Search results

There was some delay, but I finally got the ZIF-ATA adapter two weeks ago. As I assumed, the USB bridge influenced the write performance of the HU040HA. HU040HA ATA Performance That means that the 40GB HU040HA works perfectly if you have your data aligned to odd sectors, and use a cluster...

My point is that there is no NAND flash random write performance problem, as long as you respect the internal structure of the flash memory, just as you accept that harddisks come with a sector size of 512 bytes. It would be the same as if you would compare a RAM disk and a harddisk by...

I was refering to the irregular write results, without mentioning that, sorry. Harddisk benchmark tools assume that writes can be done in 512-byte increments at random locations without affecting the results. This is not true for most flash media.

I think the main problem is that you can't use harddisk benchmark tools on flash media. You would have to get detailed technical data for the SSD to know what you should test and what you should not test. Flash memory can only be erased in certain block sizes. If you write less than that...

That's absolutely correct. Normal flash cards assume to be filled up to the rim and then re-formatted. A modern SSD controller is supposed to keep a record for each cell, and swap static data from fresh cells on heavily used cells and try to keep all cells at about the same "age". If you want...

I could not resist and ordered the 30GB version. There seem to be different versions for the 30GB LaCie little disk, one with the 1024 bytes per sector HU030HA, and one with the 512 bytes per sector HU035HA. At least the one I got has the 512 bytes per sector HU035HA. It has a much cleaner write...

How is the controller attached to the mainboard? How many disks of which type?

Now I have more information and charts available. I tested the harddisk through another USB bridge, and it looks like I have to get a true 40-pin ZIF to full-size adapter to see the full write performance of the disk. Both USB bridges show some impact on the write performance, but the loss is...

Just an update regarding the write rate: As I assumed, the disk suffers from "misaligned writes" due to the fact that the disk stores 1024 bytes per sector, but acts as if it had the usual 512 bytes per sector. The factory format has the start of the data area (cluster 2) on an odd sector, so...

When I read about the new 1.3 inch Samsung SpinPoint A1 series, it reminded me of the old Microdrive days that were actually the reason why I started my website with harddisk information. The whole site was originally a list of small form factor harddisks that got seriously out of bounds ;-)...

Well, my VR has three screws with free access, the fourth is covered by a sticker. Maybe it would be OK to remove the three screws and try to use a thin saw to cut the fourth srew? About the claim that the VR would not get too hot: Did you notice the pads between the harddisk and the heat...

You should be aware that regular 32bit PCI has a maximum bandwidth of 127MB/s (binary), while PCIe x1 has 238MB/s (binary) - eSATA add-on cards are usually PCIe x1. That means that you would have to sacrifice your PCIe x4 slot for using a x1 card. However, I did not see any 2-port eSATA card...

I would recommend 2.5" harddisks. You can get 500GB at 12.5mm height, and the disks are designed to be carried around - unlike 3.5" harddisks. I would not trust DVD DL media, because it can be hard enough to read out a single layer media, so I would not take the risk of having two layers.

Well, that's the reason why they reduced the diameter of the platters. The latency is a *serious* problem with todays harddisks. Even the VR shows "poor" random access, simply because one full revolution takes 6ms - in words: SIX MILLISECONDS. That's more time than a full stroke! No matter...

Well, several years ago, some people thought that the 15,000RPM IBM disk would be a joke. It took an awful lot of time until they hit the market with a little less speed than originally advertized, but they made it! BTW: Mine is dated 23-DEC-01, unfortunately, it crashed too soon. Now 15kRPM...

The spindle speed has always been a sore point. If you compare harddisk development within the last 10 years or so, we have: Capacity 10GB -> 1TB: 100x Sustained data rate: 15MB/s -> 150MB/s: 10x Spindle speed: 10kRPM -> 15kRPM: 1.5x I have no values for full stroke, but I think that's...

In the past I was happy with the T166 HD501LJ. Now I tried the F1 series. I was able to get a "proper" read rate: HD103UJ Graph At least this is the actual read graph, not some quick scan or pretty-printed version. Please note that my website is completely out of date regarding the...

Then you have to try the Rescan method. This even works perfectly with the internal SATA ports of my old ICH6R chipset. It does not work on the SiI3114R ports, but I have to check whether a firmware update is available. I get the best performance using a PCIe x1 to ExpressCard adapter...

Ah, I see. Just from internal SATA connector to slot cover. If your SATA host really supports SATA-II, it should work. Try this: -Boot the system withtout the enclosure attached -Connect the enclosure -Open System Management, Device Manager -Select Scan for New Hardware If that works...

eSATA supports hot-plugging and automatic detection. What type of "eSATA" connection are you using? You can't just use a slot cover adapter with an old SATA-I host, for example. It also depends on the HAL of your OS regarding which devices it assumes to be attached and removed at "uptime".

