> Okay.. I read the article. It was an interesting mix between pat on the back science > and good old "Hey, aint NASA GREAT!" enthusiasm.
Well, at least that's a good counter to the constant "scientific research doesn't help people, so it's better to spend twenty times on the military than on space research" FUD that seethes through popular media.
> My question.. which I did not see answered, are where ARE they right now? I know they havent > cleared the SS yet, but where exactly are they? ARe we going to get pictures Pluto and Neptune > back?
http://www.vttoth.com/probes/probes.html
That's an okay list of current space probe locations, though it's not in any kind of detail.
None of the deep space probes are anywhere even remotely near Pluto or Neptune.
http://vraptor.jpl.nasa.gov/flteam/weekly-rpts/c ur rent.html http://spaceprojects.arc.nasa.gov/Space _Projects/p ioneer/PNStat.html http://www.schools.ash.org.au/ rochedale/solar9.htm
The above URLs should show you that both Voyagers as well as Pioneer 10 are at least ten billion kilometers away from the Sun, and they are leaving the solar system at greater than minimal escape velocity. Pluto, when furthest from Sol, is 7.4 billion kilometers away. So the probes are anywhere between 2.6 and 17.4 billion (or more) billion kilometers away from Pluto. Which means that we won't learn much from them about our little wanna-be planet.
> (Which would be GREAT.. and would solve that long running question of whether Pluto is even a > planet, a bit asteroid, or a half a planet that got pulled into the gravity well here).
The long-running question is not science based, really. We already know the mass and diameter of Pluto. Heck, we even have a (http://oposite.stsci.edu/pubinfo/PR/96/09.html) surface map of Pluto, thanks to Hubble (the satellite, not the man). The debate over whether Pluto is a planet is dependent on two things: (1) we don't really have a specific definition of the word "Planet", and (2) the common masses would get pissed off if Pluto got demoted, even if it was a rightful demotion, since they were taught as children that Pluto is a planet
> Does it even have the transmitting power to send real data back anymore? > or simply to weakly croak "I am here".
Well, Pioneer still performs maneuvering commands when requested. I don't know if its scientific resources are useable, but the Voyagers have nominal science instrument performance. I do recall that they are using these probes to determine where the point between solar wind and its stellar equivalent become equivalent in strength. I can't tell you much more, as I'm a bit busy at work today.:)
> How do I open videos from Xine's GUI? > I already tried the help files.
I'm about thirty miles away from my part-time linux box, and I have crappy memory skills, but....
I believe that you click the "playlist" button, which is on the left side of the control window. A window pops up, and down the right side it should say things like "DVD" and "VCD" and, most importantly, "file". Click on "file", and yet another window appears. This window basically is a file opening dialog, except that it sucks. You can't actually type in filenames and paths. You have to navigate through by double-clicking or triple-clicking (sometimes the double-clicking simply refuses to work!) directory names and soforth. But it should eventually work.
Of course, that's the Mandrake 8.2 version. I don't know if the new version is refined from this.
First off, while the Xircom SpringPort for the Visor series saved me from insanity, I cannot quite recommend it. Here's a few reasons:
* I'm told that other 802.11b cards have the capability of autodetecting SSID numbers floating around, whereas the SpringPort must have its SSID explicitly defined.
* The power saving feature on the SpringPort is broken. This problem has caused problems with many people who assume that their part is simply defective. Basically, when power saving is enabled, the SpringPort experiences difficulties connecting to some base stations
* It's bloody expensive. I was lucky enough to get one on eBay for just over a hundred bucks, which is probably comparable to current 802.11b CompactFlash cards, but it originally retailed for well over three hundred dollars and it'd likely be hard to get one for under two hundred nowadays.
My strategy was to get a handheld with an expansion option that may be found in future handhelds down the line. I originally got the HandEra 330 because it had three expansion options findable elsewhere:
* Palm III compatible serial port * Compact Flash Type I & II * SD/MMC
I didn't expect much from the Palm III port (though there is a sled that lets you use PCMCIA in limited fashion for that port). But I had a hunch that either of the CF (which I had also used for storage on my Visor Prism via a MemPlug) or SD/MMC slots would find their way into other PDAs. I did turn out correct: You can find CF on the Sharp Zaurus, which is an interesting option, and the SD/MMC option is standardized on the Palm family now. Some of those Microsoft things also support those expansion types. Anyway, there isn't much out for SD/MMC in terms of wireless communication (save for Palm's Bluetooth card, but that's not the same thing really), but there are several CompactFlash-based 802.11b options.
Here's an idea: The TRG Pro, which is the precursor to the HandEra 330, is a less expensive option that still supports CompactFlash. I cannot be certain, but maybe it'll work with the CF 802.11b cards. It's basically a Palm VII minus the built-in internet crapola but with 8MB DRAM and the expansion port. Perhaps you could get one of those on the cheap and upgrade much, much later to a higher-lever device that still supports whatever CF card you get.
That's all I got for y'all. I wish you best of luck on finding the best option for your needs. ^_^
> I know that color can help the GUI a bit in
> quicker human recognition and differentiation,
> but it sucks a lot of power.
A PDA like the Visor Prism or the Palm IIIc sucks power incredibly, but it's not because these PDAs have colour. It is because they use an active matrix screen. Colour PDAs like the Clie 710C and the Palm m505 have active matrix screens and therefore use much less power than that.
I should note, though, that if I were buying a PDA right now from scratch, I think my choice would be the HandEra 330.;)
> I remember the 5x86 chip that was actually an
> overclocked 486 clone. People who bought it
> were honestly tricked they bought a Pentium 133,
> when in fact it delivered performance worse than
> a Pentium 60.
Your comment is amazingly contradictory. AMD's 133MHz 5x86 was equal in frequency but lower in performance to a 133MHz Pentium. This chip had a P-rating of P75.
THINK ABOUT THIS!
You just said that AMD deceived customers by not P-rating a chip of equal MHz. And then you turned around and said that it's a deception to P-rate chips. That was a really silly comment that you made, especially since AMD was willing to market their chips as being lower performance per clock. They very openly marketed the 133MHz 5x86 as being performance equivalent to the Pentium 75. This was no secret, and it in fact was emblazoned on benchmark graphs all over the place.
P-rating, as it was and is used by AMD, is a means of comparing chips in a manner that is far more honest than MHz. AMD used it to show that they were lower performance per clock in the Pentium days, and now they are using it to show that they are higher performance per clock.
I truly don't understant why people are making such a big deal out of it.
-JC
Re:benchmarking
on
Mac Rants
·
· Score: 3, Insightful
> While it is true that most G4 to pentium
> bake-offs are done running photoshop filters, I
> don't think it is a particularly unfair test.
> After all, Photoshop really is the only standard
> application in existance that:
> a) has the same version and capabilities
> for both the PC and the Mac, and:
> b) can actually tax a current machine's
> processor.
> Other eligible apps (ie: Office) fail on
> both these counts.
I disagree with this. There are plenty of cross platform apps which can potentially be used for benchmarking both PPC and x86 computers. Here's a short list:
* Adobe Acrobat
* SETI@home
* d.net rc5-64
* POV-Ray
* Bryce
* Cinema4D
* Lightwave
* Mathematica
* Heuris MPEG Power Pro
* Deneba Canvas 7.0
* MetaCreations Canoma 1.0
* Adobe Illustrator 8.0
* ViaVoice
Additionally, if you just want to compare the hardware, you can install linux on both and run kernel compiles and likely other interesting benchmarks on that kernel.
But that's not even really the point. The problem that the author had was not that the comparison used Photoshop alone. It was that
a) the comparison alleged that the results were of "OVERALL PERFORMANCE"
b) the Photoshop test tested six hand-picked filters, possibly six of those that most prefer the G4 over the P4, instead of a wider range of filters and effects that might actually have more reasonably modeled actual Photoshop performance.
