Majority of real Phenom Users are Very Happy

Page 5 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.

coret

Distinguished
May 29, 2007
273
0
18,780


I wouldn't consider the current "fix" to be a real fix for the problem though. The current situation is like having a leaky sunroof in your car and using duct tape to solve the immediate problem until you get round to replacing the seal which is causing the problem. However, using the duct tape means you can't (quickly/easily) open the sunroof.
 



Coret, the current fix is the permanent fix for those who currently own the processor. Unless AMD is going to recall processors and replace them.

Duct tape it is!
 

keithlm

Distinguished
Dec 26, 2007
735
0
18,990


NO... not duct tape.

It is more like using super glue to stick a BIG giant square non-aerodynamic BOX to the top of the sunroof.

But at least the box is transparent.
 

coret

Distinguished
May 29, 2007
273
0
18,780
I was thinking of it in relation to the overall technology, not from an individual already-bought chip. For the latter, you are entirely correct.

Now to try and come up with a metaphor to use to promote the idea of not using a duct-tape fix... hmm...
 

yomamafor1

Distinguished
Jun 17, 2007
2,462
1
19,790


And to add another perspective:

If your top of the line, newest architecture product is not performing even close to your competitors, you wouldn't let your engineers to go on vacation, would you? I mean, AMD spent roughly about 2 years, couple million dollars, and several PR disasters to produce this product. Currently, this product is not even performing close to AMD's claim, to public expectation, yet the hardware engineers get to take a break?

Not to mention, Phenom's competitor is outperforming it by a decent margin (IPC / clockspeed). How is Phenom going to compete, when it has no IPC advantage, no clockspeed advantage, no power efficiency advantage, no financial advantage, and no marketing advantage? I suppose AMD would just sell these babies at fall out price, just to make up the cost of manufacturing?

If you were Hector, given this situation, would you let anyone in your company take an extra week of rest?



Its one thing to tout a feature, but its another to actually delivering it. Yes, I do applaud AMD's effort of making these CPUs compatible with older socket, but if the motherboard manufacturers delay on the BIOS development, then wouldn't it be just talk? Most, if not all, the motherboard manufacturers do not take those two weeks you mentioned off. This is similar to AMD's first claim of 40% performance improvement over its competition. Yes, I do applaud AMD for making the effort, but what's the use if they can't deliver it?

Now, the chipset designed for Phenom, the 790 series. I have no doubt it is a good chipset. This is probably the strongest link in the whole Spider platform. However, without Phenom being competitive, the whole platform is not competitive.
 

keithlm

Distinguished
Dec 26, 2007
735
0
18,990


I must question your logic.

If some motherboard manufacturers make a decision to NOT support the new CPU... using your logic it is then AMD's fault that they "only partially delivered."

As for the TLB problem... that is not really something that even comes up on the radar; many people trying to reproduce the problem have failed to see the "problem". So it appears to be mainly a marketing nightmare. Something only relevant in forum discussions. A real dunsel.
 

yomamafor1

Distinguished
Jun 17, 2007
2,462
1
19,790


It is AMD's fault to promise something, yet couldn't deliver. AMD made those promises, not the motherboard manufacturers. AMD should've worked out the BIOS issue with the manufacturers, prior to Phenom's launch.
 



AMD promissed compatability, not the motherboard manufactures. So yes, AMD is responsible for their own claims. Perhaps AMD was shooting for a BIOS update free upgrade and missed the mark?

Regardless, AMD is responsible for their own promises, not others.


In regards to the TLB:

If it is not serious then why did AMD pull the plug on several processors the day of the launch?
If not serious why has AMD stop-shipped all Barcelonas? (K10 Opterons)
If not serious why will AMD only ship Barcelonas to certain customers that aren't affected by the TLB issue?


AMD's own reaction doesn't support your claim that the TLB issue is not serious.
 

keithlm

Distinguished
Dec 26, 2007
735
0
18,990



It is not AMD's fault if a hardware company chooses not to support them. Spin that any way you want... but it won't change that basic fact.

Concerning the TLB... I am going by experiential data from actual people. People can not recreate the problem when they try really hard to do so. I'm sorry the facts don't fit your reality.
 

keithlm

Distinguished
Dec 26, 2007
735
0
18,990


That is about the only true statement I've see on this forum today.
 


Well, the facts sure do fit AMD's reality, as evidenced by product recalls at launch, stop ship orders, and their refusal to let tier one vendors buy\sell Barcelonas.


Are you suggesting that AMD likes to randomly recall and stop-ship products for the fun of it?



And once again, AMD PROMISED COMPATIBILITY, NOT THE MOTHERBOARD MAKERS!


Intel's supposedly "limited upgrade path" offers better dual-core to quad-core upgradable from most motherboards.
 

yomamafor1

Distinguished
Jun 17, 2007
2,462
1
19,790
Let's say I contracted an OEM to make a motherboard for me, and they promised the cost will be 30 bucks.

However, when they delivered, they said they could only do 35 bucks, because they couldn't push down the price on the chipset.

Now, should I blame the OEM for failing to comply with our agreement, or should I blame the chipset manufacturer for not dropping the price?
 

dragonsprayer

Splendid
Jan 3, 2007
3,809
0
22,780
this thread is bs

the new black is out today- newegg $239

I almost bought one when i researched i did not - why?

They do not run past 2.5-2.8ghz. you might get lucky at 3ghz. this is the "top of the line black chip"

amd messed up the press to begin with, they messed up the cpu, the messed up the bios, they where wrong on the mobo compatibility.

it like amd is liar - if you ask me???? It would appear they consistently bs - one mistake after another is not probable.

warped_systemz.png



ps: they are only selling crossfirex with 3850 gpu not 3870 on intel cpu and chipsets - how dumb! Do not buy amd!
 

OlSkoolChopper

Distinguished
Dec 15, 2007
564
0
18,980


i think that sums it up beter than a hundred Wall St. analysts.
 

Hellboy

Distinguished
Jun 1, 2007
1,842
0
19,810
What I cant understand is this.......

How can they honestly say they are happy with a processor is slower than a year old one which has bugs in it and cant reach 3Ghz unclocked where the Q6600 can reach it clocked with a standard heat sink and fan....

Maybe im missing something here but does the thread starting fanboy who has nothing to contribute to this forum than flamebating AMD against Intel.....

Face facts fan-boy, AMD has had a lucky time with the A64, but not anymore.....


Turpit , if he does any more give him a final warning please m8, I am really getting tired of Thunderman going on and on about how good AMD is when obviously they need help when it comes to replacing the K8..

So what happend to the K9 as we know the K10 sounds like its going to be a pup....

Anyway so AMD have dropped the Sempron badge - its now called a Phenom...... :)


Best for all non biased factual people out there - TC, Turpit, I-ball and anyone else whos intelligent and read my threads.

PS we need more girls on this site - :)
 

sonoran

Distinguished
Jun 21, 2002
315
0
18,790
Actually, they could generate a GUID (http://en.wikipedia.org/wiki/Guid) for each line item on the invoice, and record each in a database. Email the GUID to the customer, and require entry of a valid GUID to post a review. And yes, they could even enforce a reasonable time delay (say at least one week after purchase) before allowing the item to be reviewed. It'd take some coding, but wouldn't be all that difficult. And I'm quite sure they already have a database storing info about invoices and the line items on each - they're tracking orders somehow. The biggest downside is that GUID's are pretty ugly for us humans to type, but they could easily set up their emails with hyperlinks that would go to a page that takes the GUID as a URL parameter and goes straight to a review entry form.
 

thunderman

Distinguished
Nov 13, 2007
107
0
18,680
Once AMD get those Phenom B3 Steppings out... AMD's fortunes are going to turn around for sure. B3 Phenoms will be able to clock at higher clockspeeds and have good overclocking ability. Phenom is not hugely faster than the K8 AMDs in old 'has been' clapped out software..when software takes advantage of all four cores these Phenoms will fly. Phenoms Multi-task very well and that's going to be their strength. Yeah criticize them now...wait and you will all want one when full potential is shown. Phenom is the true enthusaists processor.