Well, I am wondering how you can get 297.530 MB/s from *two* harddisks with a *maximum* sustained data rate of about 125MB/s each? There must have been some serious data caching involved. Why does everybody want to be a "tester" these days?

I got myself a PCIe x1 eSATA card. After a few tests, I removed it: The JMicron JMB360 controller on that card is slower than using a PCIe x1 to ExpressCard adapter and an Express/34 eSATA adapter using the SiI3132 controller. It's definitely more expensive, but more versatile... Edit: The...

Trying to take this question serious, I'd call it "Possible to add an Ipod to a Velociraptor? a) Power Requirements. According to the label, the VR needs 5V plus 12V. You'd have to add a battery pack or DC/DC converter. b) Interface If it's an older iPOD, it uses a custom PATA...

Well, those 1,000,000 write cycles are usually stated for industrial grade CF. For consumer grade SLC CF, usually 100,000 write cycles are mentioned, while consumer grade MLC flash are stated with 10,000 write cycles. Using regular flash cards as SSD is like putting regular tyres on a Formula...

Looks like I changed the wording in that sentence while typing. Of course I meant "... but be prepared to see very high (fast) read rate and fast read access times ...".

For an overview regarding UDMA CF cards, see my website: http://www.hjreggel.net/cardspeed/index.html#cs_udmacf.html See the links below the first table for test results. Please note that flash media can only be erased based on a page size of about 16kB. I would strongly suggest using a...

Well, current 32GB CF are MLC flash and you will have slow writes and delay due to re-allocation. Most of the fast SLC CF are available at only 8GB at about 80-150 Euro, where the better ones are more in the 120 Euro range. I don't know about 8xSATA RAID controllers, but the 8x8GB CF with...

I think at least some older Fujitsu-Siemens UNIX based systems used to have harddisk with custom firmware. So the flasher software might be available for Solaris only - but that's just a guess...

Ouch! Regular 32bit PCI? Are you aware that it has a *shared* bandwidth of 133MB/s? You'd be better off with a PCIe x1. If 32bit PCI is all you've got, then the Adaptec 29160 is probably the best you can get. But for the price of the controller, you can get a 150GB Raptor and most likely...

If you are wondering why those cards are PCI-X, then you should better forget about the whole thing. If you still insist: Forget about Hardware RAID, get an Adaptec 29160, but select the right model for your purpose. Or asking the other way round: What slots do you have available?

Well, I'd say that this is just a matter of interpretation. Computer guys define 1kB "one kilobyte" being 1024 Byte. That's "one" - "kilobyte", not "one kilo" - "bytes" - or: (1)(kB), not (1k)(B). The kilo in kilobyte is not the SI prefix, the whole kilobyte is a new unit. It's a little bit like...

I haven't seen a labeled USB connector, most of them just have a tiny triangle mark, but that one is perfect: You can clearly see +5V, D-, D+, GND. Get your mobo manual and check the USB headers. Sometimes there are 10-pin USB headers for 2 ports.

Ah, I see. The discussion is about the "burst speed" bar, not the "Sequential Read" graph. Then I agree, it's important to know that the indicated bust speed is misleading. Sorry for the confusion...

It's not about whether it makes sense to you. It's about how harddisk work. The MAX series can't go faster than about 95MB/s. No matter if you use U160, U320 or U640 SCSI, SAS or whatever. Not even if you get a 64bit 133MHz PCI-X card. The MAX is still a LMR harddisk. Only the new PMR...

Could you explain a little what seems so shocking to you? What would you expect? The graph looks normal to me. The sustained data rate of a harddisk is limited by the RPM and the sectors per track. Using a faster interface does not make your harddisk faster.

Slimline drives should work as single drive on a channel. But there are some models where different firmware for master, slave, CS and inverted CS is available. Check rpc1.org for information about your drive.

3.5" WD drives are typically rated 65G operational shock, 250G non-operational shock. 2.5" harddisks are typicall rated 250-300G operational shock, 800-1000G non-operational shock. Do you still call my statement "greatly exaggerated". You should definitely check some harddisk datasheets in...

Hi, Thanks for the suggestions, I know that these disks are old. I already have a replacement disk and I have backup of all the data. The problem is the time to set up the new installation. I can't simply move, because I would have to merge numerous partitions from three disks to one disk...

Hi, No. Laptops use 2.5" harddisks that are built for that purpose. The OP had a 3.5" harddisk. A 2.5" harddisk has about the same or more operational shock tolerance as the non-operational shock for 3.5" harddisks. This means that you can kill a *switched-off* 3.5" harddisk with a knock...

Hi, I would say that tilting a spinning 3.5" harddisk could kill it. 3.5" harddisks are built for stationary use, not for being lugged or tossed around. But you should check all cables and fans first. Hans-Jürgen