> There is hardly difference between using a 1.8
> Ghz Pentium 4 and a 500 Mhz Pentium 3 when
> surfing the web or typing a paper in Word.
That is subjective. My 500MHz is slow as crud, and I can see an amazing difference even just going up to 800MHz. This is likely because I am more aggressive with my computer use (even with just internet stuff), but it still allows me to object to your statement that a jump from 500MHz to 1.80GHz means little.
> How come I don't see many touting that the 1.2
> Ghz Athlon is some how lacking in ability when
> compared to the 1.8 Ghz Pentium 4?
That is because you haven't looked. There are very few reputable places that show that the 1.20GHz Athlon is superior to a 1.80GHz Pentium 4. In general, the idea is that the 1.33GHz Athlon is in the same range as the 1.70GHz Pentium 4, and the 1.40GHz Athlon is fairly level with the 1.80GHz Pentium 4. Of course, there are benchmarks in which said P4 is much faster, and there are benchmarks in which said Athlon is much faster. But you do not see AMD zealots simply running one of the rare benchmarks that have the 1.00GHz Athlon outperforming the 1.80GHz Pentium 4 and then claiming that these results reflect overal performance. That would be going over the line, and Scott Wasson's point in his rant was that the comparison between the G4 and various P4 boxes went over the line in this fashion.
> This author really doesn't know what he's talking about.
This message is a case of the pot calling the kettle black. In other words, you made this claim about the person, and you then followed up by making a statement that rather conclusively proves the same about yourself.
> The G4 processor is very powerful and the Mac architecture is well designed.
> As far as speeds go, Macs use a different
> instruction set technology, so an operation on a
> PC might take it 3 instructions, while on a Mac
> it only takes one.
This is the sentence to which I object. The x86 instruction set is referred to as CISC, while the PPC instruction set (I don't know its proper formal name) is referrer to as RISC. In reality, the distinction is not so clear, since CISC and RISC are really just idealized philosophies. But to the general point, one of the basic tendencies of RISC-based instruction set are that the instructions are more basic. By this, I mean that there is often a leaning towards RISC having fewer instructions, with each instruction performing a more primitive operation. In a CISC-based instruction set, it is more likely to have a larger array of more complex instructions. Due to these philosophical differences, it is more likely for a RISC instruction set to require *more* instructions per action than a CISC instruction set. The benefit on the RISC side is that the simplified instructions sometimes, even in combination, can offer lower latencies and higher throughputs, though in modern processors of today, this advantage is not very apparent save for the current required x86 increase in decode stages, which seems to only be a significant factor in cases of branch misprediction (and, even then, the relative penalty is pretty small).
Anyway, current PPC processors do buck that particular RISC trend in some ways. The G4, for instance, has a MAC (or Multiply-ACcumulate) instruction which acts like a multiply then an add. Still, in general, in the real world, the latency of a processor is more a function of the microarchitecture than of the instruction set. And the instruction throughputs typically matter more for actual performance. In this area, I do not believe that the G4 family has a significant advantage in most cases (save for MAC and PERMUTE commands).
-JC
RISC processor performance fallacy
on
Sun's Zippy New Chips
·
· Score: 2, Interesting
> keep in mind that these are pure RISC processors
> and have always toasted any CISC or CISC-to-RISC
> processor of a much higher processor rating.
That is misleading and, in fact, bordering on the level of a total lie. The benefits of RISC architectures are not performance. They're simplicity. This simplicity, in the past, sometimes had the benefit of increasing performance, but higher performance is not a rule in and of itself.
Saying that "pure RISC processors... have always toasted any CISC or CISC-to-RISC" is a solid lie. There are plenty of occasions where a CISC processor outperforms a RISC processor.
In specfp_2000, the lowest frequency Pentium 4 scores a 516, while the highest frequency UltraSPARC III scores a 482. The slowest Pentium 4 is 7.1% faster than the fastest USPARCIII.
In specint_2000, the slowest P4 gets a 490, while the fastest Sun processor gets a 467. Here, the wimpiest current generation Intel processor is 4.9% faster than the best thing Sun can offer.
These above factors keep in mind that the Sun chips are *specifically* architected to achieve the highest performance possible, pretty much regardless of cost. They are full-on server chips. The Intel Pentium 4 series are designed with cost factors in mind. The Pentium 4 cannot be a three thousand dollar behemoth due to its target market (actually, the 750MHz USPARCIII processor module costs about $7k on Sun's website). So the USPARCIII can have the benefit of loads of added performance enhancing features while the Netburst (P4) architecture has to cut corners at every step.
the UltraSPARC III outperforms the Pentium 4 on a clock for clock basis. Of course, the original Pentium outperforms the Pentium 4 on a clock for clock basis on many benchmarks, too. This means nothing. It is merely reflective of different design strategies. I can easily point out the fact that the Pentium 4 offers higher performance per watt or higher performance per number of integer ALUs. But, like "performance per megahertz", those are also stupid measurements.
There is nothing out there which would cause me to believe that an x86 processor made with the design strategy of the UltraSPARC III ("we're gonna sell this for thousands of dollars, so throw in the kitchen sink, too!") would not outperform the UltraSPARC III at like frequency. Well, except if the the fp instructions on the USIII are three operand, but that's a special case. ^_^
> WWII. Nazis. Hitler's 'super men.' How bad would it have been had they had human cloning?
How would cloning at all affect this issue? First of all, Hitler wasn't in power for nearly long enough time to even bring a single generation of clones to maturity.
Secondly, cloning would probably have been thrown out in favour of selective breeding or dual parent genetic alteration. Frankly, Hitler would never have accepted any individual as an ideal Aryan, so producing a genetic duplicate would be seen as a useless tool in their quest for the ultimate human form.
-JC
http://www.jc-news.com/
Cloning is just another method of reproduction!
on
Send out the Clones?
·
· Score: 1
You're starting from the assumption that we want to make clones into duplicates of ourselves. But what about the prime drive of our entire species? What if somebody wants to continue their family line but has no spouse (or their spouse is infertile) and does not want to use some stranger's DNA?
What's wrong with having a child as a single parent? People forget that this area of science can have applications as a simple, straightforward means of having a normal child.
-JC
http://www.jc-news.com/
Re:The most interesting & missing parts of the sto
on
Chip News To Crunch On
·
· Score: 1
> All i see this as is a Plot to make my new mobo obsolete asap.
KT133 boards should be compatible with future Mustang-based Athlons. Only the Mustang server has been axed. The Mustang variant knows as "Palomino" will be called Athlon and the "Morgan" will be named Duron. They will work with current boards, though the Palomino's multiplier will usually be fixed for a 266/133MHz chipset connection, so you may have to change its multiplier to get it to work on your board.
BTW, the Hammer series won't really be out until 2002, and there will in fact be a 130nm version of the K7 released for you then, too.:)
> Check www.spec.org; in short, Intel just got kicked into touch. A new US III 900MHz
> workstation is getting over twice the FP performance of a roughly equivalent Intel chip.
Though the 900MHz UltraSPARC III an amazingly high performer in specfp2000 (482 peak, for those not up to date). But it is hardly twice that of current x86 chips! The top line AMD and Intel x86 chips are both above 330 peak. That makes the fastest UltraSPARC III 46% faster than the fastest x86 chips, not 100% faster.
However, the 900MHz USIII may not be available at the intro. It may only achieve server volumes around, say, a couple months from now. So, you really should be comparing it against x86 chips which will be available around the end of November. What will we have then?
Intel (barring delays) will have announced their 1.50GHz and 1.40GHz Pentium 4. Rumoured (pretty much confirmed, btw) specfp for the 1.50GHz Pentium 4 is 524, which is well above what the USIII is published at. Mind you, the P4-1.5 is not expected to achieve volume production for quite a while after intro, but this very high score means that even a lower frequency P4 should outperform the USIII-900 score.
