An SSD or any 500 GB/platter drives, which includes the Samsung F3 or any 2 TB drives.

I wouldn't worry about performance unless you have more than 100 Mbps Internet.

100 megabit/s = 12.5 megabyte/s.
The old WD Green 1 TB (WD10EADS, 3 x 333 GB/platter, 5400 RPM) has a write speed of about 70 megabytes/s.

Just buy what you like or according to your budget.

This post has been edited by everling: Mar 22 2010, 08:34 PM

If your OS on another HDD, then you have the option of using the Computer Management in the Administrative Tools to delete all the partitions in that drive.

Alternatively, the Windows install CD/DVD should provide you the option to repartition your drive. It allows you to delete existing partitions and create new partitions.

This post has been edited by everling: Mar 23 2010, 02:04 AM

You can try your luck with a third computer, but there is a good chance it is broken if two computers can't detect it. If there are data that needs to be recovered, then you'll need to engage the services of a data recovery company and it won't be cheap.

I can only guess at what might be the problem. Try this:

System Properties > Hardware > Device Manager > IDE ATA/ATAPI Controllers > Primary IDE Channel > Advanced Settings

Check if it's in PIO mode. If so, switch to DMA if available. Do the same for the Secondary IDE Channel.

I do not know about the jumpers, but the lag is because of Windows XP.

Windows XP expects 512 byte sectors. Any advanced format drives, which is the new modern standard slowly gaining in the market, uses 4K byte sectors. So when XP wants to write 128 bytes, in the previous format the HDD just writes it but the new HDD must first read the 4K byte sector and then write 4K bytes. Rough performance loss in XP is about 10% average; it's not worse because XP has no issue with reading and I think it is fine with large sequential writes.

The solution to your problem is to either upgrade to Windows 7 or any 4K sector aware OS, go back to the shop to downgrade your hard disk or ignore the performance loss. There is another possibility that does mitigate the problem a little bit, something about aligning Windows XP with the sector via a software utility from WD. I don't know enough to know if you have already done it as I'm avoiding the 4K sector HDDs for now. SSDs have the same problem and they do perform much better with 4K sector aware OSes as well.

The new 4K byte sectors is indeed an advance across the board (performance, reliability, storage efficiency), except for OSes that expects 512 byte sectors (losses on performance).

This post has been edited by everling: Mar 30 2010, 10:04 AM

Yes.

You mean about Win7 working fine on 512 byte sectors and 4K byte sectors? Yes.

But when it comes to Win XP, you might want to prefer 512 byte sector drives over the 4K byte sector drives.

Any decent 500 GB/platter @ 7200 RPM can indeed beat the Velociraptor on sequential read/write because of it's higher data density. The Samsung F3 series is an example. But the Velociraptor would still win on random access time when compared to platters. The only thing that would clearly beat the Velociraptor on all performance metrics are SSDs, with the exception of Intel SSDs on write performance.

I had forgotten about short stroking. It's not often that people would willingly throw away 80% of the capacity away just to get higher performance. I wonder if you can get similar results by benchmarking only on the first 300 GB partition. If yes, then you can still have the remaining 1.2 TB.

Air cond room is no proof. Otherwise they wouldn't bother to spend so much money on clean rooms and bunny suits.

1000 megabit/s / 8 bits = 125 megabytes/s

It's going to be a bit less in practice but should be much more than 10-20 MB/s.

Don't really know. Perhaps your router supports only 100 Mbps?

100 megabit/s / 8 bit = 12.5 megabyte/s

This post has been edited by everling: Apr 16 2010, 01:38 AM

It is not nice and neat however.

Keep in mind that the 1 TB Samsung is already notably cheaper than the 1 TBs from other manufacturers. And now we have a 750 GB Samsung that is 66% of the 1 TB's pricing when it should be closer to 75%? It's a little strange when you consider the platters.

Iianm, there are 500 GB, 333 GB and 250 GB platters with nothing else in between. To make a 750 GB, you'd need two 500 GB, three 333 GB or three 250 GB platters. This suggests either they're making a loss, using defective platters or using leftover platters. If the 750 GB is indeed three 250 GB platters, the performance will be significantly less than the F3s. If it performs the same as the F3s, then it's using the 500 GB platters, the platters are likely to have some defect and Samsung is making a loss.

Try Recuva.

1.5 TB and 2 TB drives are assuredly 500 GB platters because nobody has produced a five or six platter drive (eg: 5x 333 GB or 6 x 333 GB) in a long while. Usually tops out at four platters a drive.

Other than that, best way is to read the reviews as platter size is something usually mentioned nowadays.

Depending on your SATA drives. For average hardware, it's okay. For those like Samsung's F3, it's a bit slow. For SSDs, it may be slow.

I don't know what the ST means. But depending on which model of the 500 GB, it can be much slower or a bit faster than then WD Green 1 TB 64MB cache (WD10EARS). WD is very annoying in that way.

False.

Maximum theoretical transfer rates:

• SATA 1.5 Gbps: 150 megabyte/s
  
• SATA 3.0 Gbps: 300 megabyte/s
  
• SATA 6.0 Gbps: 600 megabyte/s

HDD: WD1002FAEX maximum transfer rate to disk: 126 megabyte/s
HDD: ST320005N1A1AS-RK sustained data transfer rate: 138 megabyte/s
SSD: Intel X-25M G1 160 GB sequential read: 259.2 megabyte/s

That's probably because of the burst transfer. Instead of reading from or writing to the disk, your computer reads from and write to the HDD's limited cache. That can go much higher than the sustained transfer rate, but only for short bursts as the cache will get exhausted pretty quickly.

This post has been edited by everling: May 6 2010, 11:26 AM

Nice report. How did you get it?

How did you generate the report? I don't recognise the output format.