Intel may have canned Tejas project?

[Sunburn]

Intel may have canned Tejas project? (update: Confirmed)

That is what the inquirer suggests. Anyway, I think this would be an excellent idea.

Intel may have canned Tejas project

The rush to 64-bit Pentium Ms?

By Mike Magee: Thursday 06 May 2004, 11:00
PERSISTENT RUMOURS from people close to Intel engineers are suggesting that the chip firm may have canned its work on Tejas, the chip that is supposed to follow Prescott.

These sources tell the INQUIRER, and as yet this is unconfirmed, that the project has been shelved before tape out, and layout resources are no longer working on it.

The tape out was supposed to have happened round about now.

The speculation on the semiconductor street is that Intel is frantically attempting to take the Pentium M to the next shrink and move it as rapidly as possible into the pole position.

And the legendary skunk works in Oregon, the same sources suggest, have already designed a 64-bit wide P6/Pentium M core, perhaps dropping the legacy stuff and including 32-bit compatability.

Which would strongly suggest, as we�ve reported before, that Intel might well bring its Conroe and its Merom projects closer to reality than first anticipated.

 

[Albuquerque]

Wouldn't hurt my feelings any -- I have a great deal of respect for AMD, even more now that they're brought 64 bit processing to the mainstream market. Intel is learning from their mistakes (and their competition) and is now moving towards the mainstream 64 bit environment along with producing lower-clocked, lower-heat much higher efficiency CPU's rather than simply ramping clockspeed.

Two very good moves that will benefit the entire computing populace everywhere.

 

[Sunburn]

Intel to formally confirm Tejas canned today

The future is the Pentium M

By Mike Magee in Barcelona: Friday 07 May 2004, 05:37
A REPORT says Intel will confirm today that its future Tejas microprocessor, the successor to Prescott, will never appear.

See Intel may have canned Tejas project.

The INQUIRER understands that another future Intel project, Tulsa, has also hit the chip equivalent of Skid Row.

The Wall Street Journal reckons that Intel today will confirm that Tejas will never appear because of problems with power consumption.

Tejas was to succeed Prescott at the end of this year, but there are considerable technical difficulties continuing with the P4 microarchitecture. Prescott itself generates rather a lot of heat.

We reported from the Intel Developer Forum earlier this year that the company was considering making dual core chips based on the Pentium M microprocessor. And that future plans for Intel desktop chips will be based on adaptations of the versatile and cool running Pentium M chip.

 

[Sophisticated]

i really dont care what the "inquirer" has to say...as soon as they here things they just post it then things change and they post that...they post what people wanna hear im tired of them...the real info comes from the company themselves

 

[doormat]

I severly doubt the Tejas project is canned. Just a few weeks ago, I talked to a chip designer from intel who was working on it... just seems really weird to can it that quick...

 

[Albigger]

i really dont care what the "inquirer" has to say...as soon as they here things they just post it then things change and they post that...they post what people wanna hear im tired of them...the real info comes from the company themselves

I also saw an article (as was mentioned in the quote above) in the Wall street journal that Intel would cancel the program.

I severly doubt the Tejas project is canned. Just a few weeks ago, I talked to a chip designer from intel who was working on it... just seems really weird to can it that quick...

Corporations can work in strange ways.

Anyway, the Pentium M class of chips seems to have a lot of potential, and if these rumors are true, I think it is better - as Albuq. stated above.

 

[Sunburn]

lol

 

[Stratcat]

Dumping Tejas for a Pentium-M variant is an excellent move, IMHO. I kust hope it's HT'ed, too...

Strat

 

[doormat]

Hmmm, I was wrong....

http://us.rd.yahoo.com/mymod/hdln/r.../nm/20040507/tc_nm/tech_intel_cancellation_dc

 

[Stratcat]

And the Inquirer re-confirming itself...

HERE

Strat

 

[cack01]

Tejas needed to be dumped. Intel is moving towards the Pentium M being their flagship cpu. It would basically be pointless and costly to continue with tejas. I doubt Tejas's tech is fully dead though, as Intel needs to continue to ramp up it Xeon line of chips.

 

[FIZZ3]

I don't agree that Tejas 'had to be' dumped. On paper, it is supposed to be the chip that fixes most of Prescott's current flaws. Starting with actually using those ~50M inactive Prescott transistors for SMT or 64bit computing. If it is indeed being canned, it means that Intel is unable to improve the current design (heat?); surely that is bad news.
While the Pentium M chip design is indeed looking to be very nice even for desktop chips, the level of performance needed is not there quite yet, not to mention 64bit capability. Getting the M up there will probably require considerable effort; leading to more delays before it arrives and leaving the consumer stuck with nothing but Prescotts in the mean time...

