Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×

AMD Interview 25

Brandon Bell writes "As a follow-up to our K7 Preview we've just conducted an interview with AMD. Topics discussed include the K7, 3DNow!, overclocking, and AMD's future plans. " Its mostly marketroid stuff, but there are a few bits worth reading.
This discussion has been archived. No new comments can be posted.

AMD Interview

Comments Filter:
  • I'm running a AMD K-6.2 350 at 100mhz bus, I bought it the week before christmas and I love. It was dirt cheap at $138. I just wish because it was so cheap I could have dual processers. Does any one know if the K-7 will be able to do SMP? Also any Ideas on the starting price. I would guess around $300 but I'm probably totally wrong.
  • very true. The thru-put on this motherboard would be greatly improved resulting in a much more effiecient bus. If this is the way they implement it. That would be uber c00l.

    Installing memory in pairs would let you interleave it at 128 bits, but it would be an super-engineering accomplishment to get it running at 200MHz. We call a dual PII system a "Dual 450MHz PII computer" not a "900MHz PII system". 100MHz 128bit bandwith would be enough for a system like this...
    It's far easier to forgive your enemy after you get even with him.
  • Looks like the FSB is going to actually be around 100MHz vs. 200MHz. The 200MHz bus people talk about is to chipset and other processors. This will certainly improve the efficiency of dual processor systems, but the motherboard still connects to the RAM at ~100MHz and the PCI bus at 33 or 66MHz. One thing i am wondering about is overclocking the motherboard bus speed. If you OC the memory bus to 133, this will also affect the PCI bus, OCing it to either 44.4MHz with a 3X multiplier or, if you can adjust the multiplier (many motherboards cannot), perhaps 33.3MHz with a 4X multiplier. This has been one major problem with OCing the bus, and inturn, the PCI bus. It messes with PCI cards and can cause instability.
    I would assume that the 200MHz bus to the CPU would be affected also by the main bus overclock. This would result in complications and most likely an unstable system.

    But, I'm sure Abit will work around this. ;)
    BTW: the G4 bus will perform a similar trick by doubling the bus width between processors and memory instead of MHz speed. Its called the MaxBus and will transfer data at 128 bits instead of 64. I'm not sure SDRAM will be used, though.
    It's far easier to forgive your enemy after you get even with him.
  • Unless you're foolish enough to be using a proprietary OS, in which case you had no guarantees of privacy to start with. IMO, so-called "privacy advocates" who complain about the chip id simply lack any clue.

    That said, my first AMD was also a 386/40, and I've never looked back. Is Intel still in business? :-)
  • (Sorry about the mangled subject line if there is one. Blame NSN4.5 for S/Linux)

    K6-2 chips compete pretty favorably with the PII in price, as AMD chips have always tended to be cheaper than their Intel counterparts. However, I'm not convinced that the K6-2 is a better buy than the Celeron, particularly and especially the slower Celerons than can be overclocked. I'm very happy with my 300A oc'd to 450. Since I needed a new mobo *anyway*, whether I went to Cyrix, AMD, or Intel for the new chip, the Celeron was a great value.

    This isn't intended to bash AMD at all - I've owned systems with CPUs made by AMD *and* Cyrix in the past (and one other Intel) - with no complaints except for the Cyrix running hot on occasion. It's important that AMD and Cyrix are around - else I wouldn't have been able to afford that cheap Celeron!

    Maybe by the time the K7 is out and affordable I'll be ready to upgrade again.
  • From what's been going around, this is almost correct. The 200 MHz bus is useful though in the following way.

    Even if you talk to memory at 100 MHz (due to memory manufacturer limitations), you still have 1/2 your bus free to talk to the rest of the system, such as the PCI/AGP cards. So while a PIII or similar processor can only burst to memory, a K7 in theory should be able to transfer data across AGP at 66 MHz while talking to memory at 100 MHz.

    None of this has been verified, but its the assumption being made.

    There's also the possibility that you'll have to install memory in pairs and then it'll interleave them to get 200-MHz memory performance. This seems reasonable, but only time will tell.
  • According to this article [] a company called Poseidon Technology is working on an up to 8-way K7 chipset. I submitted this article to Slashdot a while back, but it never got posted. Oh well.
  • I have a friend that works at AMD and according to him it will be SMP and have a floating point unit comparable to Intels(about damn time!). Of course, you still need a motherboard that supports multiple K-7's. But as soon as that comes out I'm snagging it!
  • Right now OEM's are sellng the K6 systems to provide cheaper systems. Sad thing is, these systems are so underpowered, the chips power is never seen. Gateway for example, has all AMD systems shipping on a mini tower setup, with a 90w power supply, and an integrated motherboard with no ISA, 3 PCI (one usually taken by a modem), and no AGP. I asked the AMD representitive when OEM's will take AMD seriosuly, and he hoped when the K7 comes out, as the K6 is still only seen as a cheap chip, not something that peforms well.

    One thing that should help multiple K7's be more powerful then multiple Xeons is that all will be on their own bus, unlike the shared bus method Intel uses.