AMD will likely have available for purchase by then a 1.30GHz (1.20GHz at the worst case, likely) Athlon, possibly based on their incrementally improved Mustang core, and probably outfitted with 2.133GB/s PC2100 Double Data Rate SDRAM. Additionally, they are slowly attaining optimized compiler support thanks to the efforts of DEC. I cannot say what the score will be, but I expect that it will be north of 400, even with current compilers.
Of course, all of this is moot. Sun doesn't need to be faster than x86. The two platforms do not compete in the same market segments. On the contrary, Sun has to compete against the Merced (Itanium). And given Intel's comments about the processor (they have openly stated that spec2000 is not a relevant benchmark for Itanium, which means that it likely performs like crap), I don't think Sun will have much of a problem competing against Intel in terms of specfp performance until the release of McKinley (IA-64 generation 2) in late 2001 or more likely 2002.
On the Coppermine (PIII 'E'), when the processor asks for a specific portion of memory, if it's in L1 then it takes 3 cycles to be retrieved. If it's in L2 instead, then it takes 7 cycles (or so?) to be retrieved.
That's basically the difference. They both 'tick' at the same clock rate, but one just happens to be able to deliver data in less than half the time.
This is why I'm always pissed at people who ignore every other factor when they refer to "full speed" cache. I mean... if you had L2 cache running at 800MHz on an 800MHz processor, but that L2 cache was only 8 bits wide and took eighty cycles to retrieve a piece of memory (eg, making it probably even slower than SDRAM), should you really refer to it as "full speed"?
PS: Apologies... in an earlier post, I referred to Willamette's bus as being 128-bit. I was very incorrect. The correct width is 64-bit (incidentally making Willamette's bus less cool than I thought).
Actually, it's simpler than even that. All Northwood is is the mobile version of the Pentium IV on a 130nm process. Saying that Intel is whipping this up in a panic is silly since this is a perfectly natural evolution in their roadmap.
We've known about their 130nm process, and we know that all their x86 chips get put into a mobile format. Also, we know that Willamette's core will be release a la the Pentium IV product very late this year.
FWIW, the Willamette is Intel's first new core since the P6 back in the mid 90s, which found itself inside the Pentium Pro, II, III, and Celeron. Rumours and admissions declare the below to be its likely improvements:
A deeper (or smoother, at least) instruction pipeline for stronger frequency ramping
Added execution pipelines and functional units (eg, allowing it to issue more instructions per cycle)
A "trace cache", to optimize the order in which instructions are fed into the pipelines (I admit, I could be screwing up this particular explanation)
variable frequency units -- this one's a leap of faith, but tentatively according to some sources, different parts of the cpu will be clocked at different rates. A 2GHz Willamette chip might have a 2GHz integer unit, a 1.7GHz fpu, and a 1.5GHz SSE unit (mind you, this is a speculative example, with numbers picked out of the air)
An improved motherboard bus, capable of 200MT/s (100MHz, double pumped) but also 128-bit, twice the width of the Athlon's EV-6, allowing for twice the peak bandwidth. Also, quadruple pumped (400MT/s) for the later server version, codenamed Foster. On the other hand, it will still be a shared bus, which is supposedly less "clean" than a point to point protocol that the Athlon uses, meaning that high-way SMP may get lots of collisions and degraded performance on Willamette.
The 130nm process basically will decrease the size of the features on the processor. Basically, imagine drawing stuff with a big fat marker, then getting a nice, fine pen. You can make much more detailed drawings, right? Basically same thing here. Benefits of going from 180nm to 130nm process:
Area of processor will be chopped in half, allowing for more than twice the amount of raw processor die to be fit onto one of those round fab wafers.
Because defects increase exponentally with die area, yields per die will improve
More on-die cache can be added with less risk of yield crash
processor can use lower voltage and much lower power dissipation
processor can be clocked much higher, generally a boost of 50-70% in the long run, but that's just my WAG.
You may be surprised to find out that most 500MHz and 550MHz Athlons are actually 600MHz or 650MHz chips clocked down to meet demand at the "low end". There are reasonably cheap devices (GFDs -- GoldFinger Devices) which allow you to modify the multiplier and voltage of the Athlon, so you could take it upon yourself to simply run your part at a higher clock.
I mean, granted, you should be careful and test it thoroughly, but it is notable that I know many people who've attempted this sort of "overclocking", and none of those whom I've met (who have tried it) have yet failed to run a Athlon-500 at 700MHz, and a substantial fraction of these people run theirs at 750MHz or higher.
Just a suggestion. If you're not too timid and you're willing to take every precaution, then you could avoid spending five hundred or so dollars on an upgrade.
-JC
PS: Since I don't post here often, I should put up a disclaimer or something: Overclock at your own risk -- burnouts happen about one out of every hundred thousand attempts for careful overclockers, so please don't kill or blame me if bad things happen as a result of an overclocking attempt (eg, I take no responsibility).
With everything else sucked out of memory save for the shell and systray, I'm doing a tad above 732,000 keys per second on my K6-2-300 @ 300MHz (using DCypher cli Win32 client).
Aargh, this is annoying. I mean, I heard rumours that the beta client was a little slowish, but I just benchmarked the d.net CSC client on my machine, and I got about 380.4kkeys/sec cracking rate (it's a K6-2-300).
It there a problem with some systems, software-wise, or is there a bug in the d.net implementation of CSC? I tried out the DCypher.net CSC client (it's been out for a week or so, I think) and for 732.6kkeys/sec on the same system, under the same circumstances!
Actually, I checked and I'm finding similar comparisons from various people I know, with d.net's CSC client being about half as fast in cracking compared to DCypher.net's (a friend of mine tried both on an Athlon-650 and got 2,023,437keys/sec for DCypher and 1,040,189.47keys/sec for d.net, for example).
My testing is being done with all programs except the shell and systray (and a dos box, as I'm using the command line clients) closed and out of memory.
Is this an optimization issue? Will d.net release improved clients in a few days? I'm really getting worried and annoyed. I had planned to do a DCypher-CSC/d.net-CSC comparison on my website to show which was faster on which of a variety of cpu cores. But this is insane!
Oh... ummm, I guess it's only fair that I link to DCypher, as they're kinda the underdog here and not as many people know about them.
I fear I must viciously attack you on several points, some merely semantical, others very deep in context:
(semantical) VIA isn't making modifications to the AMD750 chipset. They have their own chipset design that happens to work with the same processor. Saying that VIA's KX133 is a mod of the AMD750 is like saying the VIA MVP3 is a modification of the i430VX chipset, and we DON'T want to think that!
(serious) AMD is perfectly capable of handling most (or perhaps all) of the chipset production for Slot-A. They're doing it right now, and they could continue to do it if VIA totally gave up.
(more serious) AMD lets other companies make chipsets for their platforms *because* they believe it to be beneficial in the long run. Intel tries their darndest to prevent other companies from competing in their market, and they get burned because of it. 440NX, i810, i820, Profusion -- Intel constantly brings out new products that require instant recall for one reason or another, or that get delayed ad infinitum. It is only sheer luck that VIA happened to have a 133MHz chipset out to fit the PIII 600B -- Intel has been on and off suing them to not do that very thing! AMD, on the other hand, is allowing ALi, SiS, VIA, and whomever else to offer their own chipsets for the platform. This way, you'll get competition in this submarket. Competition can allow for companies to innovate their products at a faster rate, and they're motivated a bit more. On top of that, if VIA's production stops, AMD can make up for it. If AMD's production stops, SiS can make up for it. If SiS's productions stops... and so on. Allowing other chipset makers to produce for your platform isn't a bad idea, it's a PHENOMALLY GOOD idea.
(serious) AMD isn't moving their Fab to Dresden. They're making an additional fab in Dresden to complement the Austin fab. Merely "moving" their fabrication plant would be an amazingly boneheaded move on their part, since it would not benefit them at all and would incur an insane cost. By making an *extra* fab, the cost is still insane, but AMD is doubling their capacity.
