“Intel has started shipping its 65nm dual-core mobile Pentium processor, Yonah, along with Centrino bundles based on the new processor. The five processors are clocked at 2.16GHz, 2GHz, 1.83GHz, 1.66GHz and 1.66GHz (single core), and operate on a 667MHz frontside bus, connecting across a FCBGA6 or FCPGA6 pin-out. All processors sport 2MB of L2 cache.” In addition, “Intel has debuted its latest dual-core desktop Pentium D processors. The chip giant this week added four models to its price list; all containing 4MB of L2 cache split into two 2MB chunks, one for each of the two cores in the chip package. All four operate on an 800MHz frontside bus. The CPUs are clocked at 2.8GHz, 3GHz, 3.2GHz and 3.4GHz.”
Intel Launches Yonah, New Pentium Ds
45 Comments
2006-01-03 5:05 pm
Thom Holwerda
“Did anyone else notice that this was posted anon? How’d that happen?”
The comment could have only been posted by an admin. Hmmm… Which admin posted this article?
Oh, nevermind. There goes that theory… Thom NEVER trolls.
Yeah, or it might be just the fact that we still allow anonymous posting from mobile devices because we haven’t written a login form yet for mobile and text browsers. The user agent at the bottom of the post might be a hint at this. Oh, and we’ve only said that like, 2398473957 times.
But yeah, your option sounds like a big possibility too. An admin has done it. Sure. 
.
Edited 2006-01-03 17:08
-
2006-01-03 7:53 pm
2006-01-03 3:26 pm
Kroc 
When can we expect the first laptops with these though? I got a cheque through for a replacement laptop (insurance) and have been waiting for this announcement.
2006-01-03 3:48 pm
-
Kroc
First line of the summary, 6th word if you need further guidence.
(karma to hell, I needed to say that)
-
2006-01-03 4:36 pm
rayinerTitle line of the guy’s question. Second word if you need further guidence
To answer the OP: no, the new Pentium D’s are 90nm chips, and not 65nm chips.
-
2006-01-03 6:13 pm
2006-01-03 4:51 pm
-
2006-01-03 4:59 pm
-
-
-
Thom Holwerda
Now, what Apple will do with the new chips is another issue. I don’t expect Apple to be able to get away with shipping its top-end iBook with anything less than the dual-core 1.67GHz model. That’s a $243 chip (sans Apple discount), and is appropriate for a $1300 machine. Certainly, that’s what its competitors will have in similar-class machines.
Yes, I’m really looking forward to how the Intel iBooks will turn out, assuming they will turn out on the 10th, of course, which is still speculation.
I’m in the merket to replace my iBook, it’s from October 2004, and now that I have this iMac G5 for review (expect the article this week), you really notice the difference– and how utterly slow the iBook G4 actually is. It’s a sad thing I have to return this G5
. -
2006-01-03 6:22 pm
-
2006-01-04 1:03 pm
-
-
2006-01-04 4:42 pm
2006-01-03 6:17 pm
2006-01-03 6:21 pm
-
-
2006-01-03 10:04 pm
molnarcsdual-core Yonah vs. Single core.
Well, I think single core Yonah will use less power than dual core Yonah, and it will certainly produce less heat. What’s more, it won’t be much faster
– unless, of course, you run concurrent jobs: compiling openoffice while watching movies and ripping your CDs to mp3, all at the same time. Even then, the only difference I see while doing all these things concurrently (on my single core sempron 3100+) is that mplayer starts in 2 secs instead of instantaneously, and of course, these jobs will be finished later. Unless you really need massive concurrent power (I can imagine that on source-based distroes btw) – dual core is simply marketing, for desktops at least, and for laptops or notebooks even more so. For professional workstations, servers handling lots of concurrent jobs (heavily loaded mysql servers for instance) this is an entirely different issue. However, the pathetic 667MHz FSB will become quickly saturated, so there goes your second core waiting for data on the bus to process. That’s what you see on all tests comparing the X2 with Intel’s solutions.
In short: dual core yonah with 667FSB is ridiculous, especially if these low voltage models are intended for laptops. Most of the processes described above (ripping audio/video, compiling software) will have your CPU running @100% all the time, defeating the purpose of mobility, and are typical desktop jobs (you won’t leave your laptop on for 11 hours just to compile openoffice…) It’s marketing: dual core sounds cool and looks great in your dmesg … but the 667Mhz FSB doesn’t look so cool, especially when you see low-end semprons running circles around your shiny dual core when it comes to tasks that are memory bandwidth bound…
At this time (and in the upcoming years)
2006-01-03 6:27 pm
-
-
-
-
2006-01-03 6:42 pm
Hamman
5) Should Celeron still be avoided for anything more than web surfing and email?
Well, the Pentium M-based Celeron M is basically as fast as the Pentium M. A bit slower in apps that use the cache a lot, but the reason Celeron M is a bad choice for laptops is because they don’t support frequency scaling, so they draw more power and use up the battery faster.
All IIRC, of course 
2006-01-04 2:23 am
2006-01-04 2:33 am
2006-01-04 7:01 pm
What does this mean for Apple and the forthcoming Macs?
It means they’ll be awesome, is what.
Yonah is still 32-bit and very unlikely to be used outside of the i-range of apple products. The G5 is 64-bit and very likely Apple will want to make the powerbooks 64-bit too so they will have to wait until June when Intel launches it’s 64-bit dual-core chip before updating those lines.
