[Intel] There is NO TLB bug in Nehalem.

Quote:

---

Originally Posted by freeloader

America's definition of libel is not the same in different parts of the world. Here in Canada, it's almost impossible to find someone libel and get an awarded conviction. Very rare.

---

as i said, this is far more complex than a few lines at XS is going to justify, individual countries have limited laws to no laws concerning this subject,

but it seems there are 'some' cases in canada that go through ok,

TORONTO, Jan. 18 /CNW/ - A libel case brought by the Honourable Justice Dianne Pettit Baig, stemming from a February 8, 2000 article in the Globe and Mail and other media, has been settled. The defendants, who include The Globe and Mail, The Canadian Press, The Fort Frances Times, seven Fort Frances lawyers (Clare Brunetta, Ian McLennan, Wesley Derksen, Ken Koprowski, Lawrence Phillips, Emery Ruff and Donald Taylor), two members of the Fort Frances court staff (Margaret Katona and Donna Anderson) and a Fort Frances locksmith (Ken Rogoza) have all contributed to the settlement totalling $568,500.00. The amount is paid to Justice Baig for damages, including legal costs. The Globe and Mail paid $200,000.00 of the settlement amount. All of the defendants have made written apologies to Justice Baig who is a judge of the Ontario Court of Justice in the Thunder Bay and Rainy River Districts. For further information: contact her lawyer: Robert Rueter at Rueter Scargall Bennett LLP, (416) 869-3363. .

Quote:

---

Originally Posted by soundood

to OP;
well they would say that ehh? what you expect them to say,

"yes there is an erattum, and if you buy ci7, all your family will die of smallpox?"

which as all the AMD fans know is true, just hope all you intel guys have a good supply of bodybags.

i with the Fud all the way. :up:

---

Quote:

---

Originally Posted by Ghostbuster

That is un-called for. Is this how your mind works? We proved that this is a non-event and that FUD was merely posting some sensasionalist BS.. And you tell prople will die of smallpox? :down:

---

it seems the irony of my post was lost in your non humour day, lighten up dude :up:

Quote:

---

Originally Posted by LOE

commie

---

Would you like more infraction points already? :rolleyes:


EDIT: You guys are unbelievable... can you not just have a discussion without the sniping and BS. Seriously? Everyday I am amazed at the caveman attitude of people in these threads. I MIGHT understand if the content of this thread was something really important! Like [enter market leader of any harware type] going out of business or something... but this?

Remember we're again disucussing a 'misleading interpretation' from Fuad. Not the end of the world.

Quote:

---

Originally Posted by Oliverda

There are TLB bugs in Nehalem:

AAJ42 X Plan Fix: Incorrect TLB Translation May Occur After Exit From C6

AAJ59 X Plan Fix: Writes to IA32_CR_PAT or IA32_EFER MSR May Cause an Incorrect ITLB Translation

AAJ54 X No Fix: Core C6 May Clear Previously Logged TLB Errors

source: 10-12 pages

---

That was already covered by Intel's response. See first post.

Quote:

---

Originally Posted by soundood

it seems the irony of my post was lost in your non humour day, lighten up dude :up:

---

Urmm... Sorry I do not call that humour in any sense... especially the line "all your family will die of smallpox", sounds more like cursing. :nono:

EDIT: Telling people's family to go and die a slow and horrible death is NOT funny and its certainly NOT humour! You can compare this kind of statements with those of terrorists and extremist threats. Seriously! :mad:

Quote:

---

Originally Posted by kl0012

That was already covered by Intel's response. See first post.

---

Intel just talked about the AAJ1 erratum. :rolleyes: These erratums are still exist. See the official Intel PDF.

Quote:

---

Originally Posted by Oliverda

Intel just talked about the AAJ1 erratum. :rolleyes: These erratums are still exist. See the official Intel PDF.

---

Guess how many TLB "issus" are still on 65nm C2 or X2. :rolleyes:

Quote:

---

Originally Posted by Oliverda

Intel just talked about the AAJ1 erratum. :rolleyes: These erratums are still exist. See the official Intel PDF.

