Slashdot Mirror
ArsTechnica Compares the P4 and G4e: Part II
Deffexor writes "It looks like Hannibal of ArsTechnica fame has put Part 2 of his original comparison article between Intel's P4 and the Apple/Motorola G4e. In a nutshell, this second article covers the execution core, the AltiVec unit and SSE2, as well as a myriad of other interesting factoids. An interesting read, if not a little technically intense for those of us with less than a CE/EE degree. Have at it boys!"
This is exactly what I've been trying to find out for some time now. I've been increasingly upset with the x86 line of chips since it seems that there is hardly any diffrence between 600Mhz and 1.2Ghz.
Here's another comparison: Joy Of Tech (and the next 6 pages as well)
This is the place where you write something that will make you seem like a complete idiot.
>An interesting read, if not a little technically intense for those of us with less than a CE/EE degree.
Tell me about it, I do have more then CE, two letters even, namely MCSE and even I had to stop when they started throwing around the heavy stuff. I mean, A = A + B is supposed to make sense even if B isn't equal to zero.
I intend to live forever, so far so good.
Big article, only had time to glance over it (and I'm not technically qualified to understand it in its full detailed glory), but as far I can see, the dude isn't picking sides. Wich is a rare treat.
:)
Although he does confirm Steve Job's words of wisdom: Mhz aren't everything
(I, on the other hand, am picking sides)
You can't take the sky from me...
A more interesting comparison will be to pit the P4 against the comming G5. According to the Register, Apple has begun seeding early G5's at up to 1.6GHz to key devlopers. Other sources are claiming limited yeilds in the 2.4GHz range already.
There's still bugs to be worked out before production ramps up for release early next year, and supposedly AltiVec will not be as strong on the G5 as it is in the G4. But at 2.4GHz on an already-superior FPU, who needs it?
i personally believe that flexibility of the assembly instructions as well as the number of instructions executed per cycle contribute greatly to the dominant speed (at any given MHz/GHz) of the ppc processor. compare any intel/amd processor to a ppc at the same clock speed, and the ppc will kick its x86 ass.
the high end ppc desktops are topping out around 900MHz, while the p4's are hitting 2GHz. there has to be another explanation besides the complaint that jobs is ignorantly sitting on his thumbs. i think he knows what he's doing.
note: i am not a mac zealot.. i don't even own a mac - only 4 x86 pc's (1 athlon, 2 p133, 1 p120). i simply can appreciate the speed of the ppc.
"I just want to thank my coach Eric a.k.a. Disco for shattering my reality..."
But with the G5 around the corner, I think THAT will be THE interresting comparison.. expecially since Intel plans on keeping the P4 for a while (, ramping it up in speed, when you Read adobe saying the G5 are significantly faster than P4 (and if you go read the article, the same people do say that the P4 is faster than a G4 (exept for altivec stuff) so if they say G5 is faster than P4, it probably will be :)...it should be really nice to see something that kills the P4 in raw performance other than AMD).
--- Metamoderating abusive downgraders since my 300th post.
Note: I have a B.S. in computer science, a solid understanding of hardware issues, and have been programming for 19 years.
When I read articles like this, there's so much detail that I find myself--even willingly--losing sight of the big picture. Sure, you could read a detailed write-up about Toyota's new engine, but those details don't really matter much unless you've just made a hobby of knowing about engines. Realistically, you'll have a hard time connecting those details to your driving experience. Heck, someone could put in a different engine, tell you that its a Toyota, and you'd be saying things like "Oh, yes, this feels just like a Toyota, I can tell that the designers did blah and blah."
After the Pentium II generation of CPUs, things have gotten very, very muddled. Amazing features that are supposed to increase performance don't always do so. Sometimes they make things worse. Little compiler tweaks can make one program be twice as fast as another, given the same hardware. Chips with higher clock rates can be significantly slower than chips with 20% slower clocks. Certain applications run much faster than on previous chips, but there are others that show no increase.
It's all very chaotic and confusing, even for people in the know. I suspect that if you took a program that people claimed to need a P4 or Athlon for--something very performance sensitive--and set yourself the task of making it run faster on a PII than an Athlon, you could do it. But that doesn't matter, as everyone seems to be clamoring for newer chips.
This article is extremely informative and gives you a good insight into how these processors are designed, as well as how they compare. I disagree with the poster though, you don't need a CE or EE degree to get the idea of what's going on. I'm a CE and I had classes on this sort of thing so yes I could follow all the gritty details, but I think the author did a good job of explaining things so that most people could understand. Also, I thought the author summed things up perfectly saying:
The preceding discussion should make it clear that the overall design approaches I outlined in the first article can be seen in the execution cores of each processor. The G4e continues its "wide and shallow" approach to performance, counting on instruction-level parallelism to allow it to squeeze the most performance out of code. The P4's "narrow and deep" approach, on the other hand, uses fewer execution units, eschewing ILP and betting instead on increases in clock speed to increase performance.
