Nehalem and Conroe Thread Simularities

[WaitingForNehalem]

I find it very interesting at how the same questions and doubts given to Nehalem were first given to Conroe. And look at how Conroe turned out.

POLL: Are you going to kick your old system to the curb when Conroe ships?

Intel blocks Nvidia from running SLI on Conroe platforms?

Gaming with Conroe

Is AMD gonna have an answer to Conroe.

Is Conroe right for me?

 

[apoppin]

What is your point?

Look back further on the P4 and NetBust architecture

- how well did PressHot turn out?
- about 6.4Ghz shy of projections

We all *know* Neha will turn out OK
but i expect .. expensive and buggy for the first few months

just like conroe

 

[harpoon84]

apoppin, just what exactly do you mean by 'buggy'? I agree Nehalem will be expensive (due to high platform cost) but how would you know its buggy before its even released?

 

[dmens]

conroe was buggy? how so?

 

[myocardia]

Originally posted by: dmens
conroe was buggy? how so?

The Core 2 memory management unit (MMU) in X6800, E6000 and E4000 processors does not operate to previous specifications implemented in previous generations of x86 hardware. This may cause problems, many of them serious security and stability issues, with existing operating system software. Intel's documentation states that their programming manuals will be updated "in the coming months" with information on recommended methods of managing the Translation Lookaside Buffer (TLB) for Core 2 to avoid issues, and admits that, "in rare instances, improper TLB invalidation may result in unpredictable system behavior, such as hangs or incorrect data."[56]

Among the issues noted:

Write-protect or non-execute bit for a page table entry is ignored.
Floating point instruction non-coherencies.
Allowed memory corruptions outside of the range of permitted writing for a process by running common instruction sequences.
Intel errata Ax39, Ax43, Ax65, Ax79, Ax90, Ax99 are said to be particularly serious.[57] 39, 43, 79, which can cause unpredictable behavior or system hang, have been fixed in recent steppings.

http://en.wikipedia.org/wiki/Intel_Core_2

 

[dmens]

oh yeah, the issue which linus torvalds said was "totally insignificant" and "not an errata".

of course if you define buggy as "not perfect", then good luck finding a suitable CPU. on the other hand, c2d didn't have anything worth a recall, or any issue that required a performance crippling patch. that's satisfyingly non-buggy for just about anyone.

 

[Idontcare]

Originally posted by: WaitingForNehalem
I find it very interesting at how the same questions and doubts given to Nehalem were first given to Conroe. And look at how Conroe turned out.

Is AMD gonna have an answer to Conroe.

Not trying to pick on anyone, and I'm sure Noubourne was basing their statement on solid data available at the time, but I found this quote from this thread to be rather ill-fated given the luxury of 2+yrs of hindsight at this point:

Originally posted by: Noubourne on 06/17/2006 09:51 AM
K8L 2007. Have patience. They've been winning for 2 years now. They can take 9 months off.

 

[apoppin]

Originally posted by: dmens
conroe was buggy? how so?

everyone forgets quickly

The early adopters always run into issues. Usually with the brand new MBs and the 3 pairs of DDR3 interaction [in this case] .. one set of new BIOS flashing after another

And they always pay way more than us guys who are just behind the bleeding edge

Yes, Neha IS the future .. but *when* will it reliably run away from a highly O/C'd Penryn in gaming

i am a gamer .. and i believe in future Quad Core .. but i will let you guys pay the big bucks to be essentially beta testers for intel

i will join you in six months [unless intel sends me a sample ] .. when most of you will be thinking of upgrading to the "new neha" and the newer "II" Motherboards and cheaper DDR3 .. that will most certainly be out in Spring.

i don't think a HD4870x3 paired with a fast O/C'd e8500 will take much of a back seat to anything for the next six months
- my guess .. and i put my $$$$ where my mouth is - X48/e8500/2x2GP PC 8500 should be here day after tomorrow. You can keep waiting for your elusive Godot and your "perfect PC" that will be "future proof"
:roll:

 

[dmens]

Originally posted by: apoppin
everyone forgets quickly

The early adopters always run into issues. Usually with the brand new MBs and the 3 pairs of DDR3 interaction [in this case] .. one set of new BIOS flashing after another

And they always pay way more than us guys who are just behind the bleeding edge

Yes, Neha IS the future .. but *when* will it reliably run away from a highly O/C'd Penryn in gaming

i am a gamer .. and i believe in future Quad Core .. but i will let you guys pay the big bucks to be essentially beta testers for intel

i will join you in six months [unless intel sends me a sample ] .. when most of you will be thinking of upgrading to the "new neha" and the newer "II" Motherboards and cheaper DDR3 .. that will most certainly be out in Spring.

i don't think a HD4870x3 paired with a fast O/C'd e8500 will take much of a back seat to anything for the next six months
- my guess .. and i put my $$$$ where my mouth is - X48/e8500/2x2GP PC 8500 should be here day after tomorrow. You can keep waiting for your elusive Godot and your "perfect PC" that will be "future proof"
:roll:

you got some concrete examples of issues?

we've been validating nehalem for a year already, the only beta testers are in house. whatever, anyone who runs stuff that is HPC or similar should get a nehalem asap, that is just my (professional) opinion.

 

[myocardia]

Originally posted by: dmens
oh yeah, the issue which linus torvalds said was "totally insignificant" and "not an errata".

of course if you define buggy as "not perfect", then good luck finding a suitable CPU. on the other hand, c2d didn't have anything worth a recall, or any issue that required a performance crippling patch. that's satisfyingly non-buggy for just about anyone.