(semantic) No Slot-A chipset supports PC133 memory, and no board officially supports PC133 memory, though some can be set to take it. Also, all boards currently support UMDA66, this isn't something that VIA would need to add.
(serious) The Motorola thing is not set in stone. There has been no agreement.
(serious) "shift the whole operation to Germany"? That suggestion, and the statement that follows, doesn't even begin to make an iota of sense. You're acting as if AMD's production is in Taiwan or something! AMD's cpus, chipsets, and most of their other stuff is made in the US. Specifically, their processors in Austin, TX, and their chipsets in (mostly, I think) Sunnyvale, CA. If Taiwan suddely were wiped off the face of the planet, AMD's ability to make processors and chipsets WOULD NOT BE AFFECTED AT ALL.
(semantic) AMD could not put out a 2GHz processor by Christmas of this year. At best, they could perhaps clear a little over 1GHz.
(serious) In no way would AMD be screwed if the KX133 never appeared. There's no palpable benefit with it. With their current chipset, future Athlons would still be faster (in some cases significantly) than competing P6 (the processor family which includes the Pentium III) offerings.
(beyond serious) Transmeta? You're joking! The Transmeta cpu will only have its details announced in November! At best, you could expect it to start popping out towards the end of 2000, and we haven't actually heard any talk about its performance or anything. Mind you, I'm hoping for the best for Transmeta, but there's no reason to yet believe that, say, it'll be faster than a Cyrix MII at x86 (which is, of course, not its native instruction set architecture).
(serious) VIA is going nowhere fast with Cyrix and Centaur. VIA "let go" almost half of Cyrix's team, and a lot of the rest left on their own. Morale is at an all time low, and at the point that VIA bought Cyrix, their new "flagship" product (Cayenne: Jedi/Gobi/Josh) was working at below 250MHz. AMD's Athlon only services the high end of the market, and VIA's processors only service the low end. Right now, VIA makes the bulk of their moolah on chipsets, and Slot-A offers an immense opportunity for profit, an opportunity that their processors couldn't offer until at least a year from now, and even then I'm skeptical. VIA isn't going to drop KX133, it would be an incredibly unwise business decision, and they know that; the chances of them doing this are less than one in an infinity, and only a gwennap would suggest otherwise.
-JC PC News'n'Links http://www.jc-news.com/pc
AMD's specific financial problems
on
K8 Details
·
· Score: 3
Just a quick comment on that, AMD doesn't have a problem with finance management, at least not any more than your typical company. Their problems largely rooted in the following two factors:
1) AMD's sixth generation processor design was put together decently, but with a very shallow pipeline. This means that with your typical ramping schema, it should be at about the same MHz level as the Cyrix chips (300MHz) or the WinChips (250MHz). As it is, AMD has an immensely aggressive ramping team which has managed to bring AMD's K6 family to just under Intel's P6 family in MHz, which has a couple effects: (a) Because the K6 family has been historically about two clock bins lower than the P6 family, and because Intel's pricing schema involves tremendous gulfs between the top two clock bins and all below it, AMD's cpu Average Selling Prices could not help but drop lower and lower as time progressed. (b) Due to the K6's low pipeline and the fab team's uncomparable (and absolutely necessary) aggressiveness, the bin split of the K6 family parts are HORRENDOUS. Before AMD's recent jump to their cs44e7 hybrid process (quarter micron with some 180nm features), the top bin being produced was 475MHz and the bottom bin was still way down at 333MHz or so, with over half the parts still binning below 400MHz. This added more shame to their ASPs, as anything below 400MHz was under a hundred bucks, which means something like only fifty dollars profit per chip, at best. (c) As a result of the aggressive ramping they needed (to compete with Intel's more easily rampable design), yields were kept lower than comparable Intel parts (though for the most part not horrendous, save for the little "incident" in February). This means that they get lower quantity to sell than they could have gotten otherwise, which means that, in addition to ASPs, they're making very low amounts of revenue.
2) There really is no way to get past problem 1a without making a newer cpu core with a deeper instruction pipeline. And to get past the problem in 1b, while that newer cpu core will help, it'd really be the wiser choice to expand your capacity, so AMD has forced themselves to spend a whopping, Intel-like amount of money (in R&D and in building a whole new megafab) so that, while they hurt in current quarters, they can thrive in future quarters. Would this strategy work? It's not guaranteed, but it's a hell of a lot cooler than the old "play it safe" mentality. If AMD had played it safe and not done all this fab or R&D stuff, then they'd have easily made profits (I believe) off the K6 series in every quarter of 1998 and 1999. The only problem is that they'd be lagging in clock speed at this point and they'd have no real future technology with which to compete. In effect, though they'd be profiting, they would be writing their own tombstone. The way they're doing it now, they've lost lots of money but they *finally* have superior technology to work with. Even without that newer fab, as soon as they ramp K7 to at least 60% capacity, they'd be making a pretty solid profit. With the newer fab, they'll be able to profit very nicely and retroactively fund these projects that they so unharmoniously dumped cash into all these years. They'd also be able to afford their future plans, which is a nice byproduct.
PS: This stuff is largely my opinion, though I believe it to be largely based on fact. It isn't merely a pipe dream that leads me to believe that the K7 is the first design since the 486 that offers everything AMD needs to absolutely thrive in the market.
Re:Its Specint95 is lower than the PA-8500 at 450M
on
K7 Benchmarking
·
· Score: 1
Is the 450MHz PA-8500 out? I know one of the cardinal rules of spec95 is never compare MHz-to-MHz, since one of the architectural tradeoffs you make for higher performance is necessarily lower clock sometimes.
Apologies...I have not kept up well with HP's offerings.:(
The K7 will be released in 500, 550, and 600MHz variants. This has been heavily hinted since November, and was confirmed by the CEO of AMD (Jerry Sanders) himself at an annual shareholders meeting (I think that's when it was).
The L2 cache of the K7 will be a half the clock of the processor. The 1/3x MHz idea was put together because AMD wasn't certain that the SRAM market would be able to supply 300MHz SRAMS for the K7-600's L2. Thankfully, this is not a problem.
Incidentally, Kryotech's Super-G will be out this year, likely at 1GHz in Q4, with a hypercooled K7. It *will* be expensive, but it will be *worth it*. AMD will have two 180nm processes ready by Q4, which will make the K7 a lot cheaper to make and a lot more voluminous (eg: there will be more of them). Figure that you might see an 800MHz K7 by end of year if AMD deems it necessary, that's one great core for MHz!
The K7 is going to be priced comparatively to the Pentium III, not the Pentium III Xeon, from what I've been told. The estimates among my local group are:
$400 or slightly above for 500MHz
$550-ish for 550MHz
$700 or so for the 600MHz version, though they may want a more respectable (eg: high) premium for the fastest x86 process of all time
These prices are slightly higher, mostly, than our extrapolations of PIII pricing around late July, where K7 will start to pick up volume. Despite the performance delta, AMD will likely make the part available to high end consumers in pricing, plus they want to pummel down Intel's high end ASP so they choke on their own Celerons.
AMD's DDR L2 "Viper" version of the K7, in Slot-B, will compete against Xeon. It will also happen to destroy Xeon in spec -- even more utterly than regular K7 does. Cascades looks like it'll be toasted a bit, too, unless Intel puts up a surprise and gives it 1MB L1 on-die.
BTW: K7's integer score beats out HP's mighty PA-8500 (which has 1.5MB L1 on die), I'm told. It may be the 2nd or third highest specint95 core out there.
Re:K7 facts in this report
on
K7 Info
·
· Score: 2
More of the facts than you probably think are true on that page. I tried to place question marks next to the questionable bits or otherwise mark them as rumour.
The bit about the PLL is 100% true, for example -- it's just that for some stuff like this, I cannot reveal my sources.
I can't *believe* that I forgot to mention anything regarding SMP. I mean, that was just *wrong* of me! I'll have to update the page in full force, though it might take a bit. I do recall that there is a company making very high SMP (for x86) chipsets for K7, but I keep thinking it's Profusion (the folks who are doing similar stuff on Intel's platform).
