HD advice please...

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

kony wrote:

>>About 2 years ago, they tried to reach the ceiling of an ATA133 bandwith,
>>using 4 of the (then) fastest Maxtor (the only one supporting ATA133)
>>drives in RAID 0, and even compared that to WD and SCSI drives with 4
>>drive RAID. The SCSI was the fastest, by a slim margin and won certain
>>tests, but lost in some, and the WD drives, which were ATA100 were the
>>slowest. The highest they were able to acheive was 119MB/s sustained
>>throughput on the ATA133 setup with 4 drives. Drive throughput has
>>increased since then, but not by that much...
>
> 119MB/s is an ATA133 and/or PCI limit, not because the
> drives couldn't go any faster.

Not the PCI limit, because the SCSI drives posted faster speeds than the
133MB/s and it had to go through the PCI bus, as well.

Here's a link to a recent disk comparison done by Tom's Hardware:

http://www.tomshardware.com/storage/20040820/hitachi-maxtor-09.html

Note the minimum read transfer performance scores. The highest is the Raptor
74MB @ 42MB/s. The fastest write score was the Seagate 7200.7 Plus @ 32.8
MB/s. This is the physical transfer rate of the drive. Anything higher is
burst rate, or a combination.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,uk.comp.homebuilt (More info?)

On Thu, 07 Oct 2004 04:38:43 GMT, "Ron Reaugh"
<rondashreaugh@att.net> wrote:


>> The PCI bus can be limiting for even a median performing
>> single new drive.
>
>NOT under ordinary usage since current drives do about 60MB/sec. max.
>


Depends on what you call "ordinary", most people don't need
60MB/s for office work or email, but they do have a few PCI
devices. I wouldn't consider GB networking unusal today
either, it doesn't even make sense to buy a 10/100 card
anymore. Sure, a low-end box may use integral 100Mb NIC,
but how many other areas can you improve by 350% at a cost
of merely $18 except for the switch. I even hear of people
paying more than that for a crossover cable when the NIC
would eliminate the need for one.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,uk.comp.homebuilt (More info?)

In article <lJ39d.672417$Gx4.88642@bgtnsc04-news.ops.worldnet.att.net>,
"Ron Reaugh" rondashreaugh@att.net says...
>
> "Rob Morley" <nospam@ntlworld.com> wrote in message
> news:MPG.1bceab1320dcb34598a093@news.individual.net...
> > In article <tM%8d.671750$Gx4.535247@bgtnsc04-news.ops.worldnet.att.net>,
> > "Ron Reaugh" rondashreaugh@att.net says...
> > >
> > > "Rob Morley" <nospam@ntlworld.com> wrote in message
> > > news:MPG.1bce81d661fcb17198a08b@news.individual.net...
> > > > In article
> <jmY8d.671128$Gx4.471061@bgtnsc04-news.ops.worldnet.att.net>,
> > > > "Ron Reaugh" rondashreaugh@att.net says...
> > > > >
> > > > > "Rob Morley" <nospam@ntlworld.com> wrote in message
> > > > >
> > > > > > If your motherboard doesn't have a SATA controller on the
> southbridge
> > > > > > then the bandwidth will be limited by the PCI bus anyway.
> > > > >
> > > > >
> > > > > Well, the only time that there will be any PCI limiting is in a
> > > multi-drive
> > > > > RAID 0 configuration. The PCI bus can do about 100MB/sec. and the
> > > fastest
> > > > > of today's SATA drives do about 60MB/sec.
> > > >
> > > > But the drives don't have exclusive use of the PCI bus - gigabit
> > > > ethernet can saturate the PCI bandwidth on its own.
> > >
> > > Right, if gigabit is on the PCI bus where it is NOT on many recent
> mobos.
> >
> > Earlier you wrote "the only time that there will be any PCI limiting is
> > in a multi-drive RAID 0 configuration".
>
> Right, read the thread title. Here's what the OP said in his OP: "Q1 -
> Would i notice a speed increase if I went SATA 150 ?? are there any
> real world benefits to SATA other than thinner cables ??"
>
> In that context my answer was correct.

