Intel Makes Next Move

[Stick_e_Mouse]

As we all know, VIA had some huge balls making the first move when they decided to release their P4 P4X266 and P4M266 chip without an Intel license. Now Intel made the next move by slapping a lawsuit right in VIA's face! Wheather this is good or not, is totally up for debate.
In my opinion, however, it kind of sucks for us consumers since we will still be left with the P4 and Rambus combo for now (the whole SD-RAM idea is stupid). This results in higher prices and not-so-steller performance. On the other hand, it is a win for Intel and other motherboard manufacturers that DO have Intel licenses since they won't feel like they have been cheated into paying for licenses while VIA does not.

<A HREF="http://www.ebns.com/digest/story/OEG20010907S0084" target="_new">http://www.ebns.com/digest/story/OEG20010907S0084</A>
<A HREF="http://www.pcworld.com/news/article/0,aid,61184,tk,dn090701X,00.asp" target="_new">http://www.pcworld.com/news/article/0,aid,61184,tk,dn090701X,00.asp</A>

See a real naked pic of Britney Spears <A HREF="http://www.geocities.com/stick_e_mouse" target="_new">here</A>!!!

 

[flamethrower205]

I think Intel was being stupid there. I mean if a product is going to sell even more, why not? Are they not in teh business of making $ or something?

When I rule the world, Apple will only mean the fruit.

 

[Stick_e_Mouse]

Like Tom said, those greedy SOBs want ALL of the pie, not just a piece of it.

See a real naked pic of Britney Spears <A HREF="http://www.geocities.com/stick_e_mouse" target="_new">here</A>!!!

 

[rcf84]

Well it looks like Intel want SiS, ALi, ATi to have the DDR for the P4. VIA help amd get a alot of market space with its kx133 and kt133. Then AMD dumped them for nvidia and VIA is ready to launch a chipset for the P4. I think its one good way to say out of luck via.

Nice Nvidia and ATi users get a Cookie.... Yummy

 

[Raystonn]

DDR SDRAM chipsets will be available Q1 of 2002. Unlike VIA, some chipset makers like to debug their products. This requires time.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[peteb]

but will the "official" P4 DDR/SD-RAM chipsets support PC2100 and above or stick with the i845 PC1600 standards? For that matter is the Brookdale chipset officially only supporting PC166 (I think I read somewhere) or will it or future Intel/licensed chipsets support PC2100/2400 and above?

-* This Space For Rent *-
email for application details

 

[peteb]

Well, I cannot blame Intel for protecting it's patents and licensing - it is always wrong for any company to infringe upon another's intellectual property - but I am suspicious that Intel did not grant license to develop VIA solutions and a workable DDR technology that may have closed the gap/neared/bettered RD-RAM solutions.

Sure as Raystonn says, maybe Intel wanted to hold VIA out becuase it tends to throw 1/2 baked solutions out the door and let us debug them, but that never held Intel back in the past. It all smacks of protecting RAMBUS if you ask me. Like I said, I do not fault Intel for slapping a lawsuit on VIA if they have infringed patents but I do feel sorry for VIA and us that we are still no closer to fast cheap P4 solutions again.

-* This Space For Rent *-
email for application details

 

[Raystonn]

Currently the Pentium 4 supports only the 100MHz external clock that is quad-pumped to a 400MHz system bus. When the SDR SDRAM chipset is released we will also have Pentium 4 CPUs that are designed for a 133MHz external clock and system bus. It follows that later when DDR SDRAM chipsets are released we will still have Pentium 4 CPUs that are designed for 100MHz and 133MHz external clock speeds. PC1600 uses a 100MHz external clock. PC2100 uses a 133MHz external clock. Both forms of DDR SDRAM should be usable.