The company making for EV-6 could conceivably make specialized 8-way K7 a possibility for 1999, though it all depends on market demand, of course.
Oh...I think their name might have been Poseidon.
Anyway, that page is far from finished. It just covered a bunch of stuff that I think were important at the time, such as the concern over cost, exactly why the floating point unit is so fantastic, who might be supporting the platform. I even ignored MHz, since I'd basically assumed since November that the K7 would be coming out at 600MHz. I mean, wasn't it obvious?
Slashdot Mirror
> Celcius of[sic] Farenheit?
Well, it sure as hell isn't in Kelvins.
-JC
> Okay.. I read the article. It was an interesting mix between pat on the back science
c ur rent.htmle _Projects/p ioneer/PNStat.html/ rochedale/solar9.htm
) surface map of Pluto, thanks to Hubble (the satellite, not the man). The debate over whether Pluto is a planet is dependent on two things: (1) we don't really have a specific definition of the word "Planet", and (2) the common masses would get pissed off if Pluto got demoted, even if it was a rightful demotion, since they were taught as children that Pluto is a planet
:)
> and good old "Hey, aint NASA GREAT!" enthusiasm.
Well, at least that's a good counter to the constant "scientific research doesn't help people, so it's better to spend twenty times on the military than on space research" FUD that seethes through popular media.
> My question.. which I did not see answered, are where ARE they right now? I know they havent
> cleared the SS yet, but where exactly are they? ARe we going to get pictures Pluto and Neptune
> back?
http://www.vttoth.com/probes/probes.html
That's an okay list of current space probe locations, though it's not in any kind of detail.
None of the deep space probes are anywhere even remotely near Pluto or Neptune.
http://vraptor.jpl.nasa.gov/flteam/weekly-rpts/
http://spaceprojects.arc.nasa.gov/Spac
http://www.schools.ash.org.au
The above URLs should show you that both Voyagers as well as Pioneer 10 are at least ten billion kilometers away from the Sun, and they are leaving the solar system at greater than minimal escape velocity. Pluto, when furthest from Sol, is 7.4 billion kilometers away. So the probes are anywhere between 2.6 and 17.4 billion (or more) billion kilometers away from Pluto. Which means that we won't learn much from them about our little wanna-be planet.
> (Which would be GREAT.. and would solve that long running question of whether Pluto is even a
> planet, a bit asteroid, or a half a planet that got pulled into the gravity well here).
The long-running question is not science based, really. We already know the mass and diameter of Pluto. Heck, we even have a
(http://oposite.stsci.edu/pubinfo/PR/96/09.html
> Does it even have the transmitting power to send real data back anymore?
> or simply to weakly croak "I am here".
Well, Pioneer still performs maneuvering commands when requested. I don't know if its scientific resources are useable, but the Voyagers have nominal science instrument performance. I do recall that they are using these probes to determine where the point between solar wind and its stellar equivalent become equivalent in strength. I can't tell you much more, as I'm a bit busy at work today.
-JC
http://www.jc-news.com/
> How do I open videos from Xine's GUI?
> I already tried the help files.
I'm about thirty miles away from my part-time linux box, and I have crappy memory skills, but....
I believe that you click the "playlist" button, which is on the left side of the control window. A window pops up, and down the right side it should say things like "DVD" and "VCD" and, most importantly, "file". Click on "file", and yet another window appears. This window basically is a file opening dialog, except that it sucks. You can't actually type in filenames and paths. You have to navigate through by double-clicking or triple-clicking (sometimes the double-clicking simply refuses to work!) directory names and soforth. But it should eventually work.
Of course, that's the Mandrake 8.2 version. I don't know if the new version is refined from this.
-JC
First off, while the Xircom SpringPort for the Visor series saved me from insanity, I cannot quite recommend it. Here's a few reasons:
* I'm told that other 802.11b cards have the capability of autodetecting SSID numbers floating around, whereas the SpringPort must have its SSID explicitly defined.
* The power saving feature on the SpringPort is broken. This problem has caused problems with many people who assume that their part is simply defective. Basically, when power saving is enabled, the SpringPort experiences difficulties connecting to some base stations
* It's bloody expensive. I was lucky enough to get one on eBay for just over a hundred bucks, which is probably comparable to current 802.11b CompactFlash cards, but it originally retailed for well over three hundred dollars and it'd likely be hard to get one for under two hundred nowadays.
My strategy was to get a handheld with an expansion option that may be found in future handhelds down the line. I originally got the HandEra 330 because it had three expansion options findable elsewhere:
* Palm III compatible serial port
* Compact Flash Type I & II
* SD/MMC
I didn't expect much from the Palm III port (though there is a sled that lets you use PCMCIA in limited fashion for that port). But I had a hunch that either of the CF (which I had also used for storage on my Visor Prism via a MemPlug) or SD/MMC slots would find their way into other PDAs. I did turn out correct: You can find CF on the Sharp Zaurus, which is an interesting option, and the SD/MMC option is standardized on the Palm family now. Some of those Microsoft things also support those expansion types. Anyway, there isn't much out for SD/MMC in terms of wireless communication (save for Palm's Bluetooth card, but that's not the same thing really), but there are several CompactFlash-based 802.11b options.
Here's an idea: The TRG Pro, which is the precursor to the HandEra 330, is a less expensive option that still supports CompactFlash. I cannot be certain, but maybe it'll work with the CF 802.11b cards. It's basically a Palm VII minus the built-in internet crapola but with 8MB DRAM and the expansion port. Perhaps you could get one of those on the cheap and upgrade much, much later to a higher-lever device that still supports whatever CF card you get.
That's all I got for y'all. I wish you best of luck on finding the best option for your needs. ^_^
> I know that color can help the GUI a bit in
;)
> quicker human recognition and differentiation,
> but it sucks a lot of power.
A PDA like the Visor Prism or the Palm IIIc sucks power incredibly, but it's not because these PDAs have colour. It is because they use an active matrix screen. Colour PDAs like the Clie 710C and the Palm m505 have active matrix screens and therefore use much less power than that.
I should note, though, that if I were buying a PDA right now from scratch, I think my choice would be the HandEra 330.
-JC
http://www.jc-news.com/
> I remember the 5x86 chip that was actually an
> overclocked 486 clone. People who bought it
> were honestly tricked they bought a Pentium 133,
> when in fact it delivered performance worse than
> a Pentium 60.
Your comment is amazingly contradictory. AMD's 133MHz 5x86 was equal in frequency but lower in performance to a 133MHz Pentium. This chip had a P-rating of P75.
THINK ABOUT THIS!
You just said that AMD deceived customers by not P-rating a chip of equal MHz. And then you turned around and said that it's a deception to P-rate chips. That was a really silly comment that you made, especially since AMD was willing to market their chips as being lower performance per clock. They very openly marketed the 133MHz 5x86 as being performance equivalent to the Pentium 75. This was no secret, and it in fact was emblazoned on benchmark graphs all over the place.
P-rating, as it was and is used by AMD, is a means of comparing chips in a manner that is far more honest than MHz. AMD used it to show that they were lower performance per clock in the Pentium days, and now they are using it to show that they are higher performance per clock.
I truly don't understant why people are making such a big deal out of it.
-JC
> While it is true that most G4 to pentium
> bake-offs are done running photoshop filters, I
> don't think it is a particularly unfair test.
> After all, Photoshop really is the only standard
> application in existance that:
> a) has the same version and capabilities
> for both the PC and the Mac, and:
> b) can actually tax a current machine's
> processor.
> Other eligible apps (ie: Office) fail on
> both these counts.
