Sign in with
Sign up | Sign in
Your question

What can we expect from northwood.

Tags:
  • CPUs
  • Font
  • Product
Last response: in CPUs
Share
June 7, 2001 7:41:14 AM

There a lot of rumor.

What we know for sure or not on northwood have.

0.13 micron
512kb of L2 does is 126 bit or 256 like xeon.
Improve FPU true or not.
PC1066 support true or not.
Maybe a agp 8x
Who is the change in the new chipset only more pins.The dont add pins just for the fun.
And others who is true who is not. <P ID="edit"><FONT SIZE=-1><EM>Edited by juin on 06/07/01 03:49 AM.</EM></FONT></P>

More about : expect northwood

June 7, 2001 11:00:51 AM

The only two major advancements is the smaller process (.13) and the 512K L2 cache. As far as SDRAM support- that has to do with the chipset and not the processor. But, Intel is debuting the Northwood and i845 (DDR chipset) at the same time. If intel does a good job designing the chipset- it'll probably outperform RDRAM (judging from the performance of VIA's benchmarks with the P4 and DDR). But, it's too premature to say that for sure.

-MP Jesse

"Signatures Still Suck"
June 7, 2001 11:32:24 AM

Based on Intels track record, the Northwood will probably be a disappointment. It will certainly be faster then the P4, but not the "get all, be all" that was once promised.

But, until the 0.13 lithography equipment is delivered and working, there will be no .13 Northwood. Wait until late this year. This delay is out of Intels hands.

<font color=blue>This is a Forum, not a playground. Treat it with Respect.</font color=blue>
June 7, 2001 12:47:36 PM

Yes it will be 0.13um and have 512KB L2 cache. Better FPU, possible. The rest of the stuff is done in the chipset.

There was talk of an on-die memory controller but this is not likely as Northwood needs to support RDRAM and SDRAM so this will more like come in the next revision once the RAM wars have settled down a bit and the victor is apparent.

There should also be a few minor tweaks to improve instructions per clock but nothing major. The best Intel can hope for is more compiler support.

L
June 7, 2001 2:53:25 PM

Well the 0.13 micron and the L2 512kb are right. Well with the life intel is saying it will goto 6ghz. Well i think a 0.10 micron p4 can make it. Well northwood upto 4ghz atleast.

Nice Intel and AMD users get a Cookie.... :smile: Yummy :smile:
Anonymous
a b à CPUs
June 7, 2001 3:08:19 PM

6 GHz huh

Well the current p4 platform sure didnt last long.

<b>Anyways its nothign about MHz anymore!! anyways!</b>
Its about performance!

My god ppl
O yay!! Intel has a 10 GHz but.. even thou a 4 GHz AMD can kick its ass its still 10 GHz .. ooo 10! !!!!!

<b> Think !</b> ! <b>!</b>

--call it what you wish, with this machine I can make mercury flow in 3 directions at once--
June 7, 2001 3:15:43 PM

Im talking about a 0.10 micron p4 to hit 6ghz. Im saying northwood will only go up around 4ghz.

Nice Intel and AMD users get a Cookie.... :smile: Yummy :smile:
June 7, 2001 5:52:23 PM

I don't even think Intel has the equipment to make the Northwood right now. So speculate all you want, but it's going to be a while.

<font color=red>Amd or Intel? Who cares?? Not me...</font color=red>
June 7, 2001 7:18:37 PM

That impossable and not logic.
June 7, 2001 7:27:37 PM

The probleme i have found in all website.

0.13 micron is cause more heat in more dense place.Keep in mind that right now Intel standart for heating is higher that the one from AMD.Maybe intel can do the same as AMD in the heating probleme of cpu.I saw that intel have made a new heatsink.Maybe this the answer.I must choose on USB 2 or firewire and i think someting in the 3ICH was causing minor promble.
June 7, 2001 7:38:50 PM

I dont think DDR will work well on a P4 is suppost to work with Rimm.P4 is like a women sometime she like sometime not.Check the Nvisia benchmark for ther chipset.One line of benchmark is the same with VIA one line.
June 7, 2001 9:17:07 PM

~ 0.13 micron is cause more heat in more dense place ~

This is not correct. The power requirements are less as the die size decreases, this is due to the reduction in distance and therefore resistance between transistors. This allows lower voltages to be used, less power means less heat.

<font color=blue> The Revolution starts here... as soon as I finish my coffee </font color=blue> :eek: 
June 7, 2001 10:11:31 PM

The true is in overall is do less heat but the distance between the gate is very small.0.13 use less power but in much more dense place that cause heat problem in some place in the CPU.Intel have create something to solve the problem.

