Website, CSF, and lots of attacks.

[keeney123]

Re: Website, CSF, and lots of attacks.

Things are identical with SSDs apart from not requiring to defrag the drive.
They have wear leveling built in to spread the wear across the cells, the data is spread across the drive. It makes no difference to read speed.

I would say there is a difference as with a HD it has to spin up the read head arm has to find where the data is on the disk. With a SSD it is a matter of an electronic address.

[diif]

Re: Website, CSF, and lots of attacks.

I would say there is a difference as with a HD it has to spin up the read head arm has to find where the data is on the disk. With a SSD it is a matter of an electronic address.

The computer still has an address for the data on a hard drive, but with greater fragmentation those parts are scattered across the platter/s and it takes time to seek the data and then read it.

[keeney123]

Re: Website, CSF, and lots of attacks.

The computer still has an address for the data on a hard drive, but with greater fragmentation those parts are scattered across the platter/s and it takes time to seek the data and then read it.

I would say that even if there was 0 fragmentation a mechanical device is going to be slower than one that is electronic and having tracks on the inside platter and ones on the outside platter have to be considered on a hard drive. I think top maximum speed is 10,000 rpm's. A hard drive was part of the bottle neck and as a result we have the SSD. Complete load time for Widow 7 Pro and all other associate drivers and programs on my Legacy T400 is 53 sec. with a HD it would be a the minimum 120 sec.

[Spork Schivago]

Re: Website, CSF, and lots of attacks.

Things are identical with SSDs apart from not requiring to defrag the drive.
They have wear leveling built in to spread the wear across the cells, the data is spread across the drive. It makes no difference to read speed.

Gotcha, but it does matter, right? In the sense with a normal harddrive, you don't require the wear leveling, right? And although the user doesn't do it, if the data wasn't equally written across the drive like it is (with solid states), the drive would fail much quicker, wouldn't it?

The reason it doesn't affect read speed is because of the way they're designed, isn't it? No moving parts, no need to search for the sector, etc.

[Spork Schivago]

Re: Website, CSF, and lots of attacks.

I would say that even if there was 0 fragmentation a mechanical device is going to be slower than one that is electronic and having tracks on the inside platter and ones on the outside platter have to be considered on a hard drive. I think top maximum speed is 10,000 rpm's. A hard drive was part of the bottle neck and as a result we have the SSD. Complete load time for Widow 7 Pro and all other associate drivers and programs on my Legacy T400 is 53 sec. with a HD it would be a the minimum 120 sec.

Keeney123, I might be misreading what Diif is saying, but I don't think he's saying there's no difference in read speeds when it comes to comparing solid state hard drives with non-solid state hard drives. If I'm understanding him correctly, I think what he means is:

With a normal hard drive, reading sector 1 would be quicker than reading sector 107. With a solid state hard drive, reading sector 1 is just as quick as reading sector 107.

The latest bottleneck is (was?) the SATA interface I believe. They have M.2 now which is even faster than SATA, if it's implemented correctly. If I understand it all, I think some companies use an M.2 connector but are just using the SATA stuff, so you only get speeds up to SATA 3 or something. But if it's done properly, you can speeds faster than SATA 3. I get a little confused in that area because I haven't had a lot of time to research it much.

[diif]

Re: Website, CSF, and lots of attacks.

If I'm understanding him correctly, I think what he means is:

With a normal hard drive, reading sector 1 would be quicker than reading sector 107. With a solid state hard drive, reading sector 1 is just as quick as reading sector 107.

Yes, that's what I was trying to say. There is the seek time and rotational latency to deal with in hard drives but not with SSDs.

M.2 is a standard that can connect to PCI Express, SATA and USB.

They do 15,000rpm hard drives Keeney.

[keeney123]

Re: Website, CSF, and lots of attacks.

Yes, that's what I was trying to say. There is the seek time and rotational latency to deal with in hard drives but not with SSDs.

M.2 is a standard that can connect to PCI Express, SATA and USB.

They do 15,000rpm hard drives Keeney.

I guess I was not understanding you correctly, sorry.15,000 rpm that is impressive. The article I had read some time ago said the highest usable top end speed would be 10,000 RPM. I wonder if the 15,000 rpms is actually faster in operational use?

[stj]

Re: Website, CSF, and lots of attacks.

it is faster, they have been around for many years - if you use scsi.
i have some.

https://www.cnet.com/forums/discussi...-drive-100519/
notice the date on that!

http://www.seagate.com/gb/en/interna...mance-15k-hdd/

[Spork Schivago]

Re: Website, CSF, and lots of attacks.

...M.2 is a standard that can connect to PCI Express, SATA and USB...

This here...so M.2 is a standard and the standard sas it can connect to PCI-E, SATA or USB? It's up to the designers which interface to connect them to?

When they connect the M.2 socket to the SATA interface, they're limiting the maximum transfer rate of the drive to whatever the SATA interface is, correct? When they connect it to the USB interface, same thing. PCI-E is the fastest out of all of them, right?

Is there a good reason for buying a board that has an M.2 socket connected to the USB or SATA interface instead of the PCI-E interface, besides maybe cost?

With the PCI-E interface, we can have NVMe which fully utilizes the high speed PCI-E stuff and allow for that parallel operations, which we can't get with SATA and USB. Is that right?

