Pariah will have a point... the moment 15krpm SATA drives appear...
And to say it has to benefit the 'desktop' user to be innovative or useful is bullshite.
Calculate number of mom and pop and grandma types... using crap like AOL.. the biggest chunk of 'desktop users' by FAR.
Why have broadband? Why have SATA? Why did we ever even get away from those ST-506 based HDDs?
And to say it has to benefit the 'desktop' user to be innovative or useful is bullshite.
Calculate number of mom and pop and grandma types... using crap like AOL.. the biggest chunk of 'desktop users' by FAR.
Why have broadband? Why have SATA? Why did we ever even get away from those ST-506 based HDDs?
- Nov 11, 2004
- 12,576
- 7
- 81
Differant strokes for differant folks, the biggest bottleneck on any modern system is the HARD DRIVE, It won't be long before SCSI becomes a bottleneck of it's own. HDD's will one day hit 300 MB/s per drive considering they share a common bus that is very limited on bandwith. SAS at this point is only a preventative measure, just like when SATA arived there was no reason to ditch PATA when it came out.
I am excited about this technology and all technolgies even if I may or may not ever use it in my home system. SAS is needed to address the shortfalls of the original SCSI. 16384 Drives is more than most corporations will ever have in on system, so it is kind of nice to see a big number and comfortably say you may never run out of ID's for your drives.
By the way SAS is 3GB/s not 300MB/s
What? I ask you what new features SAS has that current SCSI doesn't that are interesting and will benefit us, and you tell me controller compatibility with SATA drives. I tell you we already have free SATA controllers on our motherboards making the feature redundant. And you respond by saying you can't run SAS drives on a SATA controller? What on earth does that mean? So in a round about way, you're telling me I should be interested in SAS because I need a SAS controller to run a SAS drive? That's exciting. I take back everything I said before. NOW, I'm excited about SAS with that knowledge in hand.
Gamingphreek
Lifer
- Mar 31, 2003
- 11,679
- 0
- 81
Doubt it. Do you know what kind of shielding they have to use to protect an Ultra Centrifuge. When one broke i literally bent the IIRC 3 Inch thick steel shielding. Ill eventually find the story.
Additionally unless we start using a different bearing and different platters it isn't happening as currently they cannot go that fast.
SAS sounds interesting. New technologies are so cool
-Kevin
We'll never see 30k RPM drives. We may or may not even see faster spindle speeds than 15k, as the point or diminishing returns has pretty much been reached with 15k. Anything faster won't really be faster. Continuing to increase areal density, and physically shrinking the platter (the shift to 2.5" drives has already begun with Savvio) is a far cheaper and mechanically less challenging way of increasing performance than continuing to try an increase the spindle speed.
- Nov 11, 2004
- 12,576
- 7
- 81
Originally posted by: Gamingphreek
Doubt it. Do you know what kind of shielding they have to use to protect an Ultra Centrifuge. When one broke i literally bent the IIRC 3 Inch thick steel shielding. Ill eventually find the story.
Additionally unless we start using a different bearing and different platters it isn't happening as currently they cannot go that fast.
SAS sounds interesting. New technologies are so cool
-Kevin
There are a ton of non computer devices that spin at 30k or more RPM's, so I see no reason why a HDD could not do the same. Nothing worse than a pessamistic doubter, they never acomplish much in this world.
Devices that spin at 30k rpms or more
Routers (not related to networking but woodworking)
Dremel Mototools
Turbochargers spin at 100,000 rpms or more
Gas Turbines; TurboJet and TurboProp engines rev past 30k rpms
Making a platter spin at 30k RPM's is not a challenge. Of the items you listed above, how many of them have a read head floating above them at a fraction of the height of a finger print trying to read magnetic fields that are less than 1/60 billionth of a inch^2? And again, even if they thought it was possible, which eventually it will be, the minimal performance increase would make it pointless to even try in the first place.
- Nov 11, 2004
- 12,576
- 7
- 81
30,000 rpms may be many years away, but 20,000 is perfectly feasable and attainable. Also I believe it is air from the platters spinning that keeps the heads floating so close to the platters. Improvements in head technology will alow for greater arieal densities regardless of the rpm.