This is exactly the case. Unfortunately the popular masses don't understand all of this wide vs narrow stuff, so they go for the higher clock speeds. In reality, Intel is really pulling one over on us, charging more money and all we're getting is a higher clock rate, not a whole lot of performance gain. PPC has proven itself time and time again to be the better processor, but unfortunately they aren't used in very popular machines (mostly Macs,) so we don't get to reap the benefits.
On a related note, this article touches on one of the many reasons why the Gamecube will run circles around the Xbox. GameCube's processor is a 485Mhz PPC designed specifically for video games, while the Xbox just uses a common Pentium running at 733 MHz.
This all brings up a good question: why haven't Macintosh's or GameCube's marketers come up with a bench mark to put next to the processor speed? Maybe I missed it, but I've never seen a Macintosh commercial saying "comes with a G4 800 MHz, comparable to a P4 1.5 MHz." There might be too many legalities involved to do something like that, but it seems like they need to educate people somehow of the non 1 to 1 relationship between clock speeds of P4s and PPCs.
~ now you know
Like with a Tootsie Pop, you start licking, and finally get too impatient and just bite the damn thing.
Mmm...chocolatey goodness...
Carl G. Jung
--
"With one breath, with one flow, You will know Synchronicity" -La Policia
The two operand Intel architecture does not allow the fused multiply add, so that the latency of such an operation is the latency of a multiply plus the latency of an add (and the destination register has to be one of the operands, although the other operand can be in memory, saving you a load). There are plenty of practical algorithms which benefit greatly from the fused multiply-add, for example polynomial evaluations, matrix multiplications, etc, a feature pioneered by IBM in the RS6000 series and that Intel is using in Inanium.
And people who claim that you can do loop unrolling to hide the latencies should check their math: with only 8 registers, there is no way to hide the latencies of a multiply plus an add on a P4, while it is almost trivial on a G4 (32 registers and shorter latencies between accumulates). Furthermore many transcendental function evaluations are evaluated in libraries through polynomial approximations, which cannot be unrolled nor easily sped up: the number of coefficients is usually large enough to make the routine limited by the latency of the back to back floating point operations, but not large enough to take a divide and conquer approach.
While the G4 is clearly the better architecture (not having double precision Altivec is not that important, I consider vector processing is only worth if you can do more than 4 elemnts per vector), the memory susbystem of the P4 is far superior. Hopefully the G5 will be comparable in this area (and I can't buy a desktop Power4 system :-().
----snip----
add A, B
mov C, A
The first command adds the two numbers, and the second command moves the result from A to C. Of course, you still have the potential problem that the original value of A was erased by the add command, so if you wanted preserve A's value then you'd have to insert even more instructions to store A in a temporary register and then restore its value once the addition has been performed.
----snip----
Not quite. I'm sure even people who _dont_ know x86 assembly language will realise all you don't need any extra instructions at all. Simply reorder them:
mov C, A
add C, B
Obviously, the example was being used to show how much nicer it would be to have three or more operands in your instructions, but it was a lousy example.
On a sidenote, we've been able to specify more than two operands with certain instructions since the 80386. Look up the syntax for the "imul" instruction.
Today's weirdness is tomorrow's reason why. -- Hunter S. Thompson
This all brings up a good question: why haven't Macintosh's or GameCube's marketers come up with a bench mark to put next to the processor speed? Maybe I missed it, but I've never seen a Macintosh commercial saying "comes with a G4 800 MHz, comparable to a P4 1.5 MHz." There might be too many legalities involved to do something like that, but it seems like they need to educate people somehow of the non 1 to 1 relationship between clock speeds of P4s and PPCs.
Cyrix used to sell PR parts, PR133 might have been a 116Mhz chip, but it was as fast as a 133Mhz pentium. So there's precedent, and it's probably legally OK, but I suspect the reason is it doesn't really matter.
What really matters is that the CPU is fast enough for what you want to do. I run OS9, OSX and linux on my machines. My home machine is a G3/350, and it's plenty fast for running OS9 for everthing but compressing MPEG1 video. It's not fast enough for running OSX. My work machine is a G4/400 and it's just fast enough for running OSX. But it's not fast enough for compressing MPEG1 video. If I had a dual-800 G4 it would be more than fast enough for OSX, but it would still be too slow for compressing MPEG1 video. My linux machine is a Dual-800, P3. It's just fast enough for running linux with all the crap I have running. It's still too slow for compressing MPEG1 video, though. I also use a 1.2GHz Athlon machine occasionally, and I consider that just fast enough to run Windows 2000. I assume XP is similar. But it would still take a long time to compress MPEG1 video.
