“Intel dual-core notebook processors, (codenamed Yonah), will be priced similarly to the current prices for the 700-series Pentium M processors, according to sources in the Taiwan notebook industry,” Charles Chou and Jessie Shen report for DigiTimes. “Details of the dual-core notebook processors were revealed at IDF (Intel Developer Forum) in San Francisco in September 2004 and the processors will be launched in the first quarter of 2006.
Chou and Shen report, “The X-series Yonah processors will be part of Intel’s Napa dual-core notebook platform, which is also expected to hit the market in the first quarter of next year, and will represent Intel’s transition from 90nm technology to 65nm. In addition, the Napa notebook architecture will be the basis for Intel’s digital home products. In addition, Yonah-series processors will be adopted by Apple Computer, which agreed to use Intel microprocessors for its Macintosh computers starting in 2006.”
Full article here.
MacDailyNews Note: In a chart accompanying the DigiTimes article, Q1 2006 prices run from US$241 for the Intel “Yonah” X20 dual core processor running at 1.66GHz with a 667MHz frontside bus to US$637 for the Intel “Yonah” X50 dual core running at 2.16GHz witha 667MHz frontside bus.
Related MacDailyNews articles:
Analysts think ‘Yonah’ Pentium M may power Apple’s first Intel-based Mac – June 08, 2005
The frontside bus only needs to fast enough to not be a bottleneck – kind of like memory. If you keep adding memory, you will eventually eliminate your memory bottleneck and your computer will no longer work faster if you add more.
“However, still what are the advantages of having a fast Frontside bus?” – macintosh16tx
Huge throughput. Applications operate on data. Without it, applications are worthless. Some applications only need to process data over and over and only require additional data or need to ouput data occasionally. For these apps, you won’t see much improvement from high FSB. Some applications, however, requires moving chunks of data in and out to be processed. The faster you can get the data in and out, the better performance you see. For this type, FSB can be a bottleneck and the processor is starved of data. No data, no computation. So, even if you have a fast CPU, it’s doing nothing mostly.
For example, video apps. They require streams of data and if you have a bottleneck in your FSB, the video may not play smoothly even if you have enough CPU power.
“The G4 and earlier PPC cpu’s are the same endian as the Intel platform. On the other hand the G5 is opposite.”
Actually fenman, you have that a bit wrong (no pun intended). The 68×00 and PowerPC chips have always been reversed from the x86 chip. That’s why Virtual PC and other low-level apps had to do some memory flipping. I believe with the introduction of the G5 chip, there was an option to go either direction and that caused some low level problems with Virtual PC.
Now with the adoption of the Intel architecture, we can erase one more roadblock to fast emulation.
Naraa-
These aren’t the prices Apple is getting. These prices are for lots of 1000. If Apple isn’t getting a significant discount on these prices, they’re getting hosed. That’s one of the major reasons for the switch (other than the roadmap) – significant volume discounts from Intel. These are the prices that “Joe’s Computer Shack” will be expected to pay.
Everyone is wrong about endianess. (and it wouldn’t be the first time for me, but…) The G3/4 had a special mode where it could use little endian variables, but its natural state was big endian. The G5 didn’t have this, and that’s one of several reasons why it didn’t do VPC as well as one might expect. As far as data format on a disk, Mac could still use HFS with little endians all over it, as long as the processor that reads it knows this. If they do switch to little endian on HFS, then something is gonna have to flag this modified HFS format so that older Macs won’t stumble and fall if you hooked up a drive from a MacIntel to a PowerPC Mac. So, they could stay with HFS (lil endian HFS that is) if they wanted to.
There are a lot of misconception on filesystem, it seems.
Mac OS X can already read and write to different filesystems. It supports HFS, HFS+(Mac), UFS (Unix), FAT12, FAT 16, FAT32 (MSDOS), etc.. You can read/write to Windows floppies for years before Apple dropped floppy drives even when Mac OS was still called System x.x. You can install MacDrive and read Mac formatted disks on Windows.