PowerBooks have no need for 64-bit. A Yonah PB would be much more powerful than the current models, so hopefully Apple will ship it sooner rather than later.
Prosumers have a *genuine* need for 64-bit. Apple’s powerbooks are aimed at serious users, the same users who use and need 64-bit G5 Macs.
Please list the 64-bit applications for MacOS that you use.
OSX is 64-bit
That alone gives benefits. Matematica is 64-bit and that’s a serious app. Intel’s next chip in June is dual-core 64-bit and likely a lot of software not currently 64-bit will be compiled to it straight away as it has to be an intel binary anyway. Which is a better and more sooner situation than widespread Win-x64 usage.
The kernel is a 32-bit binary that uses a 64-bit datatype for bookkeeping addresses. It does not itself have a >4GB address space.
Mathematica does have a 64-bit kernel if you need to use it. Do you use Mathematica? Do all “Prosumers” use Mathematica? Give me a rough estimate of the number of people who use MacOS that you believe uses Mathematica with a 64-bit kernel.
There is very little software for MacOS X that ships with 64-bit binaries. One reason for this, is that 64-bit code is slower on the G5 than 32-bit code and unless a larger address space is necessary for the execution of the program there is no advantage. Another reason that there is very little software for OS X that ships with 64-bit binaries is that only libSystem and Accelerate ship with 64-bit versions. This requires the separation of the interface from the portion of the program that must work in a 64-bit address space. This introduces latency and complexity that unless you genuinely need to use a larger address space, is not really worth dealing with. Mathematica for example is already developed as a separate kernel process and an optional interface process for all of the platforms for which it is available.
The future direction of Intel’s platform doesn’t support your statement. Your statement was that 64-bit addressing was necessary for the “Prosumer” that purchases Powerbooks and G5s. The current Powerbook uses a G4, so it clearly isn’t an example of a userbase that requires 64-bit addressing. That leaves the people that own and use G5s, which I suspect you will have a difficult time painting as avid users of 64-bit programs.
Being a serious user with a PowerBook, a G5 at home and another G5 at the office, I know all about 64-bit. I don’t use it, because OS X doesn’t use it.
“PowerBooks have no need for 64-bit.”
Today, maybe.
But usually is the user that define the needs, or at least the expectations, not the machine nor, in a correct customer/vendor relationship, the vendor.
In that perspective a 2006 buyer that expect to use the machine for next 3-5 years may clearly see a 32 bit machine as a DOA for the 2006-2009 or 2006-2011 lifespan, and the same would be true even for a low end machine.
In that lifespan the home user should *resonably* expect:
– 2 or even 4 GB memory module become popular;
– newer games (so, non professiona apps) that will benefits of >4 GB of RAM (some recent ones are jet near to that limit in present days);
– newer popular videoediting application, the new frotier of RAM eating software, that will seriously benefits of demolishing the 4GB wall;
– virtualization software (don’t smile, Apple itself is working hardly on that!) becoming popular, eating up ram for each independent machine (i.e. Win + OSX + maybe Linux for most curious users) you’ll start in order to don’t bother in rebooting while changing from different systems, or simply running more instances of OSX for many family users in order to not have to reboot at user change (the son wants to play a memory hungry game while the father session in backgruond is videoediting the dl-dvd of the summer holidays while the mon’s session is busy with tons of frozen productivity apps waiting the mon’s return to home, while the grandfather session in background is recording his favourite soap-opera…);
– since disks will continue with all probability the trend of last 30 yers (being the perirheral that increase speed vastly slower than other ones) chaching in RAM more GB as possible (far more than 4 GB…) will become (and still is from many years) the best strategy in overcoming the most serious computing bottleneck, important also for home users that simply want a responsive system and doesn’t want, nor understand why, to wait long loading times;
– number crunching application like compression, encoding and encryption, today used in most popular software, that will seriously benefit of processing chunks of 64 bit instead of 32 bit.
– release of completely 64 bit OSes, many linuxes are released since last year in full 64 bit form and users are quite enthusiast of thir capabilities.
so even an home user should definitely expect the 4 GB limit of 32 bit addressing and the 32 bit width integer math become soon a nag.
to bad their cpus suck more energy that my mini fridge….
Browser: Mozilla/4.0 (compatible; MSIE 6.0; Windows 95; PalmSource; Blazer 3.0) 16;160×160
Did anyone else notice that this was posted anon? How’d that happen?
“Did anyone else notice that this was posted anon? How’d that happen?”
The comment could have only been posted by an admin. Hmmm… Which admin posted this article?
Oh, nevermind. There goes that theory… Thom NEVER trolls.
Why is this modded up? What proof do you have. Why is it that because you know his name you think you have some kind of executive power to be pointing fingers. Any one who’s been here the last week will recognise the user who always posts his UA at the bottom of every post.
Parent should be modded down on Personal Attack, because it’s exactly that.
Its modded up because its interesting. (Yes, I helped.)
The “proof” is that we can’t post anon, yet this person did.
Actually, he pointed no fingers. He actually obsolved Thom of any wrongdoing. If anything you’re attacking him. You should be modded down as a result.
“He actually obsolved Thom of any wrongdoing.”
Like this?
“Oh, nevermind. There goes that theory… Thom NEVER trolls.”
That was clearly sarcasm. How am *I* attacking Tom for defending him from an ill informed troll –>