I take this information on Intel developper forum.

If we take who toms and others site say intel is still the leader and there a whole story around the northwood/brookdale.I love that Intel wait to be sure at 100% that all work at full potencial.AMD have change the intel work.Good thing for all.I just hope AMD will do some R&D and stop to copy the concept from intel.
June 7, 2001 10:21:51 PM

<font color=blue>"I love that Intel wait to be sure at 100% that all work at full potencial."</font color=blue>

Come on now? Haven't you heard of the P3 1.13Ghz chip? They certainly did not "wait to be sure at 100% that all work at full potential". The chip did not work, and had to be recalled.

The P4 is another chip released before it was ready, just as a stopgap, as AMD had jumped too far ahead in the GHZ race. Think of the current P4 as a pre-release. The new P4, the "real" P4, the Northwood is incompatible with the current P4, requires more pins. Something must have been left out of the original P4 to get it to market quickly, and we all know the items that were left out.

Yes, Intel makes sure it's products are 100% ready before going to market. Give me a break, Juin!

<font color=blue>This is a Forum, not a playground. Treat it with Respect.</font color=blue>
June 8, 2001 1:18:18 AM

Up until the PIII 1 Ghz Intel was doing fine. It's been downhill since then. I wouldn't even have THOUGHT about buying an AMD processor before then.

<font color=red>Amd or Intel? Who cares?? Not me...</font color=red>
June 8, 2001 5:30:53 AM

Don't want to bore you guys with all the details, but let me clarify a few things regarding CMOS technology scaling:
1. While the channel resistance does go down with technology scaling, the interconnect (e.g. metal lines) resistance usually does not. The distance becomes smaller but so does the cross section of a metal line. In fact, that is the major reason to migrate to copper metalization -- to cut down interconnect resistance and capacitance.
2. While lowering supply voltage can drastically help power dissipation (V^2 dependence), the key is the scaling of the gate oxide of MOSFET so the high drive current can be maintained at lower voltage. When the silicon oxide scales to no more than a few atoms (monolayers) thick, the reliability and integrity of such oxide become serious issues.
3. The power dissipation proportionally scale with clock speed. This is true regardless of the chip size. How to get the extra heat out is a serious issue. From the manufacturing technology point of view, it would be a life saver if one can design chips to do more work per clock cycle so we don't have to go to ultra high clock speed. In that respect, the P4 is really the pits.

**Spin all you want, but we the paying consumers will have the final word**
June 8, 2001 6:03:50 AM

P3 1.13 and all P4 was made a earlier release to close the gap between the P3 and t-bird after to try to re-take the market with the P4.Look when windows make a earlier release full of bug is the same.Northwood will be later that expected.
June 8, 2001 10:43:32 AM

So you are saying Intel was justified in releasing the 1.13ghz P3 which did not work (and embarressed Intel)?

A complete contridiction to your previous post, that said Intel only releases a product when it is 100% reliable.

<font color=blue>This is a Forum, not a playground. Treat it with Respect.</font color=blue>
June 9, 2001 4:57:49 PM

Sorry for the confusion not on all cpu but in norhtwood case it seen to be the case.
June 9, 2001 5:06:35 PM

copy what concept from intel?

-* This Space For Rent *-
email for application details
June 9, 2001 6:46:44 PM

speedstep/Powernow
Eve6 buy from compaq
Instruction 3D now /sse anyway AMD have buy the sse and sse2.The whole x86 cpu in a mega lawsuit from intel.
June 10, 2001 1:35:06 AM

As a consumer, track records are important to me. The failed 1.13ghz P3, the P4 being released early and crippled in doing so, makes me hesitant that the Northwood will be perfect.

Intel has failed to deliver on its two latest products (1.13ghz P3, and P4). Hopefully Northwood will be all it is supposed to be, but Intels latest track record speaks for itself.

<font color=blue>This is a Forum, not a playground. Treat it with Respect.</font color=blue>
June 10, 2001 1:45:00 AM

there was no lawsuit - it is called licensing.

you cannot have things both ways - people cannot argue that AMD is not fully compatible with all pc apps and systems, and then moan because they licence the technology that does make them compatible.



-* This Space For Rent *-
email for application details
June 10, 2001 10:33:02 AM

Phelk is correct! It will not produce more heat!! Due to the fact that the traces are smaller they will require less voltage to transmit the signal acroos them!!! Less votage = less heat!!!!