Actually, it did require a microcode update. But shouldn't you know that? Luckily for the early C2D adopters, a BIOS update was enough. Not all early adopters are that lucky, though.

we've been validating nehalem for a year already, the only beta testers are in house. whatever, anyone who runs stuff that is HPC or similar should get a nehalem asap, that is just my (professional) opinion.

You're right, it is it an awesome server chip. What does that have to do with us, though? BTW, I hear from a non-inhouse beta tester (that I suppose don't exist, according to you) that they're wickedly fast at things like folding proteins.

 

[apoppin]

Originally posted by: dmens

Originally posted by: apoppin
everyone forgets quickly

The early adopters always run into issues. Usually with the brand new MBs and the 3 pairs of DDR3 interaction [in this case] .. one set of new BIOS flashing after another

And they always pay way more than us guys who are just behind the bleeding edge

Yes, Neha IS the future .. but *when* will it reliably run away from a highly O/C'd Penryn in gaming

i am a gamer .. and i believe in future Quad Core .. but i will let you guys pay the big bucks to be essentially beta testers for intel

i will join you in six months [unless intel sends me a sample ] .. when most of you will be thinking of upgrading to the "new neha" and the newer "II" Motherboards and cheaper DDR3 .. that will most certainly be out in Spring.

i don't think a HD4870x3 paired with a fast O/C'd e8500 will take much of a back seat to anything for the next six months
- my guess .. and i put my $$$$ where my mouth is - X48/e8500/2x2GP PC 8500 should be here day after tomorrow. You can keep waiting for your elusive Godot and your "perfect PC" that will be "future proof"
:roll:

you got some concrete examples of issues?

we've been validating nehalem for a year already, the only beta testers are in house. whatever, anyone who runs stuff that is HPC or similar should get a nehalem asap, that is just my (professional) opinion.

are you kidding?

for Neha? of course not. If so, i would be a paid beta tester with you
-however, do you really thing Neha will work flawlessly with all the new MBs?

No new CPU previously has .. they all have their issues for the early adopters

i rest my case and i will have to say "wait and see"

 

[Aberforth]

We don't really know how nehalem performs with the wide range of hardware. The best thing to do would be to wait for the refesh "sandy bridge" core because if you look at the Core 2 line up, there were a couple of issues with it. I guess they were fixed with bios updates.

 

[dmens]

Originally posted by: myocardia
Actually, it did require a microcode update. But shouldn't you know that? Luckily for the early C2D adopters, a BIOS update was enough. Not all early adopters are that lucky, though.

was it performance crippling? i know it wasn't.

BTW, I hear from a non-inhouse beta tester (that I suppose don't exist, according to you) that they're wickedly fast at things like folding proteins.

i dont call them beta testers because 99% of bugs are caught in house... guess they are more like external qualification. imho.

 

[myocardia]

Originally posted by: dmens

Originally posted by: myocardia
Actually, it did require a microcode update. But shouldn't you know that? Luckily for the early C2D adopters, a BIOS update was enough. Not all early adopters are that lucky, though.

was it performance crippling? i know it wasn't.

I'm having a hard time remembering, but it seems to me that it was causing BSOD's. Now, while it didn't lower performance, I'd personally rather have lower performance than continual BSOD's. But, admittedly I'm having to go on hearsay; I refuse to ever be an early adopter. I like waiting until any and all major bugs are worked out, and the "good" steppings are released.

i dont call them beta testers because 99% of bugs are caught in house... guess they are more like external qualification. imho.

Well, semantics aside, I wholehearted agree that Intel is the best in the industry at releasing silicon bugfree. Unfortunately, Intel doesn't make any motherboards (the ones that bear their name are made by Foxconn, not Intel), and that's the source of almost all of the early problems with any new chipset release, from what I've seen over the last 20 years or so.

edit: Umm, that was supposed to be "wholeheartedly agree"...

 

[apoppin]

Originally posted by: dmens

Originally posted by: myocardia
Actually, it did require a microcode update. But shouldn't you know that? Luckily for the early C2D adopters, a BIOS update was enough. Not all early adopters are that lucky, though.

was it performance crippling? i know it wasn't.

BTW, I hear from a non-inhouse beta tester (that I suppose don't exist, according to you) that they're wickedly fast at things like folding proteins.

i dont call them beta testers because 99% of bugs are caught in house... guess they are more like external qualification. imho.

Ok, external quality control; they do basically the same things. These testers would include the MB manufacturing partners of Intel; *those* guys only get 98% ..

98,99% of a million bugs .. er, errata is still not worth it for me to be an early adopter .. and the later steppings *always* OC better at a lower price with less frustration ..

OK, i really rest my case

i am just hoping i get e0 for my 8500 from Newegg; what do i do, send it back if it's not and eat the restocking and shipping?

 

[OCGuy]

I just hate the fact that the OP keeps us in the dark about his next move. Is he going Deneb? Bulldozer? I CANT STAND THE SUSPENSE!!!!!!!

 

[Idontcare]

Originally posted by: Ocguy31
I just hate the fact that the OP keeps us in the dark about his next move. Is he going Deneb? Bulldozer? I CANT STAND THE SUSPENSE!!!!!!!

HoldingOutforWestmere?

 

[OCGuy]

Originally posted by: Idontcare

Originally posted by: Ocguy31
I just hate the fact that the OP keeps us in the dark about his next move. Is he going Deneb? Bulldozer? I CANT STAND THE SUSPENSE!!!!!!!

HoldingOutforWestmere?

Its better than my original idea for a name:

ShouldISpendtheMoneyToBeAnEarlyi7AdopterOrKeepMyExtremelyFastCore2