So, how would you structure a comparison benchmark? SPECint? BYTEMark? PhotoShop duals? I think the answer is that you don't. It doesn't matter, as long as the computer is fast enough to do what you want it to do. The semi-annual MacWorld Photoshop duals are interesting since they actually show that the computer is too slow for designers but the Windows machines aren't any better. Perhaps they need to enunciate more, but I think their current stand of , "it's fast enough," is the mature one.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)
Next time, I'll include those humour tags for those who do not recognize it :)
:-)
As for the math part, (yes I knew it was about computer instructions) there isn't a single problem with A = A + B with B not equal to 0. Modulo arithmic for example. But then again, I am not sure what grade math level you need to have for that
I intend to live forever, so far so good.
If you are a consumer who wanted a comparison to decide which kind of computer to buy, you are right the article was (mostly) useless.
BUT, for the audience the article is intended for - geeks, technophiles, nerds & propeller heads it was not pointless at all. On this forum in particular there are a lot of people that use neither Windows nor MacOS but other operating sytems which run happily on either processor. Even if there is no *practical* point there is always sheer geek curiosity - alot of us find such articles entertaining.
Might as well have Car and Driver running a comparison of a Jaguar S-type and a 10-ton dumptruck.
I don't think that the difference between a P4 and a G4 is quite as wide as that - and they are being marketed by both sides as roughly equivalent products. Most techie people may know which is the "dumptruck" and which the "Jaguar" but it is still interesting to see a technical explanation of WHY and precisely HOW they are so different.
The whole market for motherboard upgrades comes from this situation. Apple does not support motherboard that it did not manufacture, and the OS used to check for the precence of genuine ROMs. So third parties could not build replacement boards. By upgrading only the CPU subsystem, the rest of the motherboard would remain genuine Apple and therefore run the system without problems.
Also remember that Mac hardware tends to be more expensive and last longer than PCs. While the performance boost you win by upgrading only the CPU system is lower, the impact on the workstation is also lower. Changing a motherboard means changing the system, having new drivers, so basically more maintainance work.
This situation might change with darwin, theoretically, nothing prevents some company from producing PPC motherboards, recompile Darwin for it and then build a installer that instals OS X on top of Darwin. Old machines that Apple does not support can run OS X this way.
Yes XP will run on it using VPC. The question is why.
On the G4 you can run OSX, OS9, XP and rootless XWindows all at the same time. The only problem is you have to reboot to run Linux. But then you can run the MacOS from within Linux.
Flexibility of the Mac is one of its strong suits. Check out the different Gnu Darwin, Darwin, and Xon X sites. That is where the action is
Yes I am running BSD, you still running Windows?
photosMy Photostream
A good general resource for this kind of advanced computer architecture is the book Computer Architecture by David Patterson and John Hennessy. It's quite dense. For the latest in processor architecture, the IEEE Micro magazine is useful.
The two operand Intel architecture does not allow the fused multiply add, so that the latency of such an operation is the latency of a multiply plus the latency of an add (and the destination register has to be one of the operands, although the other operand can be in memory, saving you a load). There are plenty of practical algorithms which benefit greatly from the fused multiply-add, for example polynomial evaluations, matrix multiplications, etc, a feature pioneered by IBM in the RS6000 series and that Intel is using in Inanium.
And people who claim that you can do loop unrolling to hide the latencies should check their math: with only 8 registers, there is no way to hide the latencies of a multiply plus an add on a P4, while it is almost trivial on a G4
Actually, it turns out that you can still mask the loop latency with a limited register set.
First, you can use "software pipelining" to mask quite a bit of the loop latency without having to unroll (it's a clever reshuffling of the loop instructions; for brevity, I won't describe it here). This requires one extra FP register over the straightforward implementation of an x86 dot-product loop (four instead of three, because I can no longer re-use scratch registers between steps).
Second, branch prediction will to a limited extent perform unrolling for you. While the architectural register file has only 8 registers, there are many more internal registers on the chip. Register renaming allows the processor to run several iterations of the loop in parallel without having to worry about namespace conflicts (though true dependencies remain intact). This works as long as the total number of iterations being unrolled fits within the scheduler's window (usually 8-16 instructions; I don't know how big the P4's window is).
In summary, for something as straightforward as a dot product, it's certainly possible to write x86 code that will avoid the penalty of having separate add and multiply instructions.
[You'll really be bound by the memory subsystem for both chips, but that's moot point for this discussion.]
Thats a very amusing comic, much more so than UF. Any idea why its never mentioned on SD?
Yes, if you have to wait for a computer it's too slow. Realtime is too slow. If anything takes more than 10 seconds it's too slow.
But MPEG2 implementations typically do *way* less compression than MPEG1. MPEG2's bitrate is typically at least twice as high as MPEG1, sometimes 10x more, and when you're doing discrete cosine transforms that greatly increased size along with motion vectors and variable bitrate encoding allow you to do alot less work during encoding. That's why it's so much faster. You also need alot more space to store the resultant file (DVD vs. CD) Just because 2>1 doesn't mean it's compressing more, it just means it came after.
My God, it's Full of Source!
OUTSIDE_IP=$(dig +short my.ip @outsideip.net)