[Spork Schivago]

Re: Website, CSF, and lots of attacks.

I guess I was not understanding you correctly, sorry.15,000 rpm that is impressive. The article I had read some time ago said the highest usable top end speed would be 10,000 RPM. I wonder if the 15,000 rpms is actually faster in operational use?

The 15,000 RPM drives are really fast. I think they're more or less used in servers and usually come in SCSI form, but don't quote me on that. They're also expensive, or at least when I was looking at them they were.

Regardless though, M.2 implemented on the PCI-E bus should be the fastest now I believe.

A 15,000 RPM hard drive is theoretically 50% faster than a 10,000 RPM hard drive. It's theoretically around 100% faster than a 7,200 RPM drive.

[Spork Schivago]

Re: Website, CSF, and lots of attacks.

it is faster, they have been around for many years - if you use scsi.
i have some.

there are also sata ones now - raptor series i think they are called.

You beat me to it. I didn't see your post. Sorry. They have SATA 15,000 RPM drives now? I'll check out the Raptor series.

[diif]

Re: Website, CSF, and lots of attacks.

I thought Raptors were 10,000.
An SSD that is fully compatible with PCI-E 3.0 x 4 will be the fastest type with a theoretical speed of 32Gb/s (4GB/s).

[Spork Schivago]

Re: Website, CSF, and lots of attacks.

I thought Raptors were 10,000.
An SSD that is fully compatible with PCI-E 3.0 x 4 will be the fastest type with a theoretical speed of 32Gb/s (4GB/s).

Diif, you know a lot about hard drives. Let's say I have a 15,000 RPM SATA 3.0 (6Gb/s) drive and a 7,200 RPM SATA 3.0 drive. The 15,000 RPM drive spins more than twice as fast, but isn't it still limited to the 6 Gb/s? I'd be able to retrieve files faster from the drive but the transfer rate would never go over the 6 Gb/s, right?

[stj]

Re: Website, CSF, and lots of attacks.

in theory, a 15,000rpm harddrive could write data much faster than an SSD, but in most cases, not read as fast.

the interface is another matter, because you also have to look at the pc side.
a lot of these interfaces are already faster than the data can be dma'd into the ram.
.

[diif]

Re: Website, CSF, and lots of attacks.

There aren't any 15000rpm SATA drives as far as I am aware.
A faster drive has a faster access time and IOPS input/output operations per second.
Whilst a drive might be connected to a SATA 3, that's not its read speed. Thats the theoretical max speed.

[keeney123]

Re: Website, CSF, and lots of attacks.

So it is a hybrid drive not just a hard drive. I tried to read up on the platform but the download was so slow I could of gone to sleep.

[Spork Schivago]

Re: Website, CSF, and lots of attacks.

There aren't any 15000rpm SATA drives as far as I am aware.
A faster drive has a faster access time and IOPS input/output operations per second.
Whilst a drive might be connected to a SATA 3, that's not its read speed. Thats the theoretical max speed.

Right, I understand the theoretical max speed and you won't actually get 6Gb/s. For simplicity sakes, I didn't clarify. So, with a 7,200 RPM and a 15,000 RPM drive, the transfer rates will still be the same, right? You'll just be able to access the data quicker. But once that data starts transferring, it'd be the same speed, right? For example, let's say I have an M.2 x4 drive hooked up and then two SATA drives hooked up. Even though the 15,000 RPM might not exist for SATA, let's just pretend it does. If I timed how long it'd take to copy data from each of the SATA drives to the M.2, the 15,000 RPM drive would take less time, but if I measured the transfer speed, it'd be the same, in the perfect theoretical environment. The SATA interface would be limited the speeds to the 6Gb/s. Just the 15,000 RPM drive would be able to find the data quicker and start sending it, right? Or would the 15,000 RPM drive reach a faster transfer speed?

I'm having a hard time seeing how the speed of the drive affects overall transfer rates.

[stj]

Re: Website, CSF, and lots of attacks.

your talking about seek rates, that is in the drive specs

[keeney123]

Re: Website, CSF, and lots of attacks.

it is faster, they have been around for many years - if you use scsi.
i have some.

https://www.cnet.com/forums/discussi...-drive-100519/
notice the date on that!

http://www.seagate.com/gb/en/interna...mance-15k-hdd/

It actually is SAS using full duplex as opose to SATA using Half duplex. The SCSI is just for the device identifier. SAS operates at a high voltage then SATA. Which allow for multiple back planes. SCSI is the very old parallel bus. This parallel bus caused synchronization problem and also at the time in the 80's it did not have common specs through devices. Some printers used this connection but the the rs 232 port came out which were standard specs serial with all device except Apple and the printer Epson they used. As I remember they switched the wires of the transmit receive pins. This type of stuff that Apple did in its early years really cost them the market.

[diif]

Re: Website, CSF, and lots of attacks.

I'm having a hard time seeing how the speed of the drive affects overall transfer rates.

The data is scattered about the platter/s if it can seek it faster the IOPS increases.

Imagine a conveyor belt in an Amazon warehouse with two order pickers. One is a little faster on his feet than the other, the parcels still move along the conveyor at the same speed but there are more parcels with the faster picker.