Promise Ultra133 TX

[Gilbert]

I have a p2b-ds and installed 2 western digital eide hd's. After reading
the group, I decided to order the following from www.bestbargainpc.com (they
were the cheapest):

http://www.promise.com/marketing/datasheet/file/Ultra133tx2DS_v3.pdf


.. . . however they sent me the following:

http://www.promise.com/marketing/datasheet/file/1_U100TX2_en.pdf


There is only a $1 difference or so between the Ultra 133 TX and the Ultra
100 TX, so, I am not sure if it is worth getting an RMA #, etc. unless
anyone thinks there is a compelling reason to do so (other than just
princple only).

It looks like one difference is the 133 uses pci 2.2 and the other does not.

Thanks.

--g

 

[hyben]

I have a p2b-ds and installed 2 western digital eide hd's. After reading
the group, I decided to order the following from www.bestbargainpc.com (they
were the cheapest):

http://www.promise.com/marketing/datasheet/file/Ultra133tx2DS_v3.pdf


. . . however they sent me the following:

http://www.promise.com/marketing/datasheet/file/1_U100TX2_en.pdf


There is only a $1 difference or so between the Ultra 133 TX and the Ultra
100 TX, so, I am not sure if it is worth getting an RMA #, etc. unless
anyone thinks there is a compelling reason to do so (other than just
princple only).

It looks like one difference is the 133 uses pci 2.2 and the other does not.

Complain, change.

The practical difference is that one is running at ATA133 speed
(UDMA6) the other at ATA100 (UDMA5) when your HDs are ATA133.
I have now for several years been using U133 w 1 ATA133 and 1 ATA 100
HD. I can't say that I notice any difference (I never bothered to run
the benchmarks) but still the one is, at least, nominally 33% faster
than the other one (with the appropriate drives indeed)

If your HDs are ATA100 then it may not need to mean so much now, but
you may for one or another reason use the card even in your future
system and chances are you may run into fewer compatibility issues
with the newer one (U133)

After all, you ordered U133 and have got U100. Was it the other way
around then let's go, but why should you settle with an obsolete
product?

 

[Mercury]

If it works flawlessly on a P2B-DS then, well, I would say that somehow they
had magic for-thought and gave you the right product. For all the hassle, I
wouldn't worry too much - so long as it works 100%. To tell the truth I
never purchased a '133' device in my life & I buy a lot - thoroughly
disinterested in what is really a marketing ploy.

You would, from opinions expressed many times by others, not notice any
difference.

 

[John]

I have a p2b-ds and installed 2 western digital eide hd's. After reading
the group, I decided to order the following from www.bestbargainpc.com (they
were the cheapest):

http://www.promise.com/marketing/datasheet/file/Ultra133tx2DS_v3.pdf


. . . however they sent me the following:

http://www.promise.com/marketing/datasheet/file/1_U100TX2_en.pdf


There is only a $1 difference or so between the Ultra 133 TX and the Ultra
100 TX, so, I am not sure if it is worth getting an RMA #, etc. unless
anyone thinks there is a compelling reason to do so (other than just
princple only).

It looks like one difference is the 133 uses pci 2.2 and the other does not.

Thanks.

--g

The ATA/133 controller will be significantly faster when using some
utilities such as Norton Ghost.

 

[Paul]

I have a p2b-ds and installed 2 western digital eide hd's. After reading
the group, I decided to order the following from www.bestbargainpc.com (they
were the cheapest):

http://www.promise.com/marketing/datasheet/file/Ultra133tx2DS_v3.pdf


. . . however they sent me the following:

http://www.promise.com/marketing/datasheet/file/1_U100TX2_en.pdf


There is only a $1 difference or so between the Ultra 133 TX and the Ultra
100 TX, so, I am not sure if it is worth getting an RMA #, etc. unless
anyone thinks there is a compelling reason to do so (other than just
princple only).

It looks like one difference is the 133 uses pci 2.2 and the other does not.

Thanks.

--g

Stephan has mentioned on previous occasions, that you only
get >137GB support, if the BIOS on the U100TX2 is upgraded
to a recent enough version. AFAIK, all versions of the U133TX2
should be able to handle large disks, as large disk support
was added to the standards at the same time as U133.

Here is Stephan's advice:
http://groups.google.ca/group/alt.comp.periphs.mainboard.asus/msg/bf662e262a805a26?dmode=source

Paul

 

[Egil Solberg]

The ATA/133 controller will be significantly faster when using some
utilities such as Norton Ghost.

It would be very interesting if you could tell us why, because I can't see
what should make the big difference.

 

[Egil Solberg]

HD. I can't say that I notice any difference (I never bothered to run
the benchmarks) but still the one is, at least, nominally 33% faster
than the other one (with the appropriate drives indeed)

(...that are not yet invented, and would probably not have ATA133 interface
anyways when they do.)

