have a friend with an overpriced stereo system that actually uses fully balanced AES cables to run the digital signal from his CD transport to the DACs, but even he admits that's pure overkill.
I'd be more concerned about the picture quality loss from using analog component cables - a DVI connector would solve that, as someone else has pointed out.
As a developer (film fx software), I'd not only overlook, but probably encourage low-income students to use pirated versions of my app("third party demo versions", I call them). It's not like it's costing us a sale, and any exposure to our software will encourage them to ask for it when working at a real studio.
OTOH, I frown upon studios using cracks professionally, naturally enough. If you're using the results of my work to generate an income, you should contribute to my income. If you're just playing around, be my guest.
Of course, my employer may have different views, but that's my opinion.
Thanks, that's very helpful. From what you say, all Athlon64 boards should be created equal, short of someone not bothering to wire up the high address lines. Perhaps that's the basis of Via's supposed 4 GB limit?
I did see registered DDR400 2 GB DIMMs available here, or so they claim. A bit much for your average Athlon64 buyer, as you say, but you ought to be able to get at least 4 GB from two of those, if not 6 GB from all three slots.
There's a dually MSI workstation board that is vague about its total memory limit (Via K8T800, 4 slots, supports 2 GB registered DIMMs, 8 GB total) though I've seen at least one review that suggested it had a 3 GB limit.
As you say. Things are apparently less friendly in the Mac world than one might be led to believe, if vendors try that sort of stunt. Are they suppopsed to be custom matched to each other, like speakers?;-)
There is still this one. Insane price, but at least it's registered.
Most dual Opteron boards I've seen do support dual-channel memory - except for the Tyan K8W, which supports dual-channel memory for each CPU. It's the only one, at least that has an AGP slot. I'd heard tell (unconfirmed) that memory slots for each CPU, apart from adding to the board cost, also confused most gfx drivers, since it bumped the AGP slot from PCI device 0 to PCI device 1.
Which boards have only single-channel DDR memory? I ask because I wish to avoid buying them:-)
... which the article says supports only 2GB of RAM. Why do they say that? The MSI board specs they list show:
- Supports three 184-pin DDR SDRAMs up to 2GB memory size
Why would you get a 2 GB total limit from 3 memory slots? I read that as the DIMMs being 2 GB each, which sounds like 6 GB to me.
The Asus and Shuttle boards seem a bit firmer about their memory limits, but they claim a 3 GB limit. And the Asus board uses the same chipset as the MSI - the Via K8T800, which Via says has a 4 GB memory limit - so someone's confused. I know I am.
I have around 3 hours (two full albums) on my P800, in Ogg Vorbis format (in addition to a dozen books & a large range of useful and less-useful apps). All fits nicely on a 128MB Memory Stick Duo, and sounds just fine through the supplied stereo headset.
I listen to music through it every day. When my kids broke my laptop's line-out jack, I used my P800 for music instead. It's better than your average MP3 player because it supports other formats (such as the above mentioned Ogg Vorbis), and it's always there. Anywhere I take my phone, I'm also taking my music player (and my books, addresses, a few games, a cheapo camera, a notepad, a dictionary, a planetarium etc etc etc).
I've only seen one app that would continue to suck down power like that - the leaked beta camcorder app. It would drain my battery in about 3 hours, because it kept doing MPEG4 compression even in the background. That's why it wasn't released officially.
That aside, I can have as many apps as I like sitting idle without draining power. It's only when they're being actively used that the battery life goes. You don't need to close them, regardless of what switcher claims (the memory used to store them is powered on regardless - only accessing it uses the juice).
I don't get 400 hours by any means, but I get at least a couple of days of frequent usage (I read books & play [CPU intensive] ogg files each day). If I turn off the phone side (Flight mode), battery life goes up noticeably, or if I only use it as a phone (not a PDA), it's a lot better.
I'd say battery life was pretty good, for an equivalent PDA, though low for a simple phone. But it only takes 60-90m to charge anyway, so I have no problem with it.
Java apps can be started & used just like C++ apps - just tap the icon (or select it with the jogdial & click, or press a shortcut button or whatever).
If the phone rings, the phone app pops to the front, but the other apps are still there, running in the background. On the P800 at least, GPRS data connections are suspended and (other) sounds are muted, but the apps keep going and you can browse the address book or anything else.
Frankly, there's not much point having a higher resolution - if the P900's camera is similar to my P800's, then it's limited by more than just the resolution of the sensor. The lens is tiny (and crappy), and the sensor is slow.
The pictures are grainy and blurry round the edges, and the subject should be very well lit to have any hope of looking clear (no flash, of course). Once they put a quality lens in there, a quality sensor might be more useful.
And I'm told that even so, the P800's pictures are as good or better than the majority of cameraphones.
Slashdot Mirror
VGA has 1:1 pixel aspect, 4:3 frame aspect.
NTSC has 0.9:1 pixel aspect, 4:3 frame aspect.
</slashdotpickyness>
They claim it's supposed to compare against the the 'Thunderbird' model Athlon (the one that topped out at 1.4 GHz). Not the Duron.
Most people will keep matching it against the P4 regardless, of course. Will this continue to hold true against the Prescott (allowing AMD to hike up their PR numbers by a goodly amount), or will they stick to their supposed guns?
