jwhitman@STL-06SIMA.ARMY.MIL (Jerry Whitman) (09/20/90)
Hi all;
I saw the beginning of this thread a couple weeks ago, but it seemed to die
out before I saw answers to areas I am concerned with, so I'll try to start
it up again.
I have read with great interest the GVP ads for the new HD controller and RAM
board. The trade-in feature particularly interested me. I currently have a
stock 2000HD with a 2090A controller and the 40MB Rodime drive.
Here are the issues that concern me:
1. How much real advantage do I gain by booting up under FFS as opposed to
my current process using the Old/Slow File System? Is it significant
enough to warrent making the change?
2. Do I have to reformat only the Boot partition to switch to the FFS since
the rest of the HD is currently FFS? What is involved here?
3. I do not currently do or anticipate doing any serious graphics intensive
processing. Will the purported 'new and improved' data transfer rate
give me any noticable improvement in the more mundane programs? Am I
missing something key here?
4. This is sort of part 2 of question 2 above. If I install the GVP
controller will it be able to make sense out of my HD or will this be a
re-installation? (Re-Format and Re-Load the whole ball of wax)
5. Is there any inherant advantage/disadvantage to having the memory
expansion on the controller board, and do I lose any 2090A functionality
that I might need later? Given GVPs reputation for quality
design I would suspect not, but I would rather ask than be sorry later.
6. Well, here it is, the inevitable catch-all question!! What else do I
need to know to make an intellegent decision concerning swapping out
my 2090A for the GVP controller?
Thanks for your kind attention to a long list of question. I suspect I am
not the only one out there who would like answers to some, if not all, of
these issues.
Regards, Jerry Whitman - Keeper of the FishXref and FishConjmeissen@ogicse.ogi.edu (John Meissen) (09/21/90)
In article <9009201538.AA28646@ucbvax.Berkeley.EDU> jwhitman@STL-06SIMA.ARMY.MIL (Jerry Whitman) writes: >Hi all; > >I saw the beginning of this thread a couple weeks ago, but it seemed to die >out before I saw answers to areas I am concerned with, so I'll try to start >it up again. > [....] I decided to take advantage of the offer, so I will try to address your questions. >Here are the issues that concern me: > >1. How much real advantage do I gain by booting up under FFS as opposed to > my current process using the Old/Slow File System? Is it significant > enough to warrent making the change? My solution under the A2090(a) was to have a (very) small partition that was autoconfigured, and got used just long enough to mount and switch over to a FFS partition as the main system partition. If you are running your main partition under Old File System, switching to FFS is definitely worth it. There are other methods, though, so that in itself isn't reason enough to spend ~$100. >2. Do I have to reformat only the Boot partition to switch to the FFS since > the rest of the HD is currently FFS? What is involved here? You will have to reformat the drive anyway. See below. >3. I do not currently do or anticipate doing any serious graphics intensive > processing. Will the purported 'new and improved' data transfer rate > give me any noticable improvement in the more mundane programs? Am I > missing something key here? Under "normal" usage (non-interlaced) any improvement probably won't be noticable if you are already using C='s A2090a, since it is already DMA. If you do a lot of interlaced screen stuff then it may be noticable. Users of non-DMA controllers will probably see the 'hit' their systems take when doing disk transfers will go away. >4. This is sort of part 2 of question 2 above. If I install the GVP > controller will it be able to make sense out of my HD or will this be a > re-installation? (Re-Format and Re-Load the whole ball of wax) I have been told that you will have to reformat and reload the whole ball of wax. I, too, had an A2090, but I haven't received my GVP yet so I can't comment from experience. Naturally, I will try without reformatting first :-) >5. Is there any inherant advantage/disadvantage to having the memory > expansion on the controller board, and do I lose any 2090A functionality > that I might need later? Given GVPs reputation for quality > design I would suspect not, but I would rather ask than be sorry later. Yes, the controller can DMA into the on-board memory without competing for the Zorro BUS (so I have been told). This makes transfers faster than if all the expansion memory was on another card. >6. Well, here it is, the inevitable catch-all question!! What else do I > need to know to make an intellegent decision concerning swapping out > my 2090A for the GVP controller? The REAL reason for upgrading is that (as far as I can determine) the A2090(a) cannot be made to work with SCSI tape units, CD-ROM drives, etc. If you anticipate wanting any of these later, you will be upgrading anyway. Considering this, the upgrade makes a lot of sense on a cost basis. As for negatives................ Expect to be without a working system for quite a while. They currently have about a two week backlog, which means if you take your controller out and ship it normal UPS and they ship theirs back the same way you will be without a working system for 3 - 4 weeks. I tried working with them to get around this (I am pinched badly just being down a couple of days, and unfortunately it's stretched into over a week now), by placing the order by phone and holding my controller until they said a replacement was ready. Even still, they didn't have one when they said they did, they didn't ship when they said they would, and they almost didn't ship next day air, like I paid them to. Luckily I called they day they were shipping and caught that. With any luck it will arrive this morning. Expect to have a difficult time talking to anybody on the phone, and don't expect a return call, even if they promise they will. I never got them to return a single call in three weeks. (well, one, yesterday, finally, hours before it was shipped) These are frustrating problems, but also understandable. They are selling literally thousands of these, and it is no doubt taxing their support structure to the limit. It would have been MUCH nicer to be able to order the controller at regular price, and get a return authorization for the trade-in so you could keep your system up until the new controller arrived, then make the switch and return the old one for credit. As it is, there is no backing out, and if there are ANY problems getting the new controller to work you will be SOL. >Regards, Jerry Whitman - Keeper of the FishXref and FishCon -- John Meissen .............................. Oregon Advanced Computing Institute jmeissen@oacis.org (Internet) | "That's the remarkable thing about life; ..!sequent!oacis!jmeissen (UUCP) | things are never so bad that they can't jmeissen (BIX) | get worse." - Calvin & Hobbes
h112706@assari.tut.fi (Herranen Henrik) (09/21/90)
In article <12291@ogicse.ogi.edu> jmeissen@ogicse.ogi.edu (John Meissen) writes: >In article <9009201538.AA28646@ucbvax.Berkeley.EDU> jwhitman@STL-06SIMA.ARMY.MIL (Jerry Whitman) writes: >>3. I do not currently do or anticipate doing any serious graphics intensive >> processing. Will the purported 'new and improved' data transfer rate >> give me any noticable improvement in the more mundane programs? Am I >> missing something key here? > >Under "normal" usage (non-interlaced) any improvement probably won't be >noticable if you are already using C='s A2090a, since it is already DMA. >If you do a lot of interlaced screen stuff then it may be noticable. Users >of non-DMA controllers will probably see the 'hit' their systems take when >doing disk transfers will go away. A minor correction: It really isn't interlace that makes the A2090a slow, but hi-res with much colours. Interlace won't affect the read/write speed at all. >>Regards, Jerry Whitman - Keeper of the FishXref and FishCon > >-- >John Meissen .............................. Oregon Advanced Computing Institute >jmeissen@oacis.org (Internet) | "That's the remarkable thing about life; >..!sequent!oacis!jmeissen (UUCP) | things are never so bad that they can't >jmeissen (BIX) | get worse." - Calvin & Hobbes -- Name: Henrik 'Leopold' Herranen h112706@lehtori.tut.fi Address: TTKK/P{{rakennuksen neuvonta/PL527/33101 Tampere/Suomi Finlandia "On d{htinen daevas ja kuutamoy|, on morsiamelta katkaistu p{{" E.L.1989
leblanc@eecg.toronto.edu (Marcel LeBlanc) (09/22/90)
jmeissen@ogicse.ogi.edu (John Meissen) writes: >>5. Is there any inherant advantage/disadvantage to having the memory >> expansion on the controller board, and do I lose any 2090A functionality >> that I might need later? Given GVPs reputation for quality >> design I would suspect not, but I would rather ask than be sorry later. >Yes, the controller can DMA into the on-board memory without competing >for the Zorro BUS (so I have been told). This makes transfers faster than >if all the expansion memory was on another card. I think the person you spoke to may not have been very clear. Unless GVP has completely redesigned their SCSI controller, it DOES NOT DMA to the Amiga's memory (the 2090A does). The GVP controller that I have writes to a cache (hence their misleading 'DMA to on-board memory' claim), which must then be copied to it's final destination by the CPU. The GVP controller is not in the same speed league as the 2090A controller. Dave Haynie recently posted an excellent explanation of the DMA and non-DMA techniques used by popular HD controllers for the Amiga. I can mail you a copy if you would like. Marcel A. LeBlanc -- Electrical Eng. Computer Group, Univ. of Toronto ----------------------------------------------------------------------- leblanc@eecg.toronto.edu else: uunet!utcsri!eecg!leblanc
jmeissen@ogicse.ogi.edu (John Meissen) (09/22/90)
In article <1990Sep21.163715.3423@jarvis.csri.toronto.edu> leblanc@eecg.toronto.edu (Marcel LeBlanc) writes: >I think the person you spoke to may not have been very clear. Unless GVP >has completely redesigned their SCSI controller, it DOES NOT DMA to the >Amiga's memory (the 2090A does). The GVP controller that I have writes to a >cache (hence their misleading 'DMA to on-board memory' claim), which must >then be copied to it's final destination by the CPU. The GVP controller is Yes, GVP "has completely redesigned their SCSI controller." That is one of the important features of the Series-II controller. This version, UNLIKE the earlier version, does do true DMA anywhere in the Zorro address space. The memory on the board is autoconfigured into the Zorro address space, and is part of the Amiga Fast Ram. In addition, it is supposed to be set up in a dual-port configuration (I don't know the details) so that if your Fast Ram is on the board it can transfer directly into that memory without having to deal with bus arbitration on the Zorro bus (assuming that's where the drive's buffers are allocated...I assume they handle that somehow; the documentation is quite sparse). > >Marcel A. LeBlanc -- Electrical Eng. Computer Group, Univ. of Toronto >----------------------------------------------------------------------- >leblanc@eecg.toronto.edu else: uunet!utcsri!eecg!leblanc -- John Meissen .............................. Oregon Advanced Computing Institute jmeissen@oacis.org (Internet) | "That's the remarkable thing about life; ..!sequent!oacis!jmeissen (UUCP) | things are never so bad that they can't jmeissen (BIX) | get worse." - Calvin & Hobbes