 

[Stephan Grossklass]

I have a p2b-ds and installed 2 western digital eide hd's. After reading
the group, I decided to order the following from www.bestbargainpc.com (they
were the cheapest):

http://www.promise.com/marketing/datasheet/file/Ultra133tx2DS_v3.pdf

. . . however they sent me the following:

http://www.promise.com/marketing/datasheet/file/1_U100TX2_en.pdf

There is only a $1 difference or so between the Ultra 133 TX and the Ultra
100 TX, so, I am not sure if it is worth getting an RMA #, etc. unless
anyone thinks there is a compelling reason to do so (other than just
princple only).

The performance difference between the two is likely to be small even
with current drives, given STRs still are a good bit below the maximum
transfer rate for UDMA100. UDMA133 seems to be a good bit more critical
when it comes to cables, so the plain U100TX2 may in fact mean less
trouble. Apart from the maximum transfer mode, the chips on both
controllers are almost identical - actually they were for earlier
U100TX2s, while the current U100TX2s have a smaller BIOS area or
somesuch. PCI wise they'll behave the same, and halfway current
production U100TX2s should already ship with a 2.20 build 14 or 15 BIOS
capable of 48 bit LBA (I bought one used well over a year ago and it
apparently already had the .14 BIOS from the factory).

BTW, I'd at least try to get that $1 difference back.

Stephan

 

[John]

It would be very interesting if you could tell us why, because I can't see
what should make the big difference.

I wish I knew why.

When I bought my Promise Ultra 66 controller to go in my old BX board
(ATA/33) I never expected any performance increase, but Norton Ghost
absolutely flew on the Ultra 66 (compared to the ATA/33 controller).

Similarly, my current motherboard has an ATA/100 controller, but Ghost is
significantly faster on the ATA/133 controller card. I can't remember by
exactly what percentage, but it's huge.

I understand the theory about current drives not being able to fill even
ATA/66 bandwidth, but I can say for sure that Norton Ghost does not follow
that theory.

I don't know where we stand with SATA yet, because all my drives are IDE.

 

[edavid3001]

I agree..

My motherboard has SATA II and Ultra 133. I have a SATA II drive and a
ULTRA133. These both benchmark out at 59MB/s.

If I connected a second drive running at 59MB/s to the same ULTRA 133
controller, I could see up to 118MB/s which would be more than the
UDMA100 could handle, but less than the 133.

However, the odds of this ever happing are very low. You'll never
notice it - unless drives start getting much quicker than they are now
and you purchase two of them and hook them to the same IDE cable.

 

[John]

I wish I knew why.

When I bought my Promise Ultra 66 controller to go in my old BX board
(ATA/33) I never expected any performance increase, but Norton Ghost
absolutely flew on the Ultra 66 (compared to the ATA/33 controller).

Similarly, my current motherboard has an ATA/100 controller, but Ghost is
significantly faster on the ATA/133 controller card. I can't remember by
exactly what percentage, but it's huge.

I understand the theory about current drives not being able to fill even
ATA/66 bandwidth, but I can say for sure that Norton Ghost does not follow
that theory.

I don't know where we stand with SATA yet, because all my drives are IDE.

A bit more info.

My P4 2.7Ghz system ghosts at 1.5GB/min to 2.2GB/min (depending on NTFS or
FAT partitions and size of files etc) from motherboard's built-in ATA/100
controller to PCI ATA/133 controller.

 

[Robert Hancock]

I wish I knew why.

When I bought my Promise Ultra 66 controller to go in my old BX board
(ATA/33) I never expected any performance increase, but Norton Ghost
absolutely flew on the Ultra 66 (compared to the ATA/33 controller).

Similarly, my current motherboard has an ATA/100 controller, but Ghost is
significantly faster on the ATA/133 controller card. I can't remember by
exactly what percentage, but it's huge.

I understand the theory about current drives not being able to fill even
ATA/66 bandwidth, but I can say for sure that Norton Ghost does not follow
that theory.

I don't know where we stand with SATA yet, because all my drives are IDE.

Ghost uses the BIOS to access the drives. Likely the Promise adapter
BIOS is doing things more efficiently than the system BIOS and the
onboard controller (using DMA instead of PIO, or something).

 

[Stephan Grossklass]

Ghost uses the BIOS to access the drives. Likely the Promise adapter
BIOS is doing things more efficiently than the system BIOS and the
onboard controller (using DMA instead of PIO, or something).

The Promise BIOS does indeed turn on the fastest DMA mode possible, and
it seems to be programmed fairly efficiently, performing well even when
not being cached. It's a shame that this is not even offered as an
option for the integrated IDE controllers in most main BIOSes (only the
Phoenix BIOS already enabled DMA years ago).