Had you written "in a real-world situation you may not encounter this
limitation" then you would have been correct.
>
> > I was just pointing out that
> > this might not be so.
>
> And in context you were mistaken or trolling or both.
>
Stop wriggling.

 

[Dee]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Are you ever going to figure out how to set the date and time on your
computer? Or are you just going to continue through like with your head
buried firmly up your ass?

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Dee wrote:

> Are you ever going to figure out how to set the date and time on your
> computer? Or are you just going to continue through like with your head
> buried firmly up your ass?

Date and time? What the hell are you talking about? It's setup fine here...

Or are you just trolling?

Interesting that you didn't have a response to my post about the Maxtor
warranty thing. Did you figure it out? Or are you still clueless?

 

[Dee]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Ruel Smith wrote:

> Dee wrote:
>
>
>>Are you ever going to figure out how to set the date and time on your
>>computer? Or are you just going to continue through like with your head
>>buried firmly up your ass?
>
>
> Date and time? What the hell are you talking about? It's setup fine here...
>
> Or are you just trolling?
>
> Interesting that you didn't have a response to my post about the Maxtor
> warranty thing. Did you figure it out? Or are you still clueless?
>
And I'm not clueless about the Maxor warranty! You're the one who
wasn't comprehending what was in plain English, so I figured there was
not use trying to deal with ignorance!!!

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Ruel Smith wrote:

> Dee wrote:
>
>
>>Are you ever going to figure out how to set the date and time on your
>>computer? Or are you just going to continue through like with your head
>>buried firmly up your ass?
>
>
> Date and time? What the hell are you talking about? It's setup fine here...

The date on your posts are a day earlier than when they show up.

For example, the date on the one above is shown as 10/6/2004 but today is
Oct 7.

Yours are the only ones I see that are off like that.

>
> Or are you just trolling?
>
> Interesting that you didn't have a response to my post about the Maxtor
> warranty thing. Did you figure it out? Or are you still clueless?
>

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

David Maynard wrote:

>> Date and time? What the hell are you talking about? It's setup fine
>> here...
>
> The date on your posts are a day earlier than when they show up.
>
> For example, the date on the one above is shown as 10/6/2004 but today is
> Oct 7.
>
> Yours are the only ones I see that are off like that.

Yeah, I fixed it. I'm running Linux and recently got it installed and
everything. I didn't have it setup properly. My mistake...

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

Ruel Smith wrote:

> David Maynard wrote:
>
>
>>>Date and time? What the hell are you talking about? It's setup fine
>>>here...
>>
>>The date on your posts are a day earlier than when they show up.
>>
>>For example, the date on the one above is shown as 10/6/2004 but today is
>>Oct 7.
>>
>>Yours are the only ones I see that are off like that.
>
>
> Yeah, I fixed it. I'm running Linux and recently got it installed and
> everything. I didn't have it setup properly. My mistake...
>

Yup. Looking good now.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Wed, 06 Oct 2004 16:51:37 -0400, Ruel Smith
<NoWay@NoWhere.com> wrote:


>> 119MB/s is an ATA133 and/or PCI limit, not because the
>> drives couldn't go any faster.
>
>Not the PCI limit, because the SCSI drives posted faster speeds than the
>133MB/s and it had to go through the PCI bus, as well.

Not necessarily, often PATA drives can't burst higher than
that either, showing it to be a limit other than the STR.

>
>Here's a link to a recent disk comparison done by Tom's Hardware:
>
>http://www.tomshardware.com/storage/20040820/hitachi-maxtor-09.html
>
>Note the minimum read transfer performance scores. The highest is the Raptor
>74MB @ 42MB/s. The fastest write score was the Seagate 7200.7 Plus @ 32.8
>MB/s. This is the physical transfer rate of the drive. Anything higher is
>burst rate, or a combination.

