TLB bug still present in B3 rev Phenoms

[xlink]

I think people forgot how much a 286 costed. Or a 50MB HD back then.

Back then your value end entry PC costed 4000-5000$ and not 400-500$ or less. And back then money was more worth than they are today. So with todays money it would easily have costed 10000$.

Or perhaps what some of the early TVs costed.

People should elarn history, and not just what they wish. Price is lowering, why? Competition? Only a tiny part. But the main part is VOLUME.

With no competition Intels prices would not be much different than today. maybe 10-15% higher. And they would still innovate is almost the same pace.

If not, other technologies would catch up and make the PC into the gadget segment. And without big volume they would earn as much. It can sound hard for some people, but Intel earns more by selling CPUs at 200$ than they do at 2000$.

true.
intel's biggest competitor right now is themselves. They still have to price new chips so that they can compete with the old ones.
though chances are margins will still go way up...

[happychappy]

Not good news for AMD.

BTW, is AMD doing a K10 mobile chip?

[naokaji]

BTW, is AMD doing a K10 mobile chip?

no roadmap showed any k10 mobile parts for the near future....

[xVeinx]

Not good news for AMD.

BTW, is AMD doing a K10 mobile chip?

They will have a K10 mobile processor, but that is pending 45 nm and much lower power consumption.

[Zytek_Fan]

I think people forgot how much a 286 costed. Or a 50MB HD back then.

Back then your value end entry PC costed 4000-5000$ and not 400-500$ or less. And back then money was more worth than they are today. So with todays money it would easily have costed 10000$.

Or perhaps what some of the early TVs costed.

People should elarn history, and not just what they wish. Price is lowering, why? Competition? Only a tiny part. But the main part is VOLUME.

With no competition Intels prices would not be much different than today. maybe 10-15% higher. And they would still innovate is almost the same pace.

If not, other technologies would catch up and make the PC into the gadget segment. And without big volume they would earn as much. It can sound hard for some people, but Intel earns more by selling CPUs at 200$ than they do at 2000$.

I remember my family bought a Pentium 1 when it first came out. Had 64MB of RAM.

It was $3,500 O_o

[knightwolf654]

the tlb problem only becomes a problem on servers, there is little effect on desktops, there is a guy here figuring out how to make the tlb bug come to the surface, only way so far is 4 virtual programs running with cool and quiet on.
see if i can find a link.

edit: http://www.xtremesystems.org/forums/...d.php?t=172026

[Zytek_Fan]

The TLB bug also effects OCing...

[JumpingJack]

the tlb problem only becomes a problem on servers, there is little effect on desktops, there is a guy here figuring out how to make the tlb bug come to the surface, only way so far is 4 virtual programs running with cool and quiet on.
see if i can find a link.

edit: http://www.xtremesystems.org/forums/...d.php?t=172026

This is true, the TLB bug is not a big deal for desktop users... a person is more likely to curse out Microsoft before they would suspect a TLB bug, BSODs and random OS lockups probably occur at probabilities 10x or 100x more often than this bug would...

However, server is where the margins are, AMD really needs to execute well there...

But honestly, this is really hearsay and rumor, so while fun to banter in a debate in back alley forums, it is a pretty mean rumor that would best be left alone .... if end of Q1 comes and volume shipments are still tardy... then we can point back and say... see... Barcelona B3 did not fix the bug... until then a pragmatic perspective is in order...

i.e. this is just a rumor, we cannot be certain.

[AMDDeathstar]

TLB...TLB...TLB....

I've had two Phenom 9500's since Nov.26
One crunching at stock speeds full load-> NO TLB CRASHES

The other one been overclocked
been running at 2772Mhz since Dec.27
TLB fix disabled
running games,benchies and crunching-> NO TLB CRASHES

If I had TLB crashing I don't my Phenom would have made it to 32nd spot on Riesel Sieve Boinc for daily RAC in less then a week
Computer ID 33998 owner Frank Encruncher
http://boinc.rieselsieve.com/orig/to...=expavg_credit

Does it matter if there is TLB bug on Phenom... Answer NO

[ryboto]

"...
The manufacturing process of 45 nm AMD (who will not use high-k technology unlike that of Intel) is still scheduled for the second half of the year. Initial results were "encouraging". Hopefully ...

This is strange, from what I've read, AMD will be using a high-k technology for 45nm.

[AMDDeathstar]

The TLB bug also effects OCing...

FUD also

[Zytek_Fan]

This is strange, from what I've read, AMD will be using a high-k technology for 45nm.

They were making fun of the hafnium based high-k process on their YouTube channel

[xVeinx]

This is strange, from what I've read, AMD will be using a high-k technology for 45nm.

If I recall, they will, but not initially. IBM and AMD had been working on a high-k solution I thought, but it wasn't able to be implemented at the outset.

[JumpingJack]

This is strange, from what I've read, AMD will be using a high-k technology for 45nm.

Nope will not happen.