#1 Bill Gates never said that. #2 If you don't understand why continuing to increase the spindle speed is a waste of effort, then you don't understand how hard drives work and what effect spindle speed has on performance.
- Nov 11, 2004
- 12,576
- 7
- 81
He did say that back in the 1980's, it is well documented. I may be wrong to say 256k may be it was 384 but I know it was far less than 512k.
He said that about the internet back in 1990 or 1992.
I do understand how HDD's work, but I have not heard anything about performance diminishing after 15,000 rpms, could you point me to an article from a magazene or statement from a company ? (like IBM, WD, Maxtor, Fujitsu, Hitachi, or Seagate)
By increasing the spindle speed from 15k to 30k you cut rotational latencies by nearly half.
You read it on the internet, so it must be true, right?
Gates talks
Correct, well, sort of. If you double the rotational speed, physics says the average latency has to be exactly half. Now look at the following:
4,200 RPM -- 7.1ms average latency
5,200 RPM -- 5.8ms average latency
7,200 RPM -- 4.2ms average latency
10,000 RPM -- 3.0ms average latency
15,000 RPM -- 2.0ms average latency
30,000 RPM -- 1.0ms average latency
Do you see the problem yet? Or do I need I still need to spell it out for you?
Serious question, are you under the age of 15?
Gates talks
Correct, well, sort of. If you double the rotational speed, physics says the average latency has to be exactly half. Now look at the following:
4,200 RPM -- 7.1ms average latency
5,200 RPM -- 5.8ms average latency
7,200 RPM -- 4.2ms average latency
10,000 RPM -- 3.0ms average latency
15,000 RPM -- 2.0ms average latency
30,000 RPM -- 1.0ms average latency
Do you see the problem yet? Or do I need I still need to spell it out for you?
Serious question, are you under the age of 15?
Zebo
Elite Member
- Jul 29, 2001
- 39,398
- 19
- 81
still use all PATA here.. I'm actually quite embarresed when I recieve a PM asking me how to setup thier SATA HD's or overclcoking with SATA and can't answer.
Maybe I should switch but PATA Seagates are quieter than their SATA versions.
Ya that's what I want inside my computer, a router noise.
Maybe I should switch but PATA Seagates are quieter than their SATA versions.
Ya that's what I want inside my computer, a router noise.
I know, I used to have a bet with my MS-bashing friends to find me an article where he had even made anything
close to that statement. Not to prove them wrong mind you, but I had heard that same claim so often that
I really wanted to find out where it might have originated if true.
[/quote]
Correct, well, sort of. If you double the rotational speed, physics says the average latency has to be exactly half. Now look at the following:
4,200 RPM -- 7.1ms average latency
5,200 RPM -- 5.8ms average latency
7,200 RPM -- 4.2ms average latency
10,000 RPM -- 3.0ms average latency
15,000 RPM -- 2.0ms average latency
30,000 RPM -- 1.0ms average latency
Do you see the problem yet? Or do I need I still need to spell it out for you? [/quote]
One issue though, as you already pointed out, drives can be built with smaller platters to create an additional reduction in latency.
You already mentioned the Savvio drives, but IIRC the Cheetah line has been taking avantage of smaller platters for several
generations. Your calculations above are all based on keeping the drive volume constant, yes?
Oh, and you wanted a good reason for SAS? It will be so we can have fuel for the great SATA vs SASI
debate of 2006.
You know we're long overdue...
I don't see desktop users ever using SAS unless they get a free controler card. The only advantage that SAS has over SCSI for people who would use it is that it will take cheap SATA drives and make them usable on the same controler and in arrays w/ faster drives like Raptors and current SCSI drives making managment simpler, oh and the cables are prolly gonna be thiner and cheaper. Other than those couple points really no advantages to SAS, it's basicly a SCSI replacment that you might get if your getting a new system but otherwise shouldn't bother with. Besides theirs always Fibre Channel if you want real performance.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 23K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter