28-Apr-2024 17:57 GMT.		 UNDER CONSTRUCTION
[News] Kickflash OS4: Final product specificationsANN.lu
Posted on 18-Jun-2003 00:39 GMT by Jens Schönfeld25 comments
View flat
View list
Kickflash OS4 by individual Computers has been revised. It has 1MB flash now, and can be expanded to 1GB if necessary. Due to the high demand, Kickflash OS4, which has been announced on June 2nd, has been revised. The base version now has 1MB of flash memory, and can be expanded to 1GB on an expansion port. Transfer rates can be compared to normal harddrives, but the common access times of mechanical mass-storage media practically don't exist on flash memory: On random access, so-called solid-state memory is about tenthousand times faster than common harddrives.

Kickflash OS4 is a Zorro-card that's compatible with all Zorro-2 and Zorro-3 systems. The new base price for the board is 34,90 EUR. Prices for expansions are not yet available.
List of all comments to this article
Sorted by date, most recent at bottom
Comment 1reflect18-Jun-2003 00:05 GMT
Comment 2Atheist218-Jun-2003 00:09 GMT
Comment 3Atheist218-Jun-2003 00:15 GMT
Comment 4hooligan/dcsRegistered user18-Jun-2003 01:22 GMT
Comment 5CodeSmith18-Jun-2003 05:18 GMT
Comment 6greenboyRegistered user18-Jun-2003 05:24 GMT
Comment 7hooligan/dcsRegistered user18-Jun-2003 06:14 GMT
Comment 8Ben18-Jun-2003 07:36 GMT
Comment 9takemehomegrandmaRegistered user18-Jun-2003 08:22 GMT
Comment 10takemehomegrandmaRegistered user18-Jun-2003 08:24 GMT
Comment 11Anonymous18-Jun-2003 11:03 GMT
Comment 12MarkTime18-Jun-2003 11:18 GMT
Comment 13MarkTime18-Jun-2003 11:19 GMT
Comment 14Joe "Floid" Kanowitz18-Jun-2003 11:25 GMT
Comment 15Joe "Floid" Kanowitz18-Jun-2003 11:34 GMT
Comment 16MarkTime18-Jun-2003 12:30 GMT
Comment 17CodeSmith18-Jun-2003 16:21 GMT
Comment 18CodeSmith18-Jun-2003 16:22 GMT
Kickflash OS4: Final product specifications : Comment 19 of 25ANN.lu
Posted by Joe "Floid" Kanowitz on 18-Jun-2003 16:28 GMT
In reply to Comment 16 (MarkTime):
MarkTime said,

> oh yes, I read the anectdotal evidence. it should be laughed at.

I don't know. I think you're being somewhat reactionary.

> The person, well meaning yes...had a horrible slow old scsi configuration
> not even getting 1.0mb/sec

Yep. The question is, what *is* the peak throughput one can squeeze off the buses available, and what controllers can we hope to afford that can approach it?

If you have a PCI busboard or a PPC card with an onboard controller, it's no doubt better to use those.

> so his CF solution was zippier. Great. I doubt he is alone in having
> underperforming hdd access on an older Amiga...technology has moved ahead
> and a lot of people have slow setups.
> But the solution, is a faster modern IDE hdd, and if possible, a new
> controller (obviously not available to everyone).
> The CF solution was only better, in that anything would be better than
> what he had before.

Exactly. Why complain about a cheap stopgap?

> But a modern 7200rpm/2mb cache IDE hard drive (bought a 30Gig version for $30
> after rebates just last week)...would have blown away the performance of the
> CF card.

For sustained access, as noted.

> Only on seek times for very tiny files, would the CF be better.

Booting an OS, especially a "microkernel" that can presumably go concurrent (multitasking) long before all drivers are loaded requires:

A) Rapid sustained throughput on singlethreaded access to large files.
B) Low seek latency to avoid contention between concurrent processes fighting to load different files.

Massive buffers can help, seek-reordering (TCQ?) can help, even small buffers can help (Amiga FFS is a bit complicated to wrap one's head around; I'm not sure how many seeks are required, how much directory structure is being cached under current implementations, etc)... But silicon disks are certainly a 'brute-force' approach to point B.

> Hence a Kickflash product with 1mb.....in this limited application, they might
> have something.

No doubt. As noted above, though, I can't help but wonder if a 'raw' CF solution is actually equally flexible/potentially more supporting of "rapid-development." (Depending, of course, how Kickflash actually is designed.)

> using it to load an OS, though, is not faster, not applicable.

For an OS that consists of many small files, and basically fits on two 880k floppies...

There's a reason everyone talks about pitching it at the 'embedded' space. ;)

Really, for this *specific* application (booting AmigaOS as it stands today- or just enough of AmigaOS to cram a new Kickstart in and warm-boot- it's like arguing RDRAM versus DDR SDRAM.

> It's only a good idea if you need low power limited space...not if you have
> space for a hard drive, or the power to run a hard drive.
> in otherwords, people do the CF thing in certain types of server rooms.

Depends on the HD you have available. New drives are certainly fast, but some people (not me ;)) are afraid to ply the surplus/used market for drives, and good luck finding anything new-in-box for less than USD/Euro 100. Plus, combining machines with weak stock power supplies and modern 7200RPM drives can be fraught with.. fraughtness.

> but its not applicable here, and I can tell you, from just having bought
> a fast hdd and putting it in a towerized A1200...I don't have any noticeable
> wait for loading workbench..its mere seconds.

Good to hear; obviously, having 80+GB does have its advantages. On the other hand, we should remain glad our OS is small enough to be able to take advantage of the silent solid-state hardware... and apparently performs Well Enough with it.

Buy based on case-by-case merits, as always.
Jump...
#20 Joe "Floid" Kanowitz
TopPrevious commentNext commentbottom
List of all comments to this article (continued)
Comment 20Joe "Floid" Kanowitz18-Jun-2003 16:31 GMT
Comment 21smp26618-Jun-2003 21:47 GMT
Comment 22Atheist218-Jun-2003 22:37 GMT
Comment 23Doc Emmet Brown19-Jun-2003 04:49 GMT
Comment 24Olegil19-Jun-2003 11:09 GMT
Comment 25Don CoxRegistered user19-Jun-2003 12:08 GMT
Back to Top