They will retain the option, but they will not use it to begin with... this has also been stated in a few AMD analysts and techical day meetings:

While ramping 45-nm production early next year, AMD said that it will retain the option to include high-k/ metal gates in that generation, and will include high-k/metal gate technology in its 32-nm production generation.

http://www.pcmag.com/article2/0,1759...049TX1K0100488

Frankly they are too far behind, even IBM is not using high-K at 45 but will do so at 32 apparently.

[JumpingJack]

FUD also

How do you konw ?? Perhaps the 2.8-3.0 Limitation for stable overclocks is hampered by the bug?? Without it, maybe 3.0 - 3.2 is doable.

Though I actually agre, I don't think the bug affects clocking at all, it is more desgin and, most importantly, process limited for getting to the 3.0 Ghz mark.

[knightwolf654]

i hope AMD skips 45nm.

[Zytek_Fan]

i hope AMD skips 45nm.

They can't. You can't simply "skip" a manufacture process.

[[XC] 2long4u]

true.
intel's biggest competitor right now is themselves. They still have to price new chips so that they can compete with the old ones.
though chances are margins will still go way up...

True but back then there wern't as many people building their own computers, so the middleman got a higher %. There is almost no markup on pc's now, there can't be at $400 for a computer that includes a mouse, keyboard, monitor and printer.

[kl0012]

TLB...TLB...TLB....

I've had two Phenom 9500's since Nov.26
One crunching at stock speeds full load-> NO TLB CRASHES

The other one been overclocked
been running at 2772Mhz since Dec.27
TLB fix disabled
running games,benchies and crunching-> NO TLB CRASHES

If I had TLB crashing I don't my Phenom would have made it to 32nd spot on Riesel Sieve Boinc for daily RAC in less then a week
Computer ID 33998 owner Frank Encruncher
http://boinc.rieselsieve.com/orig/to...=expavg_credit

Does it matter if there is TLB bug on Phenom... Answer NO

The bug is very critical for servers.
Here is the problem:
http://www.x86-64.org/pipermail/disc...er/010281.html
Here is the answer:
http://www.x86-64.org/pipermail/disc...er/010282.html

[ryboto]

Nope will not happen.

They will retain the option, but they will not use it to begin with... this has also been stated in a few AMD analysts and techical day meetings:


http://www.pcmag.com/article2/0,1759...049TX1K0100488

Frankly they are too far behind, even IBM is not using high-K at 45 but will do so at 32 apparently.

Huh..I remember the news about that, but then I remember two more news articles that mentioned AMD reps speaking about High-k 45nm. Eh, I don't see how they expect to produce a competing product on their current process.

[JumpingJack]

Huh..I remember the news about that, but then I remember two more news articles that mentioned AMD reps speaking about High-k 45nm. Eh, I don't see how they expect to produce a competing product on their current process.

Yeah, the report was that either in the 2nd revision of 45 nm or in 32 nm.... I can dig up their presentation to technical analysts if you would like to see it.....

But no where has AMD officially announced high-K for any process technology on the horizon, only that the retain the option, what ever that means. But I agree, they will not be able to compete without it. I could be wrong, AMD could be 'smoke screening' in their earlier presentations for competitive reasons... i.e. ' make the other guy think we aren't going there', then surprise them.

[JumpingJack]

i hope AMD skips 45nm.

That can't happen for a few reason, but specifically because the industry typically 'moves' together more or less, going straight to 32 nm would not be possible becuse the equipment vendors do not have the appropriate equipment tuned up to support it.

[_Lone_Wolf_]

We need AMD or Nehalem will cost 2k$ per CPU until late 2009..

Might I suggest the main reason for continous price reductions has been the the dramatic improvement in process technology coupled with simple economies of scale rather than competative pressure.

With roughly equal die areas for each of the 80386 and Penryn the latter will be cheaper due to vastly more viable dies being delivered from the larger wafer being used in it's manufacture.

Frankly it seems Intel is already operating as a monoploy and has been since Core 2's introduction. They are releasing new product to their own schedule and setting compelling pricing without the slightest threat from AMD. They can't arbitrarily set prices and let the market stagnate as this would both impact on their revenue and allow other competitors in. Would most people really need a PC if a net enable, keyboard equipped games console bundle with Linux and an applications suite were available?

I'm not denying competative pressure plays a role in pricing, but I think it's current importance is overstated.

[hersounds]

[savantu]

TLB...TLB...TLB....

I've had two Phenom 9500's since Nov.26
One crunching at stock speeds full load-> NO TLB CRASHES

The other one been overclocked
been running at 2772Mhz since Dec.27
TLB fix disabled
running games,benchies and crunching-> NO TLB CRASHES

If I had TLB crashing I don't my Phenom would have made it to 32nd spot on Riesel Sieve Boinc for daily RAC in less then a week
Computer ID 33998 owner Frank Encruncher
http://boinc.rieselsieve.com/orig/to...=expavg_credit

Does it matter if there is TLB bug on Phenom... Answer NO

How do you know that a crash or data corruption problem was caused by the TLB bug ? Do you have advanced debugging tools ?