Slashdot Mirror
Intel Yonah Performance Preview
illusoryphoenix writes "Anandtech has an interesting preview of the successor to Dothan (Pentium M's second generation), Yonah, with tests run on an engineering sample. It seems like latest Pentium M is still lagging in the floating point area, but has gained some ground overall. It's also interesting to note their comparisons to the Pentium D/Netburst based dual core."
The new Macs are going to kick ass. No more stone-age G4 PowerBooks.
I really have to wonder when Intel will start using this technology in desktops. It really does seem like a good idea. From TFA "At 2.0GHz, Yonah is basically equal to, if not slightly slower than an Athlon 64 X2 running at the same clock speed in virtually all of the tests we ran. " That right there should show that Intell is should switch its R&D and support the Pentium M as a desktop chip.
Yay, I have a sig.
This is a _mobile_ chip being compared to _desktop_ chips. You _should_ be impressed. And when the next generation comes out in 2H2006, Merom, any remaining performance gap will probably be gone, plus it'll then be 64-bit, too, though of course, AMD will hopefully keep making strides in the meantime, with their upcoming socket M2-based offerings.
That this is likely the Intel chip to be used in upcoming Macs is a very good sign for future Mac owners like myself.
because intel makes the memory controllers... and they make money off the memory controllers... AMD could never make a northbridge worth a shit if you remember. Intel makes a boatload off their northbridge, they don't want to cannibalize their own sales, even when it means they lose performance... that's my theory.
So will a athlon X2. At least it has all the technical requirements. Runs fine on a hacked OSX86.
It's easier to fight for one's principles than to live up to them.
I suggest you look up the overview of how they make processors. You'll see it's an entirely "analogue" procedure.
A simpler analogy would be egg "production". They take 100s if not 1000s of animals laying eggs, a certain percentage are duds [e.g. not fit for human consumption], certain percentage are small, medium, large, etc. The same basic process is used in each case. Feed animal, wait, capture egg, rinse, repeat.
It isn't that they "shrinkray" some eggs and sell them as "small" it's that they ended up that way.
Similarly when you shoot the laser [or interference pattern] through the mask to hit the die the light may be slightly off meaning the transistor may not be entirely in place and as a result take longer to switch [or not at all, e.g. dud]. The result is a chip that in order to meet the clock period overall has to be clocked slower.
Remember that the processor is as fast as the slowest clock domain part. So if your 50K transistor ALU [pulling that # out of my ass] has one transistor that is 20% slower the entire ALU must be clocked 20% slower or it'll fail.
A way to mitigate this would be to have different clock domains for parts but that would make them slower [more latency] and harder [and larger] to produce. So they design with margins. Your 2Ghz processor has parts in it that are actually meant for 2.2Ghz [or even higher] and account for "worst case" processors their yield of 2Ghz parts ends up being profitable.
The same is true in any digital parts design. A 200Mhz AES core likely can hit 250Mhz or higher in "best case". But customers don't care for "best case" because they want a design they can mass produce reliably. I'm not an EE, I don't claim to know all of the facts but that's the "jist" of it as I got it from working at a fabless hardware firm.
Tom
Someday, I'll have a real sig.
Just my 2 cents, but sooner or later the PC world needs to break away from this fixation on legacy desktop PCs with their Heath Robinson contraptions of wires, grouchy PSUs and naked circuit boards, not to mention size and noise. The line that caught my eye in this review: "A 2.0GHz Yonah under 100% load consumes less power than an Athlon 64 X2 3800+ at idle."
Unless it is for gaming or for special and demanding applications, who needs all this muscle? A few more steps in the Yonah development line and we may be able to see PCs that are far smaller, quieter and more frugal with the juice while still packing a punch.
None of this means that the Ahtlon 64 isn't darn good, only that it is not appropriate for many computing situations. Right now, Yonah looks more like a stab at tomorrow whereas the Athlon 64 represents the apogee of yesterday.
Las qué passoun
tournoun pas maï
The review fails to mention that, unlike AMD's current mobile Turion CPUs, the upcoming Yonahs will not run 64-bit code. What is Intel thinking? With 64-bit OS and software support increasingly available, who will want to invest a lot in such a laptop? Yet dual-core laptops are supposed to be high-end, and, being a more expensive investment, ought to last longer.
BTW, I'm impressed that you actually get useful charge when running CPU and GPU at 100 %. Most systems I've seen will trickle it down, sometimes for lack of power supply, sometimes due to the temperature situation in the battery.
um, that's a desktop disk, peripherals, USB devices, and GPU... and motherboard, not that it matters. They picked them ( well, everything but the mobo ) to match their previously benchmarked desktop system. If you were to actually build a laptop, the total system draw would very likely end up being less... heck, probably that GPU is a good percentage of the power draw.
Does it matter? They benchmarked it against AMD systems running 400MHz DDR memory, and the AMD systems perform better.
Furthermore, faster RAM = more expensive RAM... why pay more money, when I could pay less, buy AMD, and get better performance than Intel?
I am government man, come from the government. The government has sent me. -- G.I.R.