I disagree with this. There are plenty of cross platform apps which can potentially be used for benchmarking both PPC and x86 computers. Here's a short list:
* Adobe Acrobat
* SETI@home
* d.net rc5-64
* POV-Ray
* Bryce
* Cinema4D
* Lightwave
* Mathematica
* Heuris MPEG Power Pro
* Deneba Canvas 7.0
* MetaCreations Canoma 1.0
* Adobe Illustrator 8.0
* ViaVoice
Additionally, if you just want to compare the hardware, you can install linux on both and run kernel compiles and likely other interesting benchmarks on that kernel.
But that's not even really the point. The problem that the author had was not that the comparison used Photoshop alone. It was that
a) the comparison alleged that the results were of "OVERALL PERFORMANCE"
b) the Photoshop test tested six hand-picked filters, possibly six of those that most prefer the G4 over the P4, instead of a wider range of filters and effects that might actually have more reasonably modeled actual Photoshop performance.
> There is hardly difference between using a 1.8
> Ghz Pentium 4 and a 500 Mhz Pentium 3 when
> surfing the web or typing a paper in Word.
That is subjective. My 500MHz is slow as crud, and I can see an amazing difference even just going up to 800MHz. This is likely because I am more aggressive with my computer use (even with just internet stuff), but it still allows me to object to your statement that a jump from 500MHz to 1.80GHz means little.
> How come I don't see many touting that the 1.2
> Ghz Athlon is some how lacking in ability when
> compared to the 1.8 Ghz Pentium 4?
That is because you haven't looked. There are very few reputable places that show that the 1.20GHz Athlon is superior to a 1.80GHz Pentium 4. In general, the idea is that the 1.33GHz Athlon is in the same range as the 1.70GHz Pentium 4, and the 1.40GHz Athlon is fairly level with the 1.80GHz Pentium 4. Of course, there are benchmarks in which said P4 is much faster, and there are benchmarks in which said Athlon is much faster. But you do not see AMD zealots simply running one of the rare benchmarks that have the 1.00GHz Athlon outperforming the 1.80GHz Pentium 4 and then claiming that these results reflect overal performance. That would be going over the line, and Scott Wasson's point in his rant was that the comparison between the G4 and various P4 boxes went over the line in this fashion.
> This author really doesn't know what he's talking about.
This message is a case of the pot calling the kettle black. In other words, you made this claim about the person, and you then followed up by making a statement that rather conclusively proves the same about yourself.
> The G4 processor is very powerful and the Mac architecture is well designed.
> As far as speeds go, Macs use a different
> instruction set technology, so an operation on a
> PC might take it 3 instructions, while on a Mac
> it only takes one.
This is the sentence to which I object. The x86 instruction set is referred to as CISC, while the PPC instruction set (I don't know its proper formal name) is referrer to as RISC. In reality, the distinction is not so clear, since CISC and RISC are really just idealized philosophies. But to the general point, one of the basic tendencies of RISC-based instruction set are that the instructions are more basic. By this, I mean that there is often a leaning towards RISC having fewer instructions, with each instruction performing a more primitive operation. In a CISC-based instruction set, it is more likely to have a larger array of more complex instructions. Due to these philosophical differences, it is more likely for a RISC instruction set to require *more* instructions per action than a CISC instruction set. The benefit on the RISC side is that the simplified instructions sometimes, even in combination, can offer lower latencies and higher throughputs, though in modern processors of today, this advantage is not very apparent save for the current required x86 increase in decode stages, which seems to only be a significant factor in cases of branch misprediction (and, even then, the relative penalty is pretty small).
Anyway, current PPC processors do buck that particular RISC trend in some ways. The G4, for instance, has a MAC (or Multiply-ACcumulate) instruction which acts like a multiply then an add. Still, in general, in the real world, the latency of a processor is more a function of the microarchitecture than of the instruction set. And the instruction throughputs typically matter more for actual performance. In this area, I do not believe that the G4 family has a significant advantage in most cases (save for MAC and PERMUTE commands).
-JC
> keep in mind that these are pure RISC processors
... have always toasted any CISC or CISC-to-RISC" is a solid lie. There are plenty of occasions where a CISC processor outperforms a RISC processor.
> and have always toasted any CISC or CISC-to-RISC
> processor of a much higher processor rating.
That is misleading and, in fact, bordering on the level of a total lie. The benefits of RISC architectures are not performance. They're simplicity. This simplicity, in the past, sometimes had the benefit of increasing performance, but higher performance is not a rule in and of itself.
Saying that "pure RISC processors
In specfp_2000, the lowest frequency Pentium 4 scores a 516, while the highest frequency UltraSPARC III scores a 482. The slowest Pentium 4 is 7.1% faster than the fastest USPARCIII.
In specint_2000, the slowest P4 gets a 490, while the fastest Sun processor gets a 467. Here, the wimpiest current generation Intel processor is 4.9% faster than the best thing Sun can offer.
These above factors keep in mind that the Sun chips are *specifically* architected to achieve the highest performance possible, pretty much regardless of cost. They are full-on server chips. The Intel Pentium 4 series are designed with cost factors in mind. The Pentium 4 cannot be a three thousand dollar behemoth due to its target market (actually, the 750MHz USPARCIII processor module costs about $7k on Sun's website). So the USPARCIII can have the benefit of loads of added performance enhancing features while the Netburst (P4) architecture has to cut corners at every step.
the UltraSPARC III outperforms the Pentium 4 on a clock for clock basis. Of course, the original Pentium outperforms the Pentium 4 on a clock for clock basis on many benchmarks, too. This means nothing. It is merely reflective of different design strategies. I can easily point out the fact that the Pentium 4 offers higher performance per watt or higher performance per number of integer ALUs. But, like "performance per megahertz", those are also stupid measurements.
There is nothing out there which would cause me to believe that an x86 processor made with the design strategy of the UltraSPARC III ("we're gonna sell this for thousands of dollars, so throw in the kitchen sink, too!") would not outperform the UltraSPARC III at like frequency. Well, except if the the fp instructions on the USIII are three operand, but that's a special case. ^_^
-JC
http://www.jc-news.com/
> WWII. Nazis. Hitler's 'super men.' How bad would it have been had they had human cloning?
How would cloning at all affect this issue? First of all, Hitler wasn't in power for nearly long enough time to even bring a single generation of clones to maturity.
Secondly, cloning would probably have been thrown out in favour of selective breeding or dual parent genetic alteration. Frankly, Hitler would never have accepted any individual as an ideal Aryan, so producing a genetic duplicate would be seen as a useless tool in their quest for the ultimate human form.
-JC
http://www.jc-news.com/
You're starting from the assumption that we want to make clones into duplicates of ourselves. But what about the prime drive of our entire species? What if somebody wants to continue their family line but has no spouse (or their spouse is infertile) and does not want to use some stranger's DNA?
What's wrong with having a child as a single parent? People forget that this area of science can have applications as a simple, straightforward means of having a normal child.
-JC
http://www.jc-news.com/
> All i see this as is a Plot to make my new mobo obsolete asap.
:)
KT133 boards should be compatible with future Mustang-based Athlons. Only the Mustang server has been axed. The Mustang variant knows as "Palomino" will be called Athlon and the "Morgan" will be named Duron. They will work with current boards, though the Palomino's multiplier will usually be fixed for a 266/133MHz chipset connection, so you may have to change its multiplier to get it to work on your board.
BTW, the Hammer series won't really be out until 2002, and there will in fact be a 130nm version of the K7 released for you then, too.
-JC
> Check www.spec.org; in short, Intel just got kicked into touch. A new US III 900MHz
> workstation is getting over twice the FP performance of a roughly equivalent Intel chip.
Though the 900MHz UltraSPARC III an amazingly high performer in specfp2000 (482 peak, for those not up to date). But it is hardly twice that of current x86 chips! The top line AMD and Intel x86 chips are both above 330 peak. That makes the fastest UltraSPARC III 46% faster than the fastest x86 chips, not 100% faster.
However, the 900MHz USIII may not be available at the intro. It may only achieve server volumes around, say, a couple months from now. So, you really should be comparing it against x86 chips which will be available around the end of November. What will we have then?