    And for the home market, the K7 will have 3dnow!, and it has already been supported by many games out there. Also is more efficient then Intels KNI. (Or SSE, or whatever you want to call it).
  • I bought a system from CyberMax recently (K6-450 Entrepreneur), and it's got AGP, a 32 Mb video card, 128 Mb RAM, etc. I'm very happy with it. It all cost about $1600. (inc. 17" monitor)
  • was abought 10 years ago. It was a 386-40 and really made my FIRST x86 computer pretty snappy for the time. The FIRST time I heard about the K5 and K6, I wanted to be the FIRST on the block to get one. However I knew a guy who didnt bother to put a CPU fan on FIRST before turning the power on. He had a little puddle of molten CPU in the bottom of the case. It was his FIRST computer and it really bummed him out. Shortly thereafter he bought his FIRST mac.

    Now I've been using Intel chips for years now, but when I FIRST heard about the K7 I really wanted that chip. Now that Intel has decided to become the FIRST manufacturer to put nasty chipID's into the x86 architecture I have no choice but to buy AMD. I care about privacy, I care about the FIRST ammendment. This isnt the FIRST time a company has been against the cause of freedom. Yet I know this wont be the FIRST boycott I participate in because business overstep thier bounds.

    Back to the K7. I want to be the FIRST in line at the store on the release date. Now, am I correct in assuming that this is the FIRST chip targeted to the consumer market using the DEC bus protocol? I've thought about buying an Alpha, but I want to see how the K7 does FIRST. Oh well, the article was a bit doubleplusmarketingspeak for my taste, but I read it anyway. I just want to be FIRST among my friends with the K7 info.
  • I wish they (the cpu makers) would break from away from this semi 64-bit crap and go a full 64-bit, or 128-bit (as seen in the new risc chips).
  • by nd ( 20186 )
    AMD has always been known for having very low prices compared to the Intel counterpart. But, with the big jump with the improved floating point, 200mhz bus, etc, won't this force them to submit to Intel's price range?
  • The bus between the processor and the main chipset will run at 200MHz (or thereabouts). The speed of the bus between the main chipset and the memory, PCI slots, etc. can/will be independent from the speed of the bus to the processor.
  • If AMD produces a chip which outperforms Intel's high end chip you can expect that AMD will charge MORE than the comprable intel chip.

    However, this will likely drive both AMD and Intel's price points down! Good news for us!
  • Question

    How exactly did you overclock your processor?

    Just curious

  • To the person who was talking bout the K-7, and asking if it was able to SMP. Out of all the reports I have read it looks like this will be the chip that AMD can SMP with, also I think they are putting it in the price range of the PIII :( , there goes my dreams down the hatch. Also Check out this weeks (4/5/99) they are doing a great write up on the K-7 and how it will compete with Intel. For prices on stuff I go to they seem to be the cheapest.
  • Check out an ongoing story over at AMD Central [] about the K7. According to the technical materials I have read, the K7 will not only do SMP, but will be more efficient than comparable Intel systems, which in the PIII peak at four-way processing. This limitation is caused by a shared system bus @100MHz. In other words, all four processors in a PIII four-way motherboard share the same 100 MHz of bandwidth to the system chipset.

    In a K7 SMP system, each CPU has its own separate 200MHz bus to the system chipset. Also, this bus is separate from the memory bus running at 100MHz+, so there will not necessarily be a big bottleneck at the memory bus. The separate bus space for each CPU will allow more efficient SMP and theoretically higher clock speeds for the chips.

    AMD, as pundits at The Register [] have pointed out, still has a huge manufacturing disadvantage. Suggestions have been increasing their deal with IBM to manufacture the K7 chip, borrowing another USD 2 billion for another fab, or looking for a suitor to save them. Compaq might have been a good choice until Pfeifer's ouster.

    The bottom line is that the K7 will be a great chip, but clock speeds will not crank up much past 500MHz until the Dresden fab gets going with the .18 micron process by the end of the year.

    Can AMD survive? I hope so, I've got some stock I'd hate to take a bath on, and they have good technology.

    Right now, I've got a Celeron 300A overclocked at 450MHz, which of course Intel has quickly rid the market of because it was eating up their PII profits. I want a K7, but it will be priced to compete with the PIII, to finally give AMD the margins it needs to turn a profit.

    Intel talks tough about competing with AMD, and they certainly have the money to back up tought talk, but I have heard talk from inside Intel that the 64 bit chip design is a disaster and is draining resources from other projects. Just a rumour ... ?

  • Having spent 20 of my 25 years as a student, I am ever grateful to AMD for pricing chips in the affordable range. My first x86 was a Tandy 1000, (which kicked my TI99/4A's ass) followed by a brief stint with a 286.

    Since my AMD 386, I have always gone with AMD, and have had no complaints whatsoever. I went to my local computer store today, and picked up not a K7, but a K6-2 333, for the outrageous price of 35 bucks. It's installed, it rocks, the only thing better than a new AMD chip, is the price of the one thats not new anymore!

    Intel chips are nice, but the xtra cost of the Intel outweighs any performance and reliability issues. I trust AMD, but hey, it's a free country.

    btw, I was the FIRST one to leave the office today!


"With molasses you catch flies, with vinegar you catch nobody." -- Baltimore City Councilman Dominic DiPietro