3 more products added Catweasel etc

[Acill]

tjaoz wrote:
@redrumloa

No way to UDMA, the Buddha does not DMA. I'd guess it probably is comparable to PIO-3 or PIO-4.

PIO-3 or PIO-4? :-o

As a distributor of the Buddha card you should rather know that it works in PIO-0 only.

Due to the limitation of the Zorro II interface this card  cannot even reach in A2/3/4000 computers the full PIO-0 speed.  

Thats a load of crap! I used one in my A3000 for a few years in the form of a catweasel s class card. They are built together on that one. It got software that lets you select your PIO mode. I had it at 4 with no problems. It was MUCH faster then my internal SCSI but not near the speeds of the SCSI on my CSPPC card.

[redrumloa]

JJ wrote:
ive got a powerflyer gold in my miggy A1200, is that not DMA, and what sort of load does it put on my cpu, just curious, never looked into it

100% CPU load, get a cpu monitoring utility and be amazed:-)

[jj]

serious, damm, i wish i never sold my blizzard IV 30/50 with scsi now, that was dma, never got to use it though, never had any scsi drives, bought the scsi bit, so i could add extra ram lol

[adolescent]

redrumloa wrote:

100% CPU load, get a cpu monitoring utility and be amazed:-)

The Buddha will take a huge CPU hit in faster modes just like the FastATA.  This is due to the lack of DMA.  The Buddha doesn' t have any magic fix for this problem.  It might be less noticable because the Buddha is limited to ~2.8M/sec being a Zorro 2 only.  The FastATA can go much faster, and thus uses more CPU.  

I would recommend just getting a SCSIDE adapter and using the internal SCSI.  Jeff- I have one that you can borrow to test if needed.  That way you maintain just about the same speed, but you save a lot of CPU.

[adolescent]

Acill wrote:
Thats a load of crap! I used one in my A3000 for a few years in the form of a catweasel s class card. They are built together on that one. It got software that lets you select your PIO mode. I had it at 4 with no problems. It was MUCH faster then my internal SCSI but not near the speeds of the SCSI on my CSPPC card.

I'm going to have to agree with tjoaz here.  The theoretical transfer rate of PIO-0 is 3.2MB/s.  I don't think I've seen anything over 3MB/s on the Z2 bus (Buhdda bencharks show a max of around 2.8MB/s on Z3).

Although the Buddha might support switching modes, it can not out run them.  

[mindprober]

adolescent wrote:

I'm going to have to agree with tjoaz here.  The theoretical transfer rate of PIO-0 is 3.2MB/s.  I don't think I've seen anything over 3MB/s on the Z2 bus (Buhdda bencharks show a max of around 2.8MB/s on Z3).

I take that to mean that the card supports Zorro 3 mode. Nice. Strange that this fact was omitted from the product web page. You would think this would be an important sales point. Although, it wouldn't appear to make much of a difference in real world performance.  

[adolescent]

I don't think it supports Zorro 3 mode.  The benchmark I saw was a Buddha on an A4000/040.  I'd expect an A3000/060 to move more data than that.  But, at 2.8MB/s it's not much faster than the internal SCSI and has much more CPU load.

[tjaoz]

@x56h34

I suppose you wanna recommend FastATA4000 at this point?

Read my post once again and you will see what I really wrote.

You make me sick, you know that?

What is wrong with my post? Don't you like hard facts?

as the speed differece was barely noticeable between IDE/ATA operation between the Buddha and FastATA4000. Sure FastATA4000 was slightly faster, but it was honestly barely noticeable.

My FastATA gives me about 9.6MB/s. Sure, if I set the drive to PIO-0 it does not exceed 2.9MB/s.
FastATA working in PIO-0 is comparable to the Buddha in performance. :-)

If you bothered to do a little research, you would have known that Buddha comes with a little speed setting utility which enables higher PIO modes than mode 0

You are wrong. All Buddha modes have NOTHING to do with PIO modes. None of them is faster than the PIO-0 mode.

It's just that you work for Elbox and you are required to slander all the other products out there.

I do not work for Elbox and have never worked for them.

I simply do not like the way of making business by Redrumloa. IMHO it is not fair if someone who distributes 3rd party products writes untrue information about these products. Even if he adds the word "probably" before the information.

[tjaoz]

@Acill

Thats a load of crap! I used one in my A3000 for a few years in the form of a catweasel s class card. They are
built together on that one. It got software that lets you select your PIO mode. I had it at 4 with no problems.

You missed up the PIO modes with the Individuals' custom modes! These custom modes have nothing to do with PIO modes. All Individuals custom modes are NOT faster than PIO-0 mode.

