There might be limitations on the max sustained transfer of any PCMCIA to Firewire card you get, though this isn’t exactly my area of expertise. For all I know, notebooks come with firewire ports these days…
firewire port indeed.
its a built in 1394 port, but its one of those little 4 pin ports. I’d never seen one before, but they are apparently common in Digital Video equipment these days. I haven’t hooked up the hard rive to it yet, but I just got a Firewire sound card (MOTU 828, if anyone is interested) and its working quite nicely. The problem is it only has ONE 1394 port, so I had to buy a 1394 mini-hub that requires either power from 1394, or an AC adaptor.
Well the mini 4 pin 1394 connector don’t have no power (AC, anyway) so i have to go out to radioshack tomorrow to look for a 12v adaptor to make this thing work, either that or I spring for a PCMCIA to 1394 adaptor and hook the less need of my firewire devides into it. I’m guessing the sound card probably wins there, although really they are equally needy in a way, since anything that the drive has to read has to go out to the sound card and any thing that the soundcard hears gets written to the drive. I’m just hoping that this hub works, and that I never have to find out.
and before anyone asks, yes, both the HD and the sound card are terminal, so no daisy-chaining, unfortunately.
Chas says:
I’m sorry, could you point out the mac-bashing in my post? I seem to have misplaced it.
One also has to wonder how the question “are laptops just as good as desktops?” could provoke the answer “A Powerbook G4/500 will … beat the pants off of any Wintel portable computer for speed.” If that was not completely off-topic, uninvited, mac-vs-pc chest-pounding, then please accept my humblest apologies. But do tell us what it was.
Post some facts, then, instead of this:
- Yes, your speed does have a lot to do with bus speeds and caches. There is a nice little tutorial on it you can read at Tomshardware.com, an excellent guide to the ins and outs of PC performance.
- 7200 rpm is not slow for a PC hard drive. Go look at any hard drive mfr’s lineup right now and you will find that about half their drives, even the high performance ones, are 5400rpm. 7200 and 10000 rpm drives do, as a rule, perform better, but not that much better. A little cache on your drive will go a long way.
KKBattousai and Bad Hat: The FireWire stuff is probably running at full PCI bus speed, just like on a desktop. On a built-in firewire port, it’s probably just a PCI device, and even if you had a PCMCIA firewire card added on, it’s almost certainly going to be CardBus (which is the 32-bit PCMCIA, potentially having 133Mbits of throughput), and therefore running at PCI speeds as well.
A quick check for PCMCIA firewire cards on pricewatch.com turned up a whole bunch, and they were all CardBus.
In my experience laptops tend to break down more easily than desktops, and are more expensive to repair.
So perhaps we ought to make a list:
Laptops are slower
Desktop hardware is more cost-effective
Laptops are portable (wait–is this a good thing?? I hate it when my work follows me home)
Desktops don’t get stolen
and so on. I don’t know a whole lot about hardware mechanics, but when it comes to comfort, ease of use, and reliability my vote goes to the desktop system.
The whole post. All I have to do is contrast Macs vs PCs and you go into attack-dog mode. Cut the crap.
It was absolutely on-topic, it addressed the biggest specific complaint associated with Wintel laptops. The biggest limit to performance in a Wintel laptop is heat production, processors must be slowed down in order to avoid heat-induced failure, or else another convoluted strategy must be adopted like liquid cooling. I will explain it again, more explicitly, so even people like yourself with weak knowledge of how microprocessors work will understand.
There are two basic limits to performance in any computer system, I/O and the CPU. Even if you only consider the CPU, and all other specs are identical (which is virtually impossible), a laptop microprocessor will always be slower than a desktop processor of identical mhz. Intel and AMD make a separate line of processors targeted at laptops with slower processing and lower heat production. Rating a chip on mhz alone is a marketing scam. A classic example is the newest Pentium 4 chips for laptops, they run at a far lower speed than their maximum rated speed, some CPU reviewers allege that the chips only run an average of 300mhz. Even worse, a 1000Mhz Pentium 4 is slower than a 1000Mhz Pentium 3.
HOWEVER, this does not apply to PowerPC chips. A PowerPC chip will run at 500Mhz in a laptop precisely as it does in a desktop. Furthermore, Macs have a “unified motherboard architecture” so the same basic motherboards are used in desktops AND laptops. So the CPU speed considerations of laptop vs. desktop basically does not apply to Macs. And that’s a fact, jack.
- CPU speed does have an effect on CPU-intensive tasks. But most people do not do CPU-intensive tasks on a routine basis and I/O performance is typically the limiting factor.
- A 7200RPM IDE drive IS slow in a world of 15kRPM and now 18kRPM SCSI drives. But we’re not talking about products that laptops can’t use. I was appalled to discover recently that some Wintel manufacturers are still shipping laptops with 2400RPM drives. You can throw all the cache you want at a slow drive like that and it will still never perform close to the specs of a faster drive.
To give a concrete example, a friend and I both have identical CPUs, but he has a 7200RPM ATA drive and I have a 10kRPM Ultra2SCSI drive. Even on relatively CPU intensive tasks like compressing video, my machine will complete a job in about half the time (yes, we benchmarked it). This can be a huge difference when you’re doing a compression job that takes 10 or 12 hours to complete.
:rolleyes: Honestly, I can’t see how you can take a comment such as mine as “attack-dog mode”. In addition, it was not “the whole post”, it was merely the first line. The rest was dedicated to commenting on the issue at hand, and wasn’t even directed at you. So you cut the crap.
We’re all very grateful. Yet a CPU with a slower bus and less cache is still going to be a performance problem, and that’s a definite contributor in laptops. Despite your vast knowledge of microprocessors, you’ve failed to demonstrate otherwise. Perhaps you can come up with another explanation for different performance in CPU’s of the same clock speed with identical instruction sets. “It’s just slower” doesn’t cut it around here.
This is fascinating. Can you point me to some consumer level drives that spin faster than 10k RPM? I looked around on several drive manufacturers’ sites, and I can’t find any. In fact, I am under the impression that even in desktop machines, where cooling isn’t so much of an issue, 10k RPM drives are pushing the limit as far as heat dissipation goes.
And if you compared a 7200RPM ATA drive to a 10kRPM ATA drive, you’d see a difference too, but it wouldn’t be nearly as dramatic. Your benchmark is fundamentally flawed, as ATA I/O both eats CPU and waits for it. If you didn’t know this, you have a “weak understanding” of PC I/O, but I’ll assume you just overlooked this detail in your benchmarking.
You’re right about disk performance being a factor, but you’re wrong when you claim that bus speed and cache size are not a factor. It is difficult to say which is a larger factor (and it obviously depends on what operations you’re doing).
Will this do? Found via Google search for “SCSI Hard Drive”.
Seagate appears to be the only one actually manufacturing 15K drives. FWIW, I have a couple of the 15K 18GB drives in a Compaq 5500 running SQL Server; they have heat sinks, and with proper cooling they appear to run just as cool as 10K drives. Of course, this is in an air-conditioned data center, but the running drives were cold to the touch.
I have not seen 18K drives anywhere, although Seagate does have a 180GB 7200 RPM disk.
Thanks, Jeff. Considering that those are hot-pluggable disks for ProLiant servers, I’m still reserving judgement on the claim that they’re standard PC equipment. 5400 and 7200 are by far the most common drive speeds in PC’s, and a desktop machine with anything over 10k rpm is something rare and extreme.
galt, may I suggest you ignore people who make it a habit of dragging reasonable threads into GD for their own political reasons?