Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
new chip with reduced performance

Really? :p , I would have thought the title of the thread was going to be 'Intel's latest chips degrade performance' , lol. That's a no brainer.

I'm guessing you weren't in the PC world when the Pentium 4 was released. Even a slower-clocked Pentium III outperformed it in 5 out of 8 tests run by Tom's Hardware Guide. :) I've never been an Intel fan, I like competition, but I am glad to see they turned things around so completely with the Core and Core 2 chips. Very impressive.
 
.

Apple needs to do a little PR work. Marketing 101: What's in it for 'us?" How about giving us a free upgrade to Leopard if we buy a new computer now... or half off of an iphone with purchase... or add an Apple Care Buyers Remorse plan for $150, that if a new hardware platform comes out within 60 days of purchase, you can upgrade to the newer computer....??? Something, anything...

Do we even get Santa Rosa???:eek:

Or will they skip that... and just go with Penrym in Q4?

why should they? you'll still buy it. you'll spend as much as you can afford. :apple:
 
You are getting into the Power User Trap. Power Users want the latest and greatest. But when it comes out they see the next latest and greatest is coming. About 5 months ago I Upgraded to a C2D MBP And it is is still faster then all the other systems I work with. When you realize that your old systems is too slow for your needs and you want to be a power user. Just say to yourself Ill get the next new version when it comes out.
Nah, he's just gone full circle. When you get hooked, you've got two choices-- lament about each new revision being better than what you have and spend a fortune trying to keep up, or get yourself so hyped up on each new rumor that you save a fortune always waiting for the next big thing.
 
why should they? you'll still buy it. you'll spend as much as you can afford. :apple:

Apple should decide what they want. I assume that they want every home and business and pro out there owning a Mac computer. Look how much money they are spending on advertising for the new ads (Get a Mac). Look at how much floor space is dedicated to computers in the Apple stores (I know, ipods are probably floating the rent). But they could have stores half their size if they were just selling ipods, etc.... which they could use to hire more people/programers :D

Yet, Apple has nothing exciting for the average Joe right now. Even with the release of the 8 core option, most people on this site (pros) will wait for Penrym, Leopard, and/or possibly new designs (2008?).

Since Apple dropped computer from thier name, and only has 5% of the computer market, something has got to get thier core-base/fans excited, which will translate into word-of-mouth promotion. Case and point: I am ready to buy a Mac ($3,300 budget) and now my little cousin, collegebound, is sold on them, and she's ready to buy a 15" MBP ($2,500 budget). We don't want to buy now if they'll update hardware in 8 weeks. It would be wise to wait. But....

A little incentive would be the least Apple could do right now when they won't tell us how long, or if, they will reslease Santa Rosa in June. But with all of the delays, problems, and people taken off of OS X for the iphone (so to speak)... give us something!!!:
 
Someone mentioned something about heterogeneous processors... Would that be a multi-core/chip setup with some of the processors design-optimized for specific tasks? Like... an 8 core setup with 4 regular cores, and maybe a dedicated h.264 encoder/decoder core, an SSE4 core, a dedicated floating point core, etc? Could that work? Would software have to be rewritten to make use, or could good OS scheduling do the trick?

in the new multi-core world, this seems like the way to go! I love the idea. are there any production systems that do this successfully out there? How long 'til it makes it into a Mac, 2010?
 
Apple should decide what they want. I assume that they want every home and business and pro out there owning a Mac computer. Look how much money they are spending on advertising for the new ads (Get a Mac). Look at how much floor space is dedicated to computers in the Apple stores (I know, ipods are probably floating the rent). But they could have stores half their size if they were just selling ipods, etc.... which they could use to hire more people/programers :D

Yet, Apple has nothing exciting for the average Joe right now. Even with the release of the 8 core option, most people on this site (pros) will wait for Penrym, Leopard, and/or possibly new designs (2008?).

Since Apple dropped computer from thier name, and only has 5% of the computer market, something has got to get thier core-base/fans excited, which will translate into word-of-mouth promotion. Case and point: I am ready to buy a Mac ($3,300 budget) and now my little cousin, collegebound, is sold on them, and she's ready to buy a 15" MBP ($2,500 budget). We don't want to buy now if they'll update hardware in 8 weeks. It would be wise to wait. But....

A little incentive would be the least Apple could do right now when they won't tell us how long, or if, they will reslease Santa Rosa in June. But with all of the delays, problems, and people taken off of OS X for the iphone (so to speak)... give us something!!!:

just get a machine dude. no matter what apple does or doesn't do should not effect your personal agenda. if you need it, get it. if you don't need it, don't worry about it untill you need it.

do you really think your cousin would notice a speed bump from 2.16 GHz to 2.33 on the low end MBP? i really doubt it.
 
