Forbes: What's Next?

[MacRumors]

Forbes posts a speculative piece about what is to come from Apple in the coming months.

Much of the information posted is simply a summary of what has been circulating around the rumor sites.

One analyst agrees with the predictions of speedbumped PowerMacs in Jan, and even speculates that a G5 PowerBook is possible.

 

[arn]

Forbes frequently summarizes stuff that's circulating around the rumor sites... so I wouldn't take this as confirmation of anything.

arn

 

[zellin]

One guy said that Apple could easily produce Powerbook G5s as early as January, saying that there were no technical boundaries. What is this guy smoking? As it is the PBG4s are as hot as blazes. Although this time of year on the east coast it comes in handy

 

[ITR 81]

A G5 PB could be done if they use IBM's liquid cooling sys. or the new 90nm processors.

 

[joelc]

Next summer??

It is mostly a summary of rumors, but the rumors about G5 powerbooks don't have much credibility, and no one really thinks that it will happen by next summer (no one says "good chance" or "no later than"). So either this analyst is totally clueless or knows something. I don't know what his firm does. I doubt he knows any actual specific info from Apple, but being in the business he might know people with knowledgable opinions.

 

[x86isslow]

ibm's new 970-based bladeServers are due to be released in spring 04

blades are small-form-factor, yes?

we can expect xServes, and g5 pb's in the spring then.
i doubt apple will wait until wwdc to update the laptops, considering the 1.33 is falling behind the centrinos.

 

[Rincewind42]

Originally posted by x86isslow
ibm's new 970-based bladeServers are due to be released in spring 04

blades are small-form-factor, yes?

we can expect xServes, and g5 pb's in the spring then.
i doubt apple will wait until wwdc to update the laptops, considering the 1.33 is falling behind the centrinos.

Blades are a small form factor, but the difference is that they can have large (or at least loud) cooling systems attached, which a PowerBook cannot. Although old rumors of power draw on 90 nm 970s would allow for a PB G5 in the 1.4-1.6 Ghz range.

Just for the fascinated their expert on 64-bit is wrong about the G5. Software compiled for the G5 can operate on 64-bit integers without breaking them up, what it cannot do is assign a single process all of the memory on the machine (but it can still access the theoretical 16 GB memory maximum).

And as for those complex calculations, they involve 64-bit integers and not too much else .

 

[mrsebastian]

Originally posted by Rincewind42
Just for the fascinated their expert on 64-bit is wrong about the G5. Software compiled for the G5 can operate on 64-bit integers without breaking them up, what it cannot do is assign a single process all of the memory on the machine (but it can still access the theoretical 16 GB memory maximum).

And as for those complex calculations, they involve 64-bit integers and not too much else .

that's what i thought as well about the 64-bit stuff.

sounds like this analyst is just trying to give apple a boost since the stock is slipping a little.

 

[iJed]

I think its possible that we will see a 64 bit version of Panther Server in January with the launch of the updated Xserve. There is more of a reason to have a 64 bit version of the server OS at this point in time than the client. This is especially true if Apple is really going to start targeting highend customers who have databases in excess of 2 gigs.

 

[crees!]

Originally posted by arn
Forbes frequently summarizes stuff that's circulating around the rumor sites... so I wouldn't take this as confirmation of anything.

arn

Well after reading the article I read something I haven't heard on any rumor sites and it sounded pretty grand.

Currently programmers use an open-source compiler called gcc. An IBM compiler tailored to the PowerPC chips is already in the beta-testing phase, Glaskowsky says. Once it's released in its final form, software written to the Mac running on PowerMac G5 machines could speed up as much as 50%, he says.

 

[arn]

Originally posted by crees!
Well after reading the article I read something I haven't heard on any rumor sites and it sounded pretty grand.

IBM's XL C and XL C++ compilers were released in August

https://www.macrumors.com/pages/2003/08/20030828190020.shtml

arn

 

[crees!]

Originally posted by arn
IBM's XL C and XL C++ compilers were released in August

https://www.macrumors.com/pages/2003/08/20030828190020.shtml

arn

Okay, now I remember seeing that being posted here.. bad memory.. but from what article says it is still in beta. Which is fine with me... as long as they're working on something that will boost performance significantly.. kudos to them.

 

[ffakr]

Originally posted by arn
IBM's XL C and XL C++ compilers were released in August

https://www.macrumors.com/pages/2003/08/20030828190020.shtml
arn

I don't think xlc was released in August. The beta was opened up in August (anyone who registers can download a copy)

Um..I don't see the link anymore though.. is it out of beta yet?

 

[e2chris]

Originally posted by arn
Forbes frequently summarizes stuff that's circulating around the rumor sites... so I wouldn't take this as confirmation of anything.

arn

lol. You mean what you do?

 

[crees!]

Originally posted by e2chris
lol. You mean what you do?

Ouch

 

[JFreak]

Originally posted by iJed
There is more of a reason to have a 64 bit version of the server OS at this point in time than the client.

there is MOST reason (for Apple) to have both server and client 64bit at the same time, because they both are taken from same code base. i very much doubt they choose to sell different configurations bitwise...

(this infobit is based on a practice called "software configuration management". have you ever wondered why server and client versions' build numbers never match? there may be server build 23 and client build 24, but never same build numbers for server and client. why? because they both use same code base, and are only packaged a little differently.)

 

[Bunzi2k4]

here's a thought...

i've noticed in games like return to castle wolfenstein i have pretty bad ping... like from 100 to 700, and i was thinking... wouldn't it be cooler to have a dual internet connection? like have an ethernet cable and airport going at the same time... kinda like a dual processor, but a dual internet connection... (i can see it now, everything is ran by gigabit ethernet cables, *imagines a laptop covered in ethernet sockets*...

 

[Sheebahawk]

forbes sucks

I meant to express my dislike for forbes magazine in the page one article, but I never even had time to read through the posts. Forbes was one of the forerunners in publishing Apple bashing opinion pieces during the Scary Times™, editorials which constantly proclaimed and predicted Apple's supposed bankrupcy and ultimate failure as a company. Unfortunately manypeople read and trust the opinions published in forbes magazine, which contributed to many jumping ship from the Apple platform. Why invest in technology that won't be around or supported next year? Its good that they've come around though, still its easier to do damage than repair it.

 

[ffakr]

Originally posted by Bunzi2k4
here's a thought...

i've noticed in games like return to castle wolfenstein i have pretty bad ping... like from 100 to 700, and i was thinking... wouldn't it be cooler to have a dual internet connection? like have an ethernet cable and airport going at the same time... kinda like a dual processor, but a dual internet connection... (i can see it now, everything is ran by gigabit ethernet cables, *imagines a laptop covered in ethernet sockets*...

This is called an aggregate connection. Bigger institutions and ISPs often use aggregate connections... joeblow ISP buys a T1 until it's capacity isn't enough any more. They then buy another T1 and they use both at the same time. Down the road, they may have a T1 (or two) and a T3...

Ping times are an indication of network latency, however. It's a measure of how long it takes to echo a packet off a remote host (generally in Miliseconds). Certain connections, like DSL, tend to have higher latency than others. Hard lines like T1s and T3s have lower latencies.. but they are $$$. I've got a friend who is drawing up a business plan based off one of my ideas and he priced T3 service at $9000 / month.