Stephan

 

[Daniel Mandic]

I understand the theory about current drives not being able to fill
even ATA/66 bandwidth, but I can say for sure that Norton Ghost does
not follow that theory.

Hi John!



One more recognition to the capability and processing speed of the
mighty i440BX Chipset. The only, fully working chipset , incl. AGP
(PCI seems also to work, as you said with Norton Ghost). For every
write-through, write-back and, of course Write behind ;-) write.




Best Regards,

Daniel Mandic

P.S.: There could be more quenched out of this chipset with an even
faster CPU then available :-(

 

[John]

IDE.

Ghost uses the BIOS to access the drives. Likely the Promise adapter
BIOS is doing things more efficiently than the system BIOS and the
onboard controller (using DMA instead of PIO, or something).

Thanks for the info Robert.

 

[John]

Hi John!



One more recognition to the capability and processing speed of the
mighty i440BX Chipset. The only, fully working chipset , incl. AGP
(PCI seems also to work, as you said with Norton Ghost). For every
write-through, write-back and, of course Write behind ;-) write.




Best Regards,

Daniel Mandic

P.S.: There could be more quenched out of this chipset with an even
faster CPU then available :-(

Yes. I still have 2 PC's with BX chipsets still going strong.

I also liked the 845E chipset.

 

[Daniel Mandic]

Yes. I still have 2 PC's with BX chipsets still going strong.

I also liked the 845E chipset.

Is this the P4B?


You can use a very wide Range of CPU from 233-1400MHz with the BX. Even
with a P2/233 there is a good feeling, knowing, that at least the MB is
not braking into to the bargain. Not to mention the stability,
expandability (USB, PS2, PCI, AGP, ISA 8/16 etc....), compatibility
(DOS5-WinXP !!SP2!!), upgradability (Tualatin) etc. etc. etc... and
fast (QpA - out of my imagination, splendid), of course. Possibly very
well for Unix and Derivates, too.

I have my P3B-F with a P2 and the P2B-F with a P3



Best Regards,

Daniel Mandic

 

[edavid3001]

My P4 2.7Ghz system ghosts at 1.5GB/min to 2.2GB/min (depending on NTFS or
1.5GB/min is 1536MB/min and 25.6MB/s. The bus is 100/133/150/300 MB/s
depending on DMA mode.
2.2GB/min is 2253MB/min and 57.5MB/s. ditto

Odds are the 25.6MB/s is throttled due to writes on the destination,
not reads from the source. 100Mb/s ethernet is ~12.5MB/s. Ghost uses
compression so it can write more than 12.5MB/s by compressing the data
before writing. It's very good at compression. I see well over
12.5MB/s on 100Mb ethernet on my LAN.

But you are going 100 to 133. So you're using 57.5MB read on one bus,
and probably less than this on the write bus due to compression. My
drives read benchmark at 59MB/s. Write is lower.

If you were to turn compression off, I bet you'd see GHOST drop from
57.5MB/s to around 25-30MB/s which is probably close to where your
drive would benchmark on writes.

 

[Daniel Mandic]

1.5GB/min is 1536MB/min and 25.6MB/s. The bus is 100/133/150/300 MB/s
depending on DMA mode.
2.2GB/min is 2253MB/min and 57.5MB/s. ditto

Odds are the 25.6MB/s is throttled due to writes on the destination,
not reads from the source. 100Mb/s ethernet is ~12.5MB/s. Ghost uses
compression so it can write more than 12.5MB/s by compressing the data
before writing. It's very good at compression. I see well over
12.5MB/s on 100Mb ethernet on my LAN.

But you are going 100 to 133. So you're using 57.5MB read on one bus,
and probably less than this on the write bus due to compression. My
drives read benchmark at 59MB/s. Write is lower.

If you were to turn compression off, I bet you'd see GHOST drop from
57.5MB/s to around 25-30MB/s which is probably close to where your
drive would benchmark on writes.

Heck the what are doing you?

From where did you reply.....



Best Regards,

Daniel Mandic

 

[John]

Is this the P4B?


You can use a very wide Range of CPU from 233-1400MHz with the BX. Even
with a P2/233 there is a good feeling, knowing, that at least the MB is
not braking into to the bargain. Not to mention the stability,
expandability (USB, PS2, PCI, AGP, ISA 8/16 etc....), compatibility
(DOS5-WinXP !!SP2!!), upgradability (Tualatin) etc. etc. etc... and
fast (QpA - out of my imagination, splendid), of course. Possibly very
well for Unix and Derivates, too.

I have my P3B-F with a P2 and the P2B-F with a P3



Best Regards,

Daniel Mandic

Asus P4B533-E