Case and point: I am ready to buy a Mac ($3,300 budget) and now my little cousin, collegebound, is sold on them, and she's ready to buy a 15" MBP ($2,500 budget). We don't want to buy now if they'll update hardware in 8 weeks. [/I][/B]

C'mon Apple... please give us an update... if not for us, then for the poor little children with no MacBook Pros to keep them warm... :D
 
I have the same feeling. The slowdown is probably b/c the OS & software isn't optimized to go so fast.

Nope. The main reason is IO (i.e reading from disk), which for that specific operation (opening a program) becomes the bottle neck. As hinted a couple of posts earlier, as long as we keep using mechanical disks don't expect to see much speed-up in that area.
 
If you want good software you have to build your own hardware

Intel is build and optimised for the MS-World. Using those chips brings you sooner or later down to that level.

This is a challenge also for Apple. If you look at e.g., SUN Solaris 10, then you see what others can already offer software wise. 3D desktop was a SUN initiative, the new ZFS file system is also from them.

By just using mainstream hardware, in the long run you will loose your best people in the company, as that is not challenging work for them.

The future Intel offerings are architecture-wise just aiming at replicating what others already do, e.g., Cell. So-far they just played their fabrications capabilities which are at the moment (still) ahead of the rest.

I hope Apple is reasonable enough to realise that.
 
Keeping up with the Moores

At what point are we going to level out with these chips. I know it's great we just keep getting faster and faster chips, but damn! Every week it seems something better is out, that is **% faster than before.

I just saw Intel's European head on CNN and he said that Intel is on track to keep pace with Moore's law for the foreseeable future. He says that in order to compete with AMD, they have shifted from a three year product cycle to a two year product cycle. In the first year, they release the more powerful chip and in the second year, they shrink its size. Then they release a new chip, shrink, etc. When questioned about how they can keep getting faster, he said that reduced size and power consumption reduces electron leakage across the chip, allowing things to keep rolling along.

In other words, no backing off for now. For the moment it appears that Apple is finally riding the right horse.
 
There is no such thing as a mobile Penryn chipset. Its a chip. The next mobile chipset after Santa Rosa is Nehalem. And like Santa Rosa. It will use the Penryn processor initally then switch to a even smaller processor (tick-tock). If the mobile Penryn processor is released in 1H2008. Then expect to see it in laptops around March/April and in Macs in like May/June :p .

Penryn in Q2 2008 at the earliest. Which means we better not skip it. Because there are a lot of benifits of Santa Rosa (personally I think the processor bump is nothing compared to Santa Rosa, Core Duo to Core 2 Duo wasnt much of a leap.) So, if they were to skip Santa Rosa, I think it would be detrimental. People already dont want to buy because they are not getting Leopard but when Santa Rosa comes rolling around next month with faster FBS, Robson caching, improved battery life, and alittle bump of speed AS well. Then I doubt Apple is going to think its a good idea to bypass the entire christmas season without one of these slick new machines.
From the website:
The update to Santa Rosa will be based on a mobile Penryn processor, the name given to the upcoming 45-nanometer shrink of Intel's current chip designs. The first Penryn chips will be produced later this year and the updated version, or refresh, of Santa Rosa will hit the market during the first half of 2008.
Emphasis mine. That's what I mean by chipset. It specifically says first Penryn chips will be produced later this year, with an updated Santa Rosa platform chipsets and mobile Penryn processor in 1H2008. Also it states later in 2008, a quad core mobile Penryn processor will be available.
 
After being a pc guy for so long, i have gotten use to intel vomiting up performance #'s before the chips released. I have no doubt that these chips will be faster, but come on...

Keep in mind that these are not just fantasy numbers though. Penryn may not be in full scale production yet (3 factories slated for 45nm within the next year, right?), but there are bound to be plenty of engineering samples for benchmarking. I'd expect those numbers to get even better as the factories get up to speed and the process evolves/improves.

Clarification time:
- A chipset is not a processor nor a "platform". It'll be either a memory controller or a combined memory/graphics controller.
- A processor is just, well, the processor... for now... Expect Intel to be putting out mainstream systems on a chip (SOC) within the next few generations (as put out at the Beijing IDF yesterday).
- A platform is a group of products designed to sell, er... work together - a processor, a chipset, and a wireless component for example.

And for those dreaming of putting "oh so powerful desktop processors" in their skimpy wimpy laptops, Intel also announced plans to reduce power consumption by a factor of 10 by 2010. That's huge! Wish granted... eventually.
 
So Tired!

just get a machine dude. no matter what apple does or doesn't do should not effect your personal agenda. if you need it, get it. if you don't need it, don't worry about it until you need it.

do you really think your cousin would notice a speed bump from 2.16 GHz to 2.33 on the low end MBP? i really doubt it.