Enough of my exclamations!!!!!!!!!!!!!!!!!!!!

<b><A HREF="http://www.seti.tomshardware.com/" target="_new">How fast is your PC</A></b>
June 10, 2001 12:41:12 PM

You are kidding right? What does EV6 (we are not talking about the band)bus from compaq how are they taking that from Intel?

PowerNow?
3DNow?

Are you just trying to piss people off....why don't you read before you post...maybe read it in onther language, so you can understand it.

Sorry man but you talk to much and don't know what you are talking about. I love the fact that you slam a company and don't know anything you are writing about.


96.3 % of Statistics are made up.
Anonymous
a b à CPUs
June 10, 2001 2:39:54 PM

What can we expect??? More overpriced, RDRAM infested crap!
Anonymous
a b à CPUs
June 10, 2001 2:55:19 PM

Juin, you should know what you're talking about before you open your mouth. AMD has been doing an incredible amount of R&D the last 2-3 years. They have outpatented Intel the last couple years, and last year by a large margin. Ever hear of Hypertransport? Ever hear of Powernow? x86-64? So aside from licensing x86 from Intel, which they will always have to do as long as they make x86 processors, there isn't much AMD's copying. If you look at all the patents AMD has been awarded I would say they're innovating more than Intel.
June 10, 2001 3:46:28 PM

>Ever hear of Hypertransport?

yeah, ever hear of Api Networks formerly Alpha processors? The Credit really goes to Alpha and Compact for LDT aka Hypertransport, also note that the chip is called AP1011(Alpha Processors) interesting isn't it?

"AMD/VIA...you <i>still</i> are the weakest link, good bye!"
Anonymous
a b à CPUs
June 10, 2001 4:09:00 PM

Ahh, the worthless peanut gallery... Yeah, ever hear of Hewlett Packard? Itanium? If you consider Itanic (Itanium) some great feat the credit for this remarkable design <insert sarcasm> goes to HP.

"Intel...you are still overpriced crap, good bye!"
June 10, 2001 5:28:05 PM

Well what i know Northwood better not be cripple like the current P4. Northwood is intels make or brake. They would be stupid to cripple it. Well PC133 will be faster PC1033 rdram since PC100 is faster then PC800 rdram.

Well the Soundblaster 2 most likely will have a APU.
Well the Radeon 2 will have a full GPU.
Well Gaming wise Processors will not matter as much as before.

Nice Intel and AMD users get a Cookie.... :smile: Yummy :smile:
June 10, 2001 7:40:04 PM

Not really. Many companies will buy out another company when desirable technology can be acquired. Why reinvent the wheel?

Who did the R & D is immaterial. Who owns the technology is.

<font color=blue>This is a Forum, not a playground. Treat it with Respect.</font color=blue>
June 11, 2001 3:58:32 AM

Amd
X86-64 bit there 1 to 2 year behind intel.Amd will probaly run old 32bit software faster.
old trusty Fpu x87 who is faster on AMD.
The 1 t-bird was in a old 0.25 micron.
Intel made a improvement in the predict branch unit amd make 1 in pal.Speepstept is introduce amd introduce a simmilar tech a couple of time after.

The you have see how much time """"old"""" come in the way.
X86-64 will take many year because of AMD.If software developper can still make a 32 bit software they will make it to save cash.

R&D intel vs amd

LOL!!!!!!!!!!!!
June 11, 2001 5:08:43 AM

juin -- If you want to talk about Intel vs AMD in R&D, let me remind you a few things:
1. Intel used to lead AMD in manufacturing technology by AT LEAST half generation. However, AMD hooked up with Motorola at the 0.18 micron technology and made great stride. AMD, not Intel, was the first one produce x86 CPUs with copper metalization. Now AMD is working with IBM to become the first to use the SOI (silicon on insulator) technology to make CPUs. Moreover, AMD announced its manufacturing 0.13 micron technology at IEDM 2000, at the same time as Intel. In other word, AMD becomes quite competitive in manufacturing technology.
2. Although I am no expert in CPU design, I think the results speak for themselves. AMD has consistently produced competitive if not significant better CPU core design and architecture for the last few years.
3. If you want to talk about R&D budget, there is NO comparison. Intel used to spend more on R&D than the AMD's total revenue. However, judging from the end results, you have to admit AMD spend its resources much more wisely and efficiently.

The bottom line is not what you put out in press release, but it is what you can produce for your customers. In that regard, AMD is doing quite well despite the fact it's a much smaller company and it has much smaller R&D.

**Spin all you want, but we the paying consumers will have the final word**
!