Prototype Quicksilver Server Summary

[arn]

Re: the evidence it isn't 1394b

Originally posted by gopher
I'd certainly like to be proven wrong, but remember until we actually see the thing, or the specs for 1394b are out, my guess is the 56k modem has been put on the motherboard next to the other ports.

The specs are out, and this is exactly what it looks like

reference: http://www.1394ta.org/Technology/About/ppt1.PDF

arn

 

[Rower_CPU]

Re: Re: the evidence it isn't 1394b

Originally posted by arn
The specs are out, and this is exactly what it looks like

reference: http://www.1394ta.org/Technology/About/ppt1.PDF

arn

Thanks arn. I was trying to see if the guy had bothered to check that link you posted. Obviously, he hadn't.

 

[gopher]

thank you.

I e-mailed Arn a thank you, I hadn't seen that link before. And normally I wouldn't check PDF documents. I'm still curious how will they mount the modem if it is off board. Very odd. Unless like the baseline eMac it won't have a modem. Wouldn't that be a curious twist of fate. The death of 56k modems? I sure hope not. Unless Apple is about to announce nationwide wireless broadband!

 

[Catfish_Man]

32 bit chips...

Originally posted by jelloshotsrule
just curious about something in general...

why is there such a "small" limit on RAM in the current macs (ie towers)? i remember hearing a guy at the maya demo at mwny last year saying the computer was a dual 800 with 1.5 gigs, but ideally the computer would have 5 gigs of ram with terabytes of hard drive space...

so is the issue of limiting it to 1.5 (as of now) a cost, space, practical (ie, not enough people would want more than 3 slots to make it worthwhile) issue?

...will support up to 4GBs of RAM. This prototype motherboard has 4 DDR slots (which is a bit odd. unless they have a *nice* memory controller, they'll have to slow the ram down and use registered ram to have all 4 slots filled), so it can have the full 4 GBs. 64 bit chips will be able to support *much* more ram.

 

[Scottgfx]

IDE Raid on MB

Others have noted the fact that there is three IDE connectors. Did anyone notice that there is a set of dip switches next to the 2 connectors near the back of the board? I'm going to bet that the chip to the immediate left of those connectors is a Highpoint IDE RAID chip. Also, to the right of the ethernet connector is a space, soder pads and silkscreen for another connection. I'm guessing that it's missing the digital audio.

If I am right that the big chip near the back is the IDE RAID, then where are the Northbridge and Southbridge chips? I guess that they are on the other side of the MB. Actually, now that I look at it, Just behind the processor socket, there looks like some passthroughs for a chip on the other side. Interesting!

What's the white connector to the left of the Airport?

 

[johnpaul191]

Re: IDE Raid on MB

Originally posted by Scottgfx
Others have noted the fact that there is three IDE connectors. Did anyone notice that there is a set of dip switches next to the 2 connectors near the back of the board? I'm going to bet that the chip to the immediate left of those connectors is a Highpoint IDE RAID chip.

i think it's safe to assume that anything with dip switches on a test motherboard will be replaced with resistors or jumpers in the final product.... as for the 2 questionable ports.... doesnt Apple always put in some sort of oddball "testing" port on it's products when they are still hammering everything out? i hear that they *usually* are gone by the final product, but sometimes stay (like the slot on the bottom of the origional Rev A iMacs).

 

[Scottgfx]

Re: Re: IDE Raid on MB

Originally posted by johnpaul191


i think it's safe to assume that anything with dip switches on a test motherboard will be replaced with resistors or jumpers in the final product.... as for the 2 questionable ports.... doesnt Apple always put in some sort of oddball "testing" port on it's products when they are still hammering everything out? i hear that they *usually* are gone by the final product, but sometimes stay (like the slot on the bottom of the origional Rev A iMacs).

I was thinking along the lines of the fact that almost all of the IDE RAID PCI cards do have a block of dip switches on them.

What happened to all of the talk about the special clustering hardware that Apple was supposed to be developing? Does anyone see it in this MB?

 

[mcrain]

Re: thank you.

Originally posted by gopher
The death of 56k modems? I sure hope not. Unless Apple is about to announce nationwide wireless broadband!

Funny you should say that...

Cingular, Verizon, Sprint and a few others are getting ready to implement the 3g wireless system. I heard on the radio that it is capable of 128k internet speeds. That's more than twice what I get with my stupid dialup modem, but not nearly as fast as dsl, cable or the other broadband connections. That being said, however, if nationwide wireless internet is available at speeds twice that of a dialup modem, it would be in Apple's way of doing things to be the first to dump the dial up.

Then again, there are a lot of people who are stuck with 1-2 year dialup contracts from those promotions they used to run at best buy and circuit city.

 

[keithcobbett]

no way....

This I don't buy for much more than .03 seconds. Why would you stay with a 133MHz bus???

 

[mischief]

you wouln't be staying with a 133 bus.

There has been a lot of speculation around Apple introducing RapidI/O at it's base speed of 250 Mhz. 133DDR has an effective datarate of 266 so it is theoretically possible to drop a few cycles and sync the two. Like tuning a Harley to "trot" like a horse. About a year ago I saw a prototype that had the instabilities I would expect from an early stage test of such a system.

This means Apple could move to DDR which has better value for money and Rapidio which goes as high as 500 Mhz. A 500 Mhz bus would allow MUCH higher processor speeds. I think a lot of this hinges on Apple suprising with a 32/64 capable chipset and CPU.