There is no way to 'lock' the external clock from within the CPU (hence it being external.) We should have two different types of Pentium 4 CPUs that are designed to run at the advertised clockspeed with either a 100MHz external clock or a 133MHz external clock. There is no way to keep someone from taking a chipset that was designed to use PC1600 DDR SDRAM with a 100MHz external clock, plugging in one of the 133MHz external clock Pentium 4 CPUs, increasing that external clock to 133MHz, and running the memory at PC2100 speeds. (I would recommend you actually have PC2100 memory instead of trying to overclock the memory.) Thus, even if one is not released, you can still do it.

You could even stay with the 100MHz external clock Pentium 4 CPUs and overclock it with a 133MHz external clock and some PC2100 memory. If it runs stable you just got a 33% increase in clockspeed and memory bandwidth at the same time. Can you imagine running a 2GHz Pentium 4 at 2.66GHz? Anything is possible with proper cooling.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Raystonn]

"For some reason I still feel the hollowness power from P4!"

Huh?


"if Dual Channel with DDR ram is use, then wouldn't take the full advantage of P4?"

There really isn't enough room on today's motherboards for dual SDRAM (SDR or DDR) channels all the way to the CPU. Even on nForce you only get a single channel's worth from the chipset to the CPU. A dual channel SDRAM (SDR or DDR) implementation would require 128 pins/traces all the way to the CPU. The reason why dual channel RDRAM was easy is it only requires 32 pins/traces.

However, if you increase the number of layers (and the cost) of your motherboard you could do a dual channel SDRAM (SDR or DDR) implementation given enough time. It's not planned though for any time soon, if at all. In 2H 2002, RDRAM moves up to PC1066. This offers 4.267 GB/s of memory bandwidth in a dual channel RDRAM chipset. In a little over a year (2003) a new form of RDRAM will be out that uses 32 pins/traces per module rather than 16. This offers 8.533 GB/s of memory bandwidth. I don't foresee SDRAM (SDR or DDR) catching up. In fact, because it's so much more difficult to increase the frequency than it is to add more pins, I foresee it falling so far behind that it is eventually scrapped.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Raystonn]

"Intel is not going to stick with Rambus forever are they?"

Yes. It will become the de facto standard.


"P4 gotta have DDR and VIA is their first chance"

There really is no need for it besides having something at a lower price point with better performance than SDR SDRAM. As far as VIA being the 'first chance', they will soon regret that move very much. They do not have a license to produce that chipset. Their mothers should have taught them better. When you want to play with someone else's toys, you ask them first.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[rcf84]

I say this "!!SiS645!!". DDR333 !!! 2.7gb of bandwith maybe enough to pass rdram on bench's. Well Also ALi's chipset is going to have DDR333.

Nice Nvidia and ATi users get a Cookie.... Yummy

 

[peteb]

Hmm - didn't know you guys were going to release an official 133Mhz P4 - I thought that was the stink over Brookdale support (at least as I recall) that it was only going to be certified up to PC1600 memory speeds. A 133Mhz bus P4 would make a lot more sense and be able to utilise the bus better. Thinking about it I recall you mentioning some time back about the PC1066, which I presume is also based on 133Mhz quad pumped (133.3Mhz incase anyone else reading thinks my maths is off) and so required for later memory/chipset iterations.

Well, I just hope we see prices come down some - and I hope AMD continue to give you guys a run for your money - if nothing else it will force you both to get your toys out to market a little earlier so we get to play with them. I'm sure that you get to play with some toys that would make the average PC enthusiast spin his propellor though...

-* This Space For Rent *-
email for application details

 

[killall]

why the hell does a company need a licence to produce something which doesnt steal any technology but just can use the same processor... im sorry you cant make paper which you can write on with that pen...

if in doubt blame microsoft...

 

[Arbee]

Because for the chipset to be able to 'talk' to the P4 cpu it must use the same bus protocol (*I think* it is the GLT+ - (update:nope, it isn't ) that is intel patented.