 

[Tyrinon]

Bum deal all around as I wouldn't mind having a nuclear reactor to crunch with.

-Tyrinon

 

[mazzy]

Well, maybe today we'll find more about it.
I think that if Tejas is canned a Prescott 4.0GHz at the end of the year will never see the light.
AMD haven't ramp up the speed that fast and will not do it in a while because it's working still on the transition to 0,09, IBM have big troubles on yields on his fab at 0.09, so Intel has not a big pressure to pump up the clock really fast.
About LGA775 at all ? Without Tejas and with Prescott on Big heat troubles, change the socket for what ? a 3.8 Prescott ?
Look like the P3 tulatin and a old "i815" board, a change for a dead road.
a NW 3.6 is really possible to fill the gap for a while.

 

[Albuquerque]

I don't think the transition to 64 bit processing is going to require a HUGE die increase on Intel's part. I'd bet a cup of coffee that Intel already has hyperthreaded Pentium M's as engineering samples, and likely in higher speeds than 2ghz.

I think what's going to happen is, a TON of people on these forums are going to be whining about the "lack of performance from the slower processor", when in fact it's likely that a 2.4+ghz Pentium M core would probably perform faster than any current production model of P4 (3.4ghz+) At least, that's my opinion / speculation.

 

[Sentential]

Good!.... That just decreased the time to my next CPU....Conroe

 

[NookieN]

I don't agree that Tejas 'had to be' dumped. On paper, it is supposed to be the chip that fixes most of Prescott's current flaws. Starting with actually using those ~50M inactive Prescott transistors for SMT or 64bit computing.

50M inactive transistors? Wherever did you get that idea? SMT and 64-bit together would consume less than 5% of the die. 50M inactive transistors would be half of the chip.


At any rate, Tejas would have brought several interesting new technologies to the table. Fortunately, most of those can be carried over to other architectures.

Tejas is not something that many people on these forums would have been interested in. It was never intended to be the direct successor to Prescott, but rather a high-end desktop and workstation chip. The P4 architecture is not dead though and a true successor to Prescott will likely still appear.

 

[theflyingrat]

I for one am quite curious and excited about this. The world doesn't need more space-heater CPUs (no matter how much fun they are to tinker with.) I have plenty of Pentium M CPUs here at work, and they work fantastically, and provide excellent efficiency and battery life. Not to mention, they don't get half as warm as our Pentium-4M laptops.

From what I gather, Dothan is supposed to be announced or released or something next Monday. I think this is fantastic - wasn't it rumored to have 2MB of on-die cache, as well as being able to ramp to higher speed? If they do this without dragging out the pipeline depth too far, it will make a smokin' fast CPU. Add an 800 MHz FSB and a dual-channel controller to a motherboard, and that would be one CPU I'd love to have, even at only 1.8 or 2.0 GHz.

 

[FIZZ3]

50M inactive transistors? Wherever did you get that idea? SMT and 64-bit together would consume less than 5% of the die. 50M inactive transistors would be half of the chip.


At any rate, Tejas would have brought several interesting new technologies to the table. Fortunately, most of those can be carried over to other architectures.

Tejas is not something that many people on these forums would have been interested in. It was never intended to be the direct successor to Prescott, but rather a high-end desktop and workstation chip. The P4 architecture is not dead though and a true successor to Prescott will likely still appear.

Half the die, exactly. The number was 30M though, not 50. I'm not sure whether inactive is the right word, but they seem to go unused in current chips. It's is speculated that it has to do with HT2.
I got this information here:

http://www.x86-secret.com/articles/cpu/prescott/p4e-8.htm

 

[cack01]

Putting 64 bit extentions on the Pentium will be quite easy. If you look at the pictures of a A64 die, the 64bit instruction area is extremily small as NookieN hinted at earlier. Once the move to .9 occurs this baby should ramp nicely I really feel we will see high end desktop speeds towards the begining of Q4 if the die shrink goes well.

I am very suprised that they canned Jayhawk too. I think this could mean one of three things:
1. They have no idea what they are going to do for Xeon or some unnamed chip.
2. The next Xeon will be a Pentium M. If this is true, then we could see a repeat of what AMD did on the A64 with cache. The last thing I want is the have my all my cache stripped away from me again.
3. Itanium is the future for servers and workstation.