The cycle times for the PIO modes are as follows: PIO 0 - 600 ns, PIO 1 - 383 ns, PIO 2 - 240 ns, PIO 3 - 180 ns, PIO 4 - 120 ns.
The cycle time is the distance between the start of one read (or write) cycle and the start of next read (or write) cycle. During each read (or write) cycle, the IDE controller reads (or writes) two bytes from (or to) the hard drive. It means that in the PIO-0 mode the controller can transfer 2 bytes per 600 ns. (More about it you can find in the ATA spec http://www.t13.org/project/d1321r3.pdf)

The read (and write) cycle of the Zorro II card fitted in the Zorro slot of A2/3/4000 is 5 x 140 ns = 700 ns. It means that the Zorro II card (Buddha in this case) cannot transfer data from/to hard drive faster than 2 bytes per 700 ns, which is slower than in PIO-0 mode!!!

In the Buddha card, Individuals defined their custom timing for some hard drive signals. These modes are adjusted in the so-called "Buddha speed-register" ($7fe). The numbers you can find at  http://www.schoenfeld.de/inside/Inside_BuddhaE.txt

Looking at these timings you can see that some of these modes slow down the hard drive much UNDER the PIO-0 mode ("Buddha speed-register" values: 1, 2, 6). If you set this register to the value 0, 3, 4, 5 or 7 (and the timing does not break your hard drive requirements) than you hard drive will work faster, but still SLOWER than the full PIO-0 speed.

You can check it taking your hard drive from Amiga to PC, setting PIO-0 and checking its speed. It will work not slower than when connected to the Buddha card, with Buddha  fastest mode selected.

[Lemmink]

What is all this babble about PIO and stuff.
Just come around with some realworld data everyone easily understands. As the ZII bus can not exeed 3 MB/s so the Buddha as a plain ZII card could not too.
My experiance is about 2.3-2.5MB/s in reallive in an 060 exuipted A2000.
CPU-Load is modearate but you sure feel the difference to SCSI.
All in all you can compare the buddha to the onboard IDE of an A1200/ A600 /A4000 only slightly faster with a little less CPU-load due to way better design and more compatible to modern drives.
Overall the Buddha is the best Zorro IDE controller you can get ( I had an FastATA4000 once, struggeld with it a few weeks and sold it off for half the price - and still felt sorry for the guy that bought it -  as I came to the conclusion that I`m way better off with an IDE-SCSI Adapter if I want speed out of an IDE drive on an Amiga)

Haveing said that I won`t use IDE for harddrives at all on an Amiga anymore, as you miss every CPU cycle you lose.
Again, if you can`t affort decent SCSI equipment go for the Buddha.

[tjaoz]

@Lemmink
 

What is all this babble about PIO and stuff. Just come around with some realworld data everyone easily
understands. As the ZII bus can not exeed 3 MB/s so the Buddha as a plain ZII card could not too. My experiance is
about 2.3-2.5MB/s in reallive in an 060 exuipted A2000.

What is so obvious, that is that Buddha supports only PIO-0, was not so clear to all here. No wonder if even this card's distributor -Redrumloa- stated that this card "probably" supports PIO3 and PIO4. :-D

Back to PIO modes, it is obvious that all PIO modes consume CPU time. Programmed I/O is performed by the system CPU. The system processor is responsible for executing instructions that transfer data to and from the drive.

And this is the reason why it is always better to have a controller supporting faster PIOs than PIO-0 alone. Remember that in PIO-4 every program to be read from the disk is loaded several times faster than in PIO-0. Of course, during its loading, the processor is heavily loaded with execution of the read operation, but this load lasts several times shorter than in PIO-0.  

If for you the most important issue is that processor load is minimum during data being read from the disk, if you have a controller supporting fast PIOs (PIO-3 or PIO-4), you should reduce the priority of this controller driver. Then reading from the disk would slow down when your CPU is busy with a more important tasks, and will run at full speed when the CPU is not loaded. Period.

[Orjan]

@tjaoz

Fine, the Elbox solution may be faster than the Buddha solution, but it costs nearly ( not quite, but nearly ) twice as much as the Buddha which you, atleast in my head, also have to take into account...

[Paul_Gadd]

@tjaoz

Redrumloa is trustworthy and his customers harddrives are safe.
Elbox is untrustworthy, no matter how good their products are.

If you like them then fine but to the real people, they are a bunch of trojan writing rip off {censorship}.

[Cyberus]

Paul_Gadd wrote:
@tjaoz

Elbox is untrustworthy.

If you like them then fine but to the real people, they are a bunch of trojan writing rip off {censorship}.

Could you please clarify/elaborate?
I just want to know what makes you say this.
Thanks

[whabang]

I'mm sure he's reffering to some certain RDB-trashing code. :\

[Piru]

@Cyberus

http://users.kymp.net/p305543a/elboxdecrypt/