How terrible is wisdom when it brings no profit to the wise<P ID="edit"><FONT SIZE=-1><EM>Edited by Arbee on 09/08/01 12:13 PM.</EM></FONT></P>

 

[Crashman]

No. Intel and VIA have a history, with VIA having cheated Intel on the P3 bus (for the Cyrix 3), Intel decided not to licence the P4 bus to them. Good decision, Intel is not involving themsleves with such a shady company. I only wish Intel had WAITED unitl AFTER VIA chipsets were in 10,000 systems or so, and then sued them for BACK LICENCE fees and ROYALTIES, causing the most financial damage to the company. Or wait till 10,000 or so had SHIPPED and the first manufacturers were JUST STARTING to release the boards, then get an INJUNCTION which would prevent them form being sold, forcing VIA to REIMBURSE motherboard manufacturers for their losses!
You see, a historically GOOD company, ALi, has a licence. And an historically questionable company who has recently tried to improve their image, SiS, also has one. So it's actually quite obvious that Intel is NOT trying to OWN the chipset market as Tom had deceptively implied, but rather that Intel does not do business with VIA. Smart move.

Back to you Tom...

 

[juin]

Tom had deceptively implied, but rather that Intel does not do business with VIA. Smart move.


Some have my point in this matter.For the P3 bus intel have sue they.The probleme have been resolve in """out court""".Normaly the same thing will happen.

 

[Stick_e_Mouse]

Yeah, I agree with you that Intel made the right decision by slapping Via with a lawsuit. It was the only "right," or maybe "legal," thing for them to have done given the situation. VIA was wrong to release the chip without proper patent and license agreements.
Although VIA maybe a "shady" company, I believe that the release of the chip would have brought good things for us consumers. Im not saying I would have bought a board with one, but the release would bring much needed competition with the Santa Clara giants. Unlike you, I hate to see businesses go out in this already dismal economy--I would actually have liked to see VIA prosper because not only would prices drop, but we would start to see actual technological advancments (rather than just improvements), such as SiS's integrated chipset.

Either way, VIA needs to start ironing out it's bug infested chips and Intel needs to start making better decisions in the interest of us consumers (lower prices and better performance).

See a real naked pic of Britney Spears <A HREF="http://www.geocities.com/stick_e_mouse" target="_new">here</A>!!!

 

[Matisaro]

Crash, you wouldnt be biased against vianow would you? Heh

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

 

[madmike]

Yeah, it took balls, but VIA claims they're legal.
<A HREF="http://www.sysopt.com/articles/intel-vs-via/" target="_new">http://www.sysopt.com/articles/intel-vs-via/</A>
They see it this way:
- VIA owns S3.
- S3 has a license to utilize the P4 bus.
- The P4X266 will have an integrated S3 graphics chip.
- The P4X266 is legal.
This kind of maneuvering has been going on for years in this market. In the end, they might just settle out of court.

 

[Crashman]

My arse still hurts from my last dealing with them.

Back to you Tom...

 

[Crashman]

Ich glaube, Intel ist mit VIA nicht doof! Aber VIA, bissen sie doof! Ich spreche besser Duetsch dan juin hat English! Und minen Duetschen ist nicht so gut!

Back to you Tom...

 

[zengeos]

Yeah I read that several places. Of course, the real determination of legality will be how the S3 license was written. If it is transferrrable then VIA is probably legal. If not, then VIA doesn't have a license. So, expect it to take about 6 months to settle and by then the other chipmakers will have a sizable portion of the market and Intel's own DDR chipset will be out. So, Intel wins.

Mark-

When all else fails, throw your computer out the window!!!

 

[Kelledin]

I know just enough German to know that's not French. :tongue:

Kelledin

"/join #hackerz. See the Web. DoS interesting people."

 

[Crashman]

I think the P4 bus licence was never really there! I know they had a P3 licence, which VIA abused to make the Cyrix 3, but except for that one article (which has no proof to back it up) I had never heard of S3 having a P4 bus licence. It is possible that someone along the lines started a rumor by confusing the two, just because he writes doesn't mean he's right!

Back to you Tom...