More info here.
Works fine on my nForce2 mobo, and my others too. Perhaps a driver issue?
If you want to spend hundreds on high-quality cables, spend it on speaker connects, or on your DACs or amp stage. Don't waste your cash on ridiculously priced digital equipment (and then boast about it) when a $5 piece of wire and a $20 CD-ROM drive will do the exact same thing.
All it has to do is decode an HDTV-resolution MPEG2 file for playback, and output the analog signal correctly - and that it can do.
Maybe with future plastic memory developments, or simply HD-DVDs, but not this thing.
Video Output
Component Y/Pr/Pb: 1080i, 720p, 480p, 480i
VGA: 1080i, 720p, 480p
No recording though.
...since this box doesn't receive or output a digital picture signal. Analog connectors only, so you avoid any evil bits (in exchange for a generation loss in picture quality).
The advantage of optical cable is it's immune to RF noise, but you'd have to live in an unusually noisy environment for it to be bad enough to corrupt a relatively low-speed signal like that. I used to run ordinary S/PDIF over 20m of cheap-ass audio cable (computer to receiver's DACs), and couldn't pick the difference between a CD played on the computer to one played on the local CD-player.
I have a friend with an overpriced stereo system that actually uses fully balanced AES cables to run the digital signal from his CD transport to the DACs, but even he admits that's pure overkill.
I'd be more concerned about the picture quality loss from using analog component cables - a DVI connector would solve that, as someone else has pointed out.
Amen to that.
As a developer (film fx software), I'd not only overlook, but probably encourage low-income students to use pirated versions of my app("third party demo versions", I call them). It's not like it's costing us a sale, and any exposure to our software will encourage them to ask for it when working at a real studio.
OTOH, I frown upon studios using cracks professionally, naturally enough. If you're using the results of my work to generate an income, you should contribute to my income. If you're just playing around, be my guest.
Of course, my employer may have different views, but that's my opinion.
It would probably be cheaper to buy a Tyan K8W mobo (8 DIMM slots) and even new CPU(s), than to buy 4 x 2 GB DIMMs :-(
I did see registered DDR400 2 GB DIMMs available here, or so they claim. A bit much for your average Athlon64 buyer, as you say, but you ought to be able to get at least 4 GB from two of those, if not 6 GB from all three slots.
There's a dually MSI workstation board that is vague about its total memory limit (Via K8T800, 4 slots, supports 2 GB registered DIMMs, 8 GB total) though I've seen at least one review that suggested it had a 3 GB limit.
There is still this one. Insane price, but at least it's registered.
These boards use the K8T800 chipset which has a 4 GB limit. I'm wondering what it is that stops them using even that much.
It sounds to me more like it's a single DIMM with 2 x 1 Mb (128M x 64) chips on it.
Which boards have only single-channel DDR memory? I ask because I wish to avoid buying them :-)
What are these then?
Why would you get a 2 GB total limit from 3 memory slots? I read that as the DIMMs being 2 GB each, which sounds like 6 GB to me.
The Asus and Shuttle boards seem a bit firmer about their memory limits, but they claim a 3 GB limit. And the Asus board uses the same chipset as the MSI - the Via K8T800, which Via says has a 4 GB memory limit - so someone's confused. I know I am.
If you consider a 0.5% lead in a synthetic benchmark to be "schooled", I suppose. And ignore the other 3 graphs on that page.
Also unfortunate is the price - AUD$1000+ down under. What good is a well-performing dollar when you still get these prices, I ask you?
I listen to music through it every day. When my kids broke my laptop's line-out jack, I used my P800 for music instead. It's better than your average MP3 player because it supports other formats (such as the above mentioned Ogg Vorbis), and it's always there. Anywhere I take my phone, I'm also taking my music player (and my books, addresses, a few games, a cheapo camera, a notepad, a dictionary, a planetarium etc etc etc).
I've only seen one app that would continue to suck down power like that - the leaked beta camcorder app. It would drain my battery in about 3 hours, because it kept doing MPEG4 compression even in the background. That's why it wasn't released officially.
That aside, I can have as many apps as I like sitting idle without draining power. It's only when they're being actively used that the battery life goes. You don't need to close them, regardless of what switcher claims (the memory used to store them is powered on regardless - only accessing it uses the juice).
I don't get 400 hours by any means, but I get at least a couple of days of frequent usage (I read books & play [CPU intensive] ogg files each day). If I turn off the phone side (Flight mode), battery life goes up noticeably, or if I only use it as a phone (not a PDA), it's a lot better.
I'd say battery life was pretty good, for an equivalent PDA, though low for a simple phone. But it only takes 60-90m to charge anyway, so I have no problem with it.
If the phone rings, the phone app pops to the front, but the other apps are still there, running in the background. On the P800 at least, GPRS data connections are suspended and (other) sounds are muted, but the apps keep going and you can browse the address book or anything else.
The pictures are grainy and blurry round the edges, and the subject should be very well lit to have any hope of looking clear (no flash, of course). Once they put a quality lens in there, a quality sensor might be more useful.
And I'm told that even so, the P800's pictures are as good or better than the majority of cameraphones.