Intel (barring delays) will have announced their 1.50GHz and 1.40GHz Pentium 4. Rumoured (pretty much confirmed, btw) specfp for the 1.50GHz Pentium 4 is 524, which is well above what the USIII is published at. Mind you, the P4-1.5 is not expected to achieve volume production for quite a while after intro, but this very high score means that even a lower frequency P4 should outperform the USIII-900 score.
AMD will likely have available for purchase by then a 1.30GHz (1.20GHz at the worst case, likely) Athlon, possibly based on their incrementally improved Mustang core, and probably outfitted with 2.133GB/s PC2100 Double Data Rate SDRAM. Additionally, they are slowly attaining optimized compiler support thanks to the efforts of DEC. I cannot say what the score will be, but I expect that it will be north of 400, even with current compilers.
Of course, all of this is moot. Sun doesn't need to be faster than x86. The two platforms do not compete in the same market segments. On the contrary, Sun has to compete against the Merced (Itanium). And given Intel's comments about the processor (they have openly stated that spec2000 is not a relevant benchmark for Itanium, which means that it likely performs like crap), I don't think Sun will have much of a problem competing against Intel in terms of specfp performance until the release of McKinley (IA-64 generation 2) in late 2001 or more likely 2002.
-JC
On the Coppermine (PIII 'E'), when the processor asks for a specific portion of memory, if it's in L1 then it takes 3 cycles to be retrieved. If it's in L2 instead, then it takes 7 cycles (or so?) to be retrieved.
... if you had L2 cache running at 800MHz on an 800MHz processor, but that L2 cache was only 8 bits wide and took eighty cycles to retrieve a piece of memory (eg, making it probably even slower than SDRAM), should you really refer to it as "full speed"?
... in an earlier post, I referred to Willamette's bus as being 128-bit. I was very incorrect. The correct width is 64-bit (incidentally making Willamette's bus less cool than I thought).
That's basically the difference. They both 'tick' at the same clock rate, but one just happens to be able to deliver data in less than half the time.
This is why I'm always pissed at people who ignore every other factor when they refer to "full speed" cache. I mean
-JC
PC News'n'Links
http://www.jc-news.com/pc
PS: Apologies
We've known about their 130nm process, and we know that all their x86 chips get put into a mobile format. Also, we know that Willamette's core will be release a la the Pentium IV product very late this year.
FWIW, the Willamette is Intel's first new core since the P6 back in the mid 90s, which found itself inside the Pentium Pro, II, III, and Celeron. Rumours and admissions declare the below to be its likely improvements:
The 130nm process basically will decrease the size of the features on the processor. Basically, imagine drawing stuff with a big fat marker, then getting a nice, fine pen. You can make much more detailed drawings, right? Basically same thing here. Benefits of going from 180nm to 130nm process:
-JC
PC News'n'Links
http://www.jc-news.com/pc
You may be surprised to find out that most 500MHz and 550MHz Athlons are actually 600MHz or 650MHz chips clocked down to meet demand at the "low end". There are reasonably cheap devices (GFDs -- GoldFinger Devices) which allow you to modify the multiplier and voltage of the Athlon, so you could take it upon yourself to simply run your part at a higher clock.
I mean, granted, you should be careful and test it thoroughly, but it is notable that I know many people who've attempted this sort of "overclocking", and none of those whom I've met (who have tried it) have yet failed to run a Athlon-500 at 700MHz, and a substantial fraction of these people run theirs at 750MHz or higher.
Just a suggestion. If you're not too timid and you're willing to take every precaution, then you could avoid spending five hundred or so dollars on an upgrade.
-JC
PS: Since I don't post here often, I should put up a disclaimer or something: Overclock at your own risk -- burnouts happen about one out of every hundred thousand attempts for careful overclockers, so please don't kill or blame me if bad things happen as a result of an overclocking attempt (eg, I take no responsibility).
With everything else sucked out of memory save for the shell and systray, I'm doing a tad above 732,000 keys per second on my K6-2-300 @ 300MHz (using DCypher cli Win32 client).
-JC
Aargh, this is annoying. I mean, I heard rumours that the beta client was a little slowish, but I just benchmarked the d.net CSC client on my machine, and I got about 380.4kkeys/sec cracking rate (it's a K6-2-300).
... ummm, I guess it's only fair that I link to DCypher, as they're kinda the underdog here and not as many people know about them.
It there a problem with some systems, software-wise, or is there a bug in the d.net implementation of CSC? I tried out the DCypher.net CSC client (it's been out for a week or so, I think) and for 732.6kkeys/sec on the same system, under the same circumstances!
Actually, I checked and I'm finding similar comparisons from various people I know, with d.net's CSC client being about half as fast in cracking compared to DCypher.net's (a friend of mine tried both on an Athlon-650 and got 2,023,437keys/sec for DCypher and 1,040,189.47keys/sec for d.net, for example).
My testing is being done with all programs except the shell and systray (and a dos box, as I'm using the command line clients) closed and out of memory.
Is this an optimization issue? Will d.net release improved clients in a few days? I'm really getting worried and annoyed. I had planned to do a DCypher-CSC/d.net-CSC comparison on my website to show which was faster on which of a variety of cpu cores. But this is insane!
Oh
-JC
PC News'n'Links
http://www.jc-news.com/pc
I fear I must viciously attack you on several points, some merely semantical, others very deep in context:
... and so on. Allowing other chipset makers to produce for your platform isn't a bad idea, it's a PHENOMALLY GOOD idea.
(semantical) VIA isn't making modifications to the AMD750 chipset. They have their own chipset design that happens to work with the same processor. Saying that VIA's KX133 is a mod of the AMD750 is like saying the VIA MVP3 is a modification of the i430VX chipset, and we DON'T want to think that!
(serious) AMD is perfectly capable of handling most (or perhaps all) of the chipset production for Slot-A. They're doing it right now, and they could continue to do it if VIA totally gave up.
(more serious) AMD lets other companies make chipsets for their platforms *because* they believe it to be beneficial in the long run. Intel tries their darndest to prevent other companies from competing in their market, and they get burned because of it. 440NX, i810, i820, Profusion -- Intel constantly brings out new products that require instant recall for one reason or another, or that get delayed ad infinitum. It is only sheer luck that VIA happened to have a 133MHz chipset out to fit the PIII 600B -- Intel has been on and off suing them to not do that very thing! AMD, on the other hand, is allowing ALi, SiS, VIA, and whomever else to offer their own chipsets for the platform. This way, you'll get competition in this submarket. Competition can allow for companies to innovate their products at a faster rate, and they're motivated a bit more. On top of that, if VIA's production stops, AMD can make up for it. If AMD's production stops, SiS can make up for it. If SiS's productions stops
(serious) AMD isn't moving their Fab to Dresden. They're making an additional fab in Dresden to complement the Austin fab. Merely "moving" their fabrication plant would be an amazingly boneheaded move on their part, since it would not benefit them at all and would incur an insane cost. By making an *extra* fab, the cost is still insane, but AMD is doubling their capacity.
(semantic) No Slot-A chipset supports PC133 memory, and no board officially supports PC133 memory, though some can be set to take it. Also, all boards currently support UMDA66, this isn't something that VIA would need to add.
(serious) The Motorola thing is not set in stone. There has been no agreement.
(serious) "shift the whole operation to Germany"? That suggestion, and the statement that follows, doesn't even begin to make an iota of sense. You're acting as if AMD's production is in Taiwan or something! AMD's cpus, chipsets, and most of their other stuff is made in the US. Specifically, their processors in Austin, TX, and their chipsets in (mostly, I think) Sunnyvale, CA. If Taiwan suddely were wiped off the face of the planet, AMD's ability to make processors and chipsets WOULD NOT BE AFFECTED AT ALL.
(semantic) AMD could not put out a 2GHz processor by Christmas of this year. At best, they could perhaps clear a little over 1GHz.