---

No. They talked about clarification AAJ1. And then other erratas were mentioned.
SPEC CLARIFICATION AAJ1 was initially added due to an issue on the Intel® Core 2 Duo processor which was previously corrected with a BIOS update; this issue does not impact the Nehalem Family of CPUs. There are errata on the Intel® Core i7 processor that relate to the TLB. These all relate to improper translations or error reporting, and all of those that impact functionality have been fixed via BIOS updates prior to Core i7 launch.

Quote:

---

Originally Posted by Hornet331

Guess how many TLB "issus" are still on 65nm C2 or X2. :rolleyes:

---

You can count: http://www.amd.com/us-en/assets/cont...docs/41322.pdf :)

It has 42 erratums. The Shanghai/Deneb has less.

So, the i7 has exactly 77 erratums. Ohh man...maybe that is the source of the brand i7. :D

Quote:

---

Originally Posted by Oliverda

You can count: http://www.amd.com/us-en/assets/cont...docs/41322.pdf :)

It has 42 erratums. The Shanghai/Deneb has less.

So, the i7 has exactly 77 erratums. Ohh man...maybe that is the source of the brand i7. :D

---

reading comprehension dude... i said TLB isuss, not erratas, also Core 2 Duo and Athlon X2. :rofl:

Quote:

---

Originally Posted by Oliverda

You can count: http://www.amd.com/us-en/assets/cont...docs/41322.pdf :)

It has 42 erratums. The Shanghai/Deneb has less.

So, the i7 has exactly 77 erratums. Ohh man...maybe that is the source of the brand i7. :D

---

Whoa, epic. Nice find:p:

Quote:

---

Originally Posted by Oliverda

There are TLB bugs in Nehalem:

AAJ42 X Plan Fix: Incorrect TLB Translation May Occur After Exit From C6

AAJ59 X Plan Fix: Writes to IA32_CR_PAT or IA32_EFER MSR May Cause an Incorrect ITLB Translation

AAJ54 X No Fix: Core C6 May Clear Previously Logged TLB Errors

source: 10-12 pages

---

Good find!!! If I was you I would send every errata on Ci7 I find to FUDZILLA so that he can make an article on it:rolleyes: We have to find something wrong with that pesky Ci7 before it makes it to servers.;)

Quote:

---

Originally Posted by Oliverda

You can count: http://www.amd.com/us-en/assets/cont...docs/41322.pdf :)

It has 42 erratums. The Shanghai/Deneb has less.

So, the i7 has exactly 77 erratums. Ohh man...maybe that is the source of the brand i7. :D

---

Quote:

---

There may be missing errata numbers. Errata that have been resolved from early revisions of
the processor have been deleted, and errata that have been reconsidered may have been
deleted or renumbered.

---

Its not entirely the same. Plus nor Intel or AMD actually list all erratas. They list what is "needed".

Quote:

---

Originally Posted by Oliverda

It has 42 erratums. The Shanghai/Deneb has less.

---

How do you know that? Specification update for Shanghai wasn't released yet.

I don't get it. The Intel guy says "there is no TLB bug" and then explicitly describes an issue where corruption occurs and states that it was fixed with a BIOS update prior to launch.

...so you're saying there's a TLB bug, but it's alright since you patched it with software? That sounds familiar.

Quote:

---

Originally Posted by kl0012

How do you know that? Specification update for Shanghai wasn't released yet.

---

guess we have a new magic crystal ball user. :ROTF:

Quote:

---

Originally Posted by kl0012

How do you know that? Specification update for Shanghai wasn't released yet.

---

Well simple AMD=good, Intel=bad. Silly wabbit:D

Quote:

---

Originally Posted by Oliverda

There are TLB bugs in Nehalem:

AAJ42 X Plan Fix: Incorrect TLB Translation May Occur After Exit From C6

AAJ59 X Plan Fix: Writes to IA32_CR_PAT or IA32_EFER MSR May Cause an Incorrect ITLB Translation

AAJ54 X No Fix: Core C6 May Clear Previously Logged TLB Errors

source: 10-12 pages

---

Just as a note: C6 state means the system is in very deep sleep. PLLs, caches etc is turned of. And system state is in main memory only.

