Where is serial ATA???

[EnragedPC]

I want my serial ata....1.5gigs of sustained data.....drooooll...say goodbye to all regular ATA drives after that day and toss out the SCSI...how long till the is finally produced?

 

[toph99]

1.5gb of bandwidth, but drive now can't even max out the ATA100 spec

i just want to be rid of ribbon cables :disgust:

 

[EnragedPC]

no its a new way of data transfer...it will be 1.5 gigs sustained so ATA/100 is nothin...i think todays way is called parallel transfer?

 

[Viper GTS]

EnragedPC:

The drive itself has to be able to pump out data that fast. Except for 2 meg cache dumps, there isn't a drive in existance (except maybe solid state drives at $30K a pop) that can fill that pipe.

Viper GTS

 

[EnragedPC]

dont you think that when the actual spec comes out the HD's to support it will to?

 

[Viper GTS]

ATA100 is out, & we don't have drives that can sustain 100 mbps transfer rates. HD's have a long way to go before they'll reach that level.

Viper GTS

 

[toph99]

i think what they need to do(correct me if i'm wrong) is to add multiple reading heads(so that 2 parts of the disk can be read/written on at once) and to increase disk rotation speeds

 

[Phil21]

they only use for a drive bus that massive is in RAID systems where multiple drives are accessed concurrently. I've maxed out the SCSI LVD bus with RAID5 arrays before, so something that could truly work the drives out would be welcome.

but, I know nothing about serial ATA. If it's anything at all like IDE/UDMA right now it will truly suck.

-Phil

 

[Viper GTS]

Phil:

The bandwidth will be MUCH higher than EIDE, & the biggest advantage I see is we can finally kiss ribbon cables goodbye. HD cables will be something like a CD audio cable. Now that's cool!

Viper GTS

 

[vi edit]

I think it would be very difficult to get data pumped out that quick so long as you have mechanical devices moving inside of a drive rotating a platter or moving a needle.

As Viper said, solid state drives can probably do it now, but they are mucho expensive.

 

[Sohcan]

EnragedPC: As others have pointed out, the speed of a harddrive is not determined by the bandwidth of the bus, unless that is the limiting factor....which it rarely is. The speed of a hard drive is largely determined by the rate at which it can read data off the surface of the disks; this is purely dependent on the rotational rate of the hard drive and the density of the data on the disks. The current fastest IDE drive (IBM 75GXP) has a max throughput of 37 MB/sec.

Also, don't forget that the serial ATA spec is 1.5 gigabits/sec (188 MB/sec), vs. ATA/100's 100 MB/sec.

 

[Phil21]

Viper: Yes, the cables will truly rock! But, I was more pointing towards the underlying technology. Is it a leap forward in protocals? Or is it using the same protocal as IDE/ATA100? If so, it's fairly worthless IMO besides the cable type thing. But, if it adresses the limitations of IDE such as pathetic number of device support (need more than 2 per bus. duh!), and the ability to access multiple devices at the same time (like SCSI), then it will seriously be a VERY welcome step up.

But, if all it is, is simply a cable change, and not much more.. It's usefullness will be short lived, and I doubt it will be adopted much by hardware vendors since it offers hardly any gains.

Anyways, I suppose I could actually go and do some research on this, instead of speculating about it.. huh?

-Phil

 

[Mday]

it's 2002 already?

 

[RGN]

I want to kiss my ribbon cables goodbye.

<-- that thought makes me...

 

[urameatball]

I like my ribbons.
they're about the cheapest things you could get with your computer

 

[Viper GTS]

Serial ATA will be even cheaper than ribbons. Ribbon cables are a pain in the ass to route, they obstruct airflow, &amp; they're ugly as hell.

Serial ATA will be a welcome change, even if it didn't offer any performance increase at all. The price drop &amp; convenience will be enough for me.

Viper GTS

 

[spidey07]

By far the limiting factor in hard drive technology is physical. Drives can only spin so fast and still be read accurately. Someone mentioned &quot;well why not have multiple read heads?&quot;. that has been done for more than ten years. A lot of your larger SCSI drives have multiple platters 6-8 that are read by 12-16 heads top/bottom.

No matter how fast the interface is the physical limitations still exist and are still the bottleneck.

Need to go into my idea of buckyballs holding ions for storage. works great. What ever happened to crystal/light computers?

 

[~zonker~]

It will be coming soon, they're still resolving the sepcifications ... check here, Seial ATA Working Group