(serious) In no way would AMD be screwed if the KX133 never appeared. There's no palpable benefit with it. With their current chipset, future Athlons would still be faster (in some cases significantly) than competing P6 (the processor family which includes the Pentium III) offerings.
(beyond serious) Transmeta? You're joking! The Transmeta cpu will only have its details announced in November! At best, you could expect it to start popping out towards the end of 2000, and we haven't actually heard any talk about its performance or anything. Mind you, I'm hoping for the best for Transmeta, but there's no reason to yet believe that, say, it'll be faster than a Cyrix MII at x86 (which is, of course, not its native instruction set architecture).
(serious) VIA is going nowhere fast with Cyrix and Centaur. VIA "let go" almost half of Cyrix's team, and a lot of the rest left on their own. Morale is at an all time low, and at the point that VIA bought Cyrix, their new "flagship" product (Cayenne: Jedi/Gobi/Josh) was working at below 250MHz. AMD's Athlon only services the high end of the market, and VIA's processors only service the low end. Right now, VIA makes the bulk of their moolah on chipsets, and Slot-A offers an immense opportunity for profit, an opportunity that their processors couldn't offer until at least a year from now, and even then I'm skeptical. VIA isn't going to drop KX133, it would be an incredibly unwise business decision, and they know that; the chances of them doing this are less than one in an infinity, and only a gwennap would suggest otherwise.
-JC
PC News'n'Links
http://www.jc-news.com/pc
Just a quick comment on that, AMD doesn't have a problem with finance management, at least not any more than your typical company. Their problems largely rooted in the following two factors:
1) AMD's sixth generation processor design was put together decently, but with a very shallow pipeline. This means that with your typical ramping schema, it should be at about the same MHz level as the Cyrix chips (300MHz) or the WinChips (250MHz). As it is, AMD has an immensely aggressive ramping team which has managed to bring AMD's K6 family to just under Intel's P6 family in MHz, which has a couple effects:
(a) Because the K6 family has been historically about two clock bins lower than the P6 family, and because Intel's pricing schema involves tremendous gulfs between the top two clock bins and all below it, AMD's cpu Average Selling Prices could not help but drop lower and lower as time progressed.
(b) Due to the K6's low pipeline and the fab team's uncomparable (and absolutely necessary) aggressiveness, the bin split of the K6 family parts are HORRENDOUS. Before AMD's recent jump to their cs44e7 hybrid process (quarter micron with some 180nm features), the top bin being produced was 475MHz and the bottom bin was still way down at 333MHz or so, with over half the parts still binning below 400MHz. This added more shame to their ASPs, as anything below 400MHz was under a hundred bucks, which means something like only fifty dollars profit per chip, at best.
(c) As a result of the aggressive ramping they needed (to compete with Intel's more easily rampable design), yields were kept lower than comparable Intel parts (though for the most part not horrendous, save for the little "incident" in February). This means that they get lower quantity to sell than they could have gotten otherwise, which means that, in addition to ASPs, they're making very low amounts of revenue.
2) There really is no way to get past problem 1a without making a newer cpu core with a deeper instruction pipeline. And to get past the problem in 1b, while that newer cpu core will help, it'd really be the wiser choice to expand your capacity, so AMD has forced themselves to spend a whopping, Intel-like amount of money (in R&D and in building a whole new megafab) so that, while they hurt in current quarters, they can thrive in future quarters. Would this strategy work? It's not guaranteed, but it's a hell of a lot cooler than the old "play it safe" mentality. If AMD had played it safe and not done all this fab or R&D stuff, then they'd have easily made profits (I believe) off the K6 series in every quarter of 1998 and 1999. The only problem is that they'd be lagging in clock speed at this point and they'd have no real future technology with which to compete. In effect, though they'd be profiting, they would be writing their own tombstone. The way they're doing it now, they've lost lots of money but they *finally* have superior technology to work with. Even without that newer fab, as soon as they ramp K7 to at least 60% capacity, they'd be making a pretty solid profit. With the newer fab, they'll be able to profit very nicely and retroactively fund these projects that they so unharmoniously dumped cash into all these years. They'd also be able to afford their future plans, which is a nice byproduct.
-JC
PC News'n'Links
http://www.jc-news.com/pc
PS: This stuff is largely my opinion, though I believe it to be largely based on fact. It isn't merely a pipe dream that leads me to believe that the K7 is the first design since the 486 that offers everything AMD needs to absolutely thrive in the market.
Is the 450MHz PA-8500 out? I know one of the cardinal rules of spec95 is never compare MHz-to-MHz, since one of the architectural tradeoffs you make for higher performance is necessarily lower clock sometimes.
:(
Apologies...I have not kept up well with HP's offerings.
-JC
"PS: K7 and mP6 look to be the fastest current cores for rc5, per MHz."
;)
Heh...I mean x86 cores, of course.
-JC
The K7 will be released in 500, 550, and 600MHz variants. This has been heavily hinted since November, and was confirmed by the CEO of AMD (Jerry Sanders) himself at an annual shareholders meeting (I think that's when it was).
The L2 cache of the K7 will be a half the clock of the processor. The 1/3x MHz idea was put together because AMD wasn't certain that the SRAM market would be able to supply 300MHz SRAMS for the K7-600's L2. Thankfully, this is not a problem.
Incidentally, Kryotech's Super-G will be out this year, likely at 1GHz in Q4, with a hypercooled K7. It *will* be expensive, but it will be *worth it*. AMD will have two 180nm processes ready by Q4, which will make the K7 a lot cheaper to make and a lot more voluminous (eg: there will be more of them). Figure that you might see an 800MHz K7 by end of year if AMD deems it necessary, that's one great core for MHz!
-JC
PC News'n'Links
PS: K7 and mP6 look to be the fastest current cores for rc5, per MHz. They may both be faster than the mighty K5, once optimized for.
The K7 is going to be priced comparatively to the Pentium III, not the Pentium III Xeon, from what I've been told. The estimates among my local group are:
$400 or slightly above for 500MHz
$550-ish for 550MHz
$700 or so for the 600MHz version, though they may want a more respectable (eg: high) premium for the fastest x86 process of all time
These prices are slightly higher, mostly, than our extrapolations of PIII pricing around late July, where K7 will start to pick up volume. Despite the performance delta, AMD will likely make the part available to high end consumers in pricing, plus they want to pummel down Intel's high end ASP so they choke on their own Celerons.
AMD's DDR L2 "Viper" version of the K7, in Slot-B, will compete against Xeon. It will also happen to destroy Xeon in spec -- even more utterly than regular K7 does. Cascades looks like it'll be toasted a bit, too, unless Intel puts up a surprise and gives it 1MB L1 on-die.
BTW: K7's integer score beats out HP's mighty PA-8500 (which has 1.5MB L1 on die), I'm told. It may be the 2nd or third highest specint95 core out there.
Also, K7 kicks ass at rc5 -- pass it along!
-JC
PC News'n'Links
More of the facts than you probably think are true on that page. I tried to place question marks next to the questionable bits or otherwise mark them as rumour.
The bit about the PLL is 100% true, for example -- it's just that for some stuff like this, I cannot reveal my sources.
I can't *believe* that I forgot to mention anything regarding SMP. I mean, that was just *wrong* of me! I'll have to update the page in full force, though it might take a bit. I do recall that there is a company making very high SMP (for x86) chipsets for K7, but I keep thinking it's Profusion (the folks who are doing similar stuff on Intel's platform).
The company making for EV-6 could conceivably make specialized 8-way K7 a possibility for 1999, though it all depends on market demand, of course.
Oh...I think their name might have been Poseidon.
Anyway, that page is far from finished. It just covered a bunch of stuff that I think were important at the time, such as the concern over cost, exactly why the floating point unit is so fantastic, who might be supporting the platform. I even ignored MHz, since I'd basically assumed since November that the K7 would be coming out at 600MHz. I mean, wasn't it obvious?
-JC
PC News'n'Links
http://www.jc-news.com/pc