I would claim this is a non issue until i7 laptops are released.

http://images.anandtech.com/reviews/...ve/cstates.jpg

Quote:

---

Originally Posted by naokaji

Well, they contradict themself so often that atleast one of the 100K 2 line stories has to be true:rofl:

---

Best line in the thread! Unfortunately I disagree. IMHO, I think FUD knows exactly what he's doing. I call it "Hit Phishing" or "Hit Mining". He knows sensationalism and outrageous-ness gets hits. Hits brings in the Bucks. Sure, there's enough folks hoping, wishing and praying Intel slips up to give more hits. Then there are the folks going there to argue the other side and so on. But shame on anyone calling themselves a n Xtreme-Systems Geek or Enthusiast even clicking on that site (or the other) more than once a week and then only for laughs. Folks flocking there for either Pro or Con only encourages him to continue to post such crap-tacular BS!

Quote:

---

Originally Posted by Shintai

Just as a note: C6 state means the system is sleeping. PLLs, caches etc is turned of. And system state is in main memory only.

http://images.anandtech.com/reviews/...ve/cstates.jpg

---

I know, but for example the AAJ59 is not a C6 state related erratum.

Quote:

---

Originally Posted by kl0012

How do you know that? Specification update for Shanghai wasn't released yet.

---

You'll see it. :)

Quote:

---

Originally Posted by Particle

I don't get it. The Intel guy says "there is no TLB bug" and then explicitly describes an issue where corruption occurs and states that it was fixed with a BIOS update prior to launch.

...so you're saying there's a TLB bug, but it's alright since you patched it with software? That sounds familiar.

---

The issue is people use overall macro generalizations for terms they dont understand. So anything with bug and TLB must mean the same as AMDs TLB bug. Tho they may not have a single bit in common besides regarding to the TLB.

Quote:

---

Originally Posted by Oliverda

You can count: http://www.amd.com/us-en/assets/cont...docs/41322.pdf :)

It has 42 erratums. The Shanghai/Deneb has less.

So, the i7 has exactly 77 erratums. Ohh man...maybe that is the source of the brand i7. :D

---

This will slowly degrade into another flame war... AMD has not published errata for Shanhai/Deneb, Barcleona had at one point over 300 errata ... the number of errata does not make or break a product. AMD removes errata entry if there is no issue or the internalize it ... the index them sequentially, the last errata entry is number 355 ... hence, a total of 355 errata were discovered since the first stepping.

These type of problems are discovered after validation, which is pretty extensive from the commentary Torvals suggests. This is just a bunch of hay over nothing.

Quote:

---

Originally Posted by qurious63ss

Well simple AMD=good, Intel=bad. Silly wabbit:D

---

QFT!:yepp::rofl:

Quote:

---

Originally Posted by Particle

I don't get it. The Intel guy says "there is no TLB bug" and then explicitly describes an issue where corruption occurs and states that it was fixed with a BIOS update prior to launch.

...so you're saying there's a TLB bug, but it's alright since you patched it with software? That sounds familiar.

---

The point in question was a spec clarification, this is not a bug ... the errata list other bugs that are either benign or have been addressed with microcode updates. ... not terribly difficult to understand.

Quote:

---

Originally Posted by Oliverda

You can count: http://www.amd.com/us-en/assets/cont...docs/41322.pdf :)

It has 42 erratums. The Shanghai/Deneb has less.

So, the i7 has exactly 77 erratums. Ohh man...maybe that is the source of the brand i7. :D

---

Barcelona could of just had 1 errata, TLB bug, and C2 could of had 100s but still barcelona's was a show stopper and that's the difference. Point being that you can't just count errata and say this process/arch is better then the other one it just doesn't work that way.

Quote:

---

Originally Posted by Shintai

The issue is people use overall macro generalizations for terms they dont understand. So anything with bug and TLB must mean the same as AMDs TLB bug. Tho they may not have a single bit in common besides regarding to the TLB.

---

Well it's apparent they shared a major thing in common. Data corruption and system hangs.