Listen up I'm so sick of the just buy it dude crap. GanleyBurger, Me and most newbies want the latest. Is that too hard to understand? I think apple is just trying to get rid of the old stuff myself. I say just wait and don't settle. Let them get Stuck with old line maybe then we'll get something out of apple!
 
Yes it is..

The 15% performance boost in "imaging-related" tasks, i.e. probably the Photoshop and Flash programs of this world, isn't really all that impressive for a processor just under 14% faster in clock speed.

Not that I wouldn't some of these chips, that's just not a very impressive statistic to quote!!

Actually, that is a very impressive number. There isn't a chip made.. ever.. that increases performance at a rate that scales faster than clock speed increases. If you take a 2 GHz Core2 and clock it to 3GHz it won't be 50% faster. There are too many other bottlenecks.
The very fact that Penryn performance is scaling faster than clock at 3+ GHz is huge. The memory bus is less than 1/3 the clock speed of that processor and it still scales faster than clock increase over the previous revision of the same processor [family].

Of course, you also picked the weakest benchmark to complain about.

Granted, this isn't a huge shift in computing but we're talking about a revision of the same design. With a die shrink and some modifications to the current generation of chips [which are already the the best consumer processors around] Intel has pulled more performance out with less power draw.. more performance per MHz even. It's a very impressive feat.

ffakr.
 
Listen up I'm so sick of the just buy it dude crap. GanleyBurger, Me and most newbies want the latest. Is that too hard to understand? I think apple is just trying to get rid of the old stuff myself. I say just wait and don't settle. Let them get Stuck with old line maybe then we'll get something out of apple!

But, honestly, "buy what you need when you need it" is pretty good advice.

Look at Intel's roadmap - you can see what's planned for the next couple of years, with increasing clarity as the date gets closer to today.

Intel will always be teasing - don't you expect that by the time Penryn is ready to ship that Intel will have shown Nehalem running at an IDF?

Ask yourself if the incremental improvement in the next chipset is worth putting up with your old system for how many more months....

My "old system" is a Napa64, so I don't really need a Santa Rosa on day 1 - but a Penryn Santa Rosa is in my forecast.
_______________________________________________________

Look at poor MultiMedia - he was drooling over the prospect of a octo-core, but then he heard of the next Stoakley-Seaburg chipset. Suddenly, the octo is untenable, and he needs an SS. I wouldn't be surprised if, after this IDF, he decides the SS is just a stopgap, and he should wait for a Penryn octo-core with "real" quad chips.

Don't fall into the trap that's claimed MultiMedia !!
 
Maybe it is because I just finished a thesis on scanning tunneling microscopy, but 45 nm is absurdly small. Does the die size mean that the entire chip is 45 nm x 45 nm, or that the smallest feature size is 45 nm? I wouldn't think that it could be the former, because that would mean that the length of the chip is only about 200 atoms. Does anyone know the smallest die size that has been done in the lab? Is this it or are they working on yet smaller versions that are just not in a preproduction phase?
 
This makes an upgrade of the current MBP line even more imminent. The longer they wait the closer they come to the Penryn release. My bet is on May or June for the new MBPs and February 08 for the Penryn MBPs.
 
Maybe it is because I just finished a thesis on scanning tunneling microscopy, but 45 nm is absurdly small. Does the die size mean that the entire chip is 45 nm x 45 nm, or that the smallest feature size is 45 nm? I wouldn't think that it could be the former, because that would mean that the length of the chip is only about 200 atoms. Does anyone know the smallest die size that has been done in the lab? Is this it or are they working on yet smaller versions that are just not in a preproduction phase?

If I am not mistaken it refers the average feature size. But it is incredible by 2009 they will be at 30nm. :eek:
 
Maybe it is because I just finished a thesis on scanning tunneling microscopy, but 45 nm is absurdly small. Does the die size mean that the entire chip is 45 nm x 45 nm, or that the smallest feature size is 45 nm? I wouldn't think that it could be the former, because that would mean that the length of the chip is only about 200 atoms. Does anyone know the smallest die size that has been done in the lab? Is this it or are they working on yet smaller versions that are just not in a preproduction phase?

LOL

the CPU is some 5cm wide. it's the transistors that are tiny.
 
Nope. The main reason is IO (i.e reading from disk), which for that specific operation (opening a program) becomes the bottle neck. As hinted a couple of posts earlier, as long as we keep using mechanical disks don't expect to see much speed-up in that area.

That's true, too. I think I said that in a previous post somewhere. I'm sure the IO & the software are both bottlenecks. Which one is more, I'm not quite sure.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.