Whether you can boot an OS from an unsupported filesystem is another thing. The bottom line is, the endianness of the CPU is not tied to a filesystem exclusively.
More on Mac OS X filesystems:
kernelthread .com
Kevin, you said something that tells me you’re an Intel guy
” width=”19″ height=”19″ alt=”smile” style=”border:0;” /> (no offense there) If you look around, and look way back in the past, you will find that it is Intel that has “always been reversed”. The Intel architecture is about the only one that is bass-ackwards. Every other processor that has any traction at all is most significant byte first.
I remember when I got my hands on my first 8080 based computer, I just could NOT wrap my head around least significant byte first. That was back when looking at memory dumps was pretty much routine for a debugging session. Like, when I write down a really big number on paper, I was taught in grade school to do “345,983,912”. But the equivalent in Intel’s memory would be “912,983,345”. I still can’t do this well.
does anyone know what the G4 chips they are currently putting in the powerBooks and macMinis cost? I keep saying the Intel chips should be cheaper than the IBM/Motorola ones, but do we know for sure?
Who knows what prices Apple will get for the processors.
A Dual core 2 Ghz PB sounds neat.
I am concerned that for this processor to be produced Intel will move from 90 nm fab to 65. Isn’t moving to a smaller process one of the reasons IBM screwed up with the G5? Of course, they had other problems as well.
I kinda wish Apple would stick to chips that are available in volume. In their lust for the next best thing, they always seem to run into supply issues.
I can just see it now:
SJ announces Intel-Mac PB with the latest and greatest CPU at the WWDC 2006. Available the following month
One month later, Apple push back release date.
Two months later, the first few PB trickle out.
Apple take 9 months to catch up with demand.
In the end a year has passed whilst Apple could have sold a ton more Macs if they hadn’t been so ambitious.
MW – test as in a test to see if Apple can get it right this time
Neil-
Intel is moving to a smaller process for a number of reasons.
The smaller processors (65nm) generate less heat and consume less power than a clock-speed-equivalent 90nm processor. This is important for laptop chip (can you say “battery life”?).
Secondly, when you’re sqeezing 2 cores onto a single die (which is what the dual-core Yonah will be doing), shrinking the process allows the use of less silicon, resulting in a much cheaper processor. If the process were to remain the same, the cost of silicon (the primary cost of the chip) for a dual-core chip would be the same as the cost for two single core chips. That would make the chip prohibitively expensive for all but the highest-end systems. With the smaller process, however, the dual-core Yonah is scheduled to be priced in-line with what we’re paying for single-core chips now.
Manufacturers have had problems switching processes in the past, but the transition from 130nm to 90nm went smoothly and it appears the 90nm to 65nm process won’t be any harder.
First of all.. the G3/4/5 were all big endian processors. In fact the Mot 68k was also big endian. Almost any processor I can think of is big endian… except for intel. Dont ask me why.
Second… I am not certain about this.. but I dont think that the endianness of a processor has anything to do with the disk format. When I save files on a PC disk on my mac.. the bytes are not all flipped around. My guess is that the different file formats explicitly specify correct endian, and each OS is responsible for correctly reading and writing this from memory.
“Kevin, you said something that tells me you’re an Intel guy”
No Wingsy! Just the opposite (I even preferred the Z80 to the 8080). I’ve always been upset with how Intel built chips that contained bugs (80286) and then they had to keep those bugs in newer designs to stay compatible.
I’m just bass-ackward myself at times.
” width=”19″ height=”19″ alt=”wink” style=”border:0;” /> Thanks for clearing up the “ends” issue. I should have looked up the articles again instead of trying rely on my memory.
The good news is, even Intel can do great technology and I’m sure Apple will help them.
“Also, let’s hope Apple can avoid the “BIOS” as we currently know it.”
-TheCrunge
BIOS as WE currently know it? Your speaking Greek to me.