Intel will be on their way out... because Penryn will not save them. Intel can move to 45nm however Penryn is the same old relic technology.If I wanted a double cheese I would have gone to McDonalds. Intel know this... so they have stolen the Phenom native quad to make a direct copy for the Nehalem processor.

AMD back on top for 2008! Like all empires Intel is failing apart and are finished! AMD fans unite! AMD4life!
 

Silverion77

Distinguished
Sep 21, 2007
2,106
0
19,790
Hes back!!!!!!!!!!

stealing.......AMD stole the whole idea of processors from Intel and Intel from others.

Chrysler stole the car from Benz and Ford stole from both of them.
I joined this forum before you so dont copy me and GO AWAY
 

4745454b

Titan
Moderator


Not true, I suggest you read the link I posted earlier. The bios fix is NOT permanent, it can be turned on and off with the AMD overdrive utility. If you want to game and want your CPU at full speed, then you can turn it off. If you need to use virtualization, one of the few tasks where the TLB shows up, then you can re-enable it in overdrive. I don't remember if the link said you need to reboot or not.


 

tamalero

Distinguished
Oct 25, 2006
1,130
135
19,470


dude.. obviusly, becausein SERVER ENVIROMENT, this TLB bug can HAPPEN.
while its very very rare in desktop stuff...

*edit*
aaa you know, I wonder if this "thunderman" is actually a bot programed to write the same kind of BS as shakyrou..
I mean.. they hvav done "emo" bots for AIM and AOL....
 

boduke

Distinguished
Oct 25, 2006
410
0
18,780


I think TC means it's a permanent fix in that there is not a microcode update for it that would get rid of the need for turning it on/off entirely. In otherwords, that's the way the B2 stepping is, like it or not, it's what you have to do to get around it.

Frankly - the fact that I'd have to turn/off a bios fix for a bug regardless of what I'm doing on my PC is a HUGE turn off. I do use virtualization, extensively, so the inconvenience of turning something off due to a bug in the processor is not tiny.

That, and the fact that they underperform, is what has made me decide to move my VMWare server to my Q6600 when I get my Q9450 sometime this quarter. Since there's no upgrade path (as stated there would be by AMD) for my AM2 board that's currently in there now and the performance/bug issues, the Intel Quad will work nicely for a long while.
 

evilr00t

Distinguished
Aug 15, 2006
882
0
18,980


Slight problem. Nehalem is 45nm, and last time I checked, Phenom is 65nm. Hence it's unpossible for Nehalem to be a direct copy. :kaola: Oh wait, you're a troll. :sarcastic:
 

keithlm

Distinguished
Dec 26, 2007
735
0
18,990


I am suggesting that AMD over reacted to this issue. It makes much more sense to stop shipment and fix a problem than to release it and then hope nobody notices the problem. (You know... like other large chip companies do.) AMD made the choice to appear pro-active and not re-active; especially since they sell a lot to the SERVER market where this would manifest more frequently. (Perhaps also since they are attempting to dive into the embedded chip market... and from a management/marketing perspective this seems like a better idea in light of that. You can bet that this decision was heavily debated.)

Guess what! AMD delivered compatibililty! The chip works in some AM2 mother boards. If the motherboard maker does not want to support the chip.. that is their choice. AMD is not involved at that point. (Actually I suspect it will be supported on many AM2 boards that currently do not support it... but getting bios updates from motherboard makers takes more than 6 weeks... it doesn't matter how many people scream and moan on forums.)

AND LAST: did you notice you typed "from most motherboards" in your last sentence which refers to Intel's socket 775? Do you have double standards?