Yes, minimum... why are we focusing on minimum? Most people
try not to completely fill a drive then put
performance-limiting data on the slowest part of a drive.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

kony wrote:
>
>
> There may've been something wrong with the test then, I can
> exceed 35GB/s even coping over a LAN with single drive,

I would be VERY interested to know what LAN you are on about here.

I doubt you exceed 1/1000th of 35GB/s (gigaBYTES and not gigaBITS)
effectively.


Odie
--


RetroData
Data Recovery Experts
www.retrodata.co.uk

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Fri, 08 Oct 2004 04:26:54 +0100, Odie Ferrous
<odie_ferrous@hotmail.com> wrote:

>kony wrote:
>>
>>
>> There may've been something wrong with the test then, I can
>> exceed 35GB/s even coping over a LAN with single drive,
>
>I would be VERY interested to know what LAN you are on about here.
>
>I doubt you exceed 1/1000th of 35GB/s (gigaBYTES and not gigaBITS)
>effectively.


Err, MB/s.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

kony wrote:
>
> On Fri, 08 Oct 2004 04:26:54 +0100, Odie Ferrous
> <odie_ferrous@hotmail.com> wrote:
>
> >kony wrote:
> >>
> >>
> >> There may've been something wrong with the test then, I can
> >> exceed 35GB/s even coping over a LAN with single drive,
> >
> >I would be VERY interested to know what LAN you are on about here.
> >
> >I doubt you exceed 1/1000th of 35GB/s (gigaBYTES and not gigaBITS)
> >effectively.
>
> Err, MB/s.

Kony, even 35MB/s is super quick.

What LAN setup are you using?

I am inquisitive, because I need that sort of transfer speed but can
only get it by getting a Gigabit network installed.


Odie
--

RetroData
Data Recovery Experts
www.retrodata.co.uk

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt (More info?)

On Fri, 08 Oct 2004 14:46:51 +0100, Odie Ferrous
<odie_ferrous@hotmail.com> wrote:

>kony wrote:
>>
>> On Fri, 08 Oct 2004 04:26:54 +0100, Odie Ferrous
>> <odie_ferrous@hotmail.com> wrote:
>>
>> >kony wrote:
>> >>
>> >>
>> >> There may've been something wrong with the test then, I can
>> >> exceed 35GB/s even coping over a LAN with single drive,
>> >
>> >I would be VERY interested to know what LAN you are on about here.
>> >
>> >I doubt you exceed 1/1000th of 35GB/s (gigaBYTES and not gigaBITS)
>> >effectively.
>>
>> Err, MB/s.
>
>Kony, even 35MB/s is super quick.
>
>What LAN setup are you using?
>
>I am inquisitive, because I need that sort of transfer speed but can
>only get it by getting a Gigabit network installed.

Gb network with jumbo frames, nothing particularly special
otherwise, slightly higher TCP receive window and MTU
setting. Don't even use expensive NICs, just generic
realtek chip type that can be bought anywhere for $15 (check
pricewatch). Don't remember CPU utilization but 3Com or
Intel NICs would probably lower it.

 

[Guest]

Archived from groups: alt.comp.hardware.homebuilt,uk.comp.homebuilt (More info?)

"JonMaC" <john.mccarthy@services.fujitsu.com> wrote...
> Need a new Hard Drive, currently have Seagate Barracuda 7200.7 ATA100 80Gb -
>
> Q1 - Would i notice a speed increase if I went SATA 150 ?? are there any
> real world benefits to SATA other than thinner cables ??

Only benefit will be if you buy a physical HD that is built to take advantage of
the bandwidth. Otherwise, the physical limitations of the HD will be the
limiter on performance.


> Q2 - What brands are recommended ?? - have been pleased with Seagate so far
> so am leaning towards that rather tham cheaper Maxtor.

WD Raptor 74 is the only choice, IMO. It is the only 10K RPM SATA HD out there
(unless there are some recent additions), so it is the only one that will render
consistently better performance.