View previous topic :: View next topic |
Author |
Message |
krinn Watchman
Joined: 02 May 2003 Posts: 7470
|
Posted: Tue Sep 19, 2017 10:29 am Post subject: |
|
|
For me, amd is doing it all right seeing thigobr's comment. A good point for amd.
But i still cannot believe they made ryzen, test it... and that nobody at amd has never test compiling under linux (that is sure not specific to compiling under linux, but expose the problem clearly). |
|
Back to top |
|
|
mir3x Guru
Joined: 02 Jun 2012 Posts: 455
|
Posted: Tue Sep 19, 2017 10:35 am Post subject: |
|
|
krinn wrote: | But i still cannot believe they made ryzen, test it... and that nobody at amd has never test compiling under linux (that is sure not specific to compiling under linux, but expose the problem clearly). |
That's obvious intel inside man sabotaged ryzen
(Its not a joke) _________________ Sent from Windows |
|
Back to top |
|
|
ct85711 Veteran
Joined: 27 Sep 2005 Posts: 1791
|
Posted: Tue Sep 19, 2017 3:45 pm Post subject: |
|
|
I'm willing to give that amd did test the cpu, but probably with standard compiling levels of the most basic settings for largest compatibility; so the cpu wasn't even running at full capabilities. |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54028 Location: 56N 3W
|
Posted: Tue Sep 19, 2017 5:21 pm Post subject: |
|
|
In the beginning was the plan, and then the specification.
... (Google if you need to)
The product will have been tested to the specification, not real world test cases.
That won't change. There are just too many real world test cases. _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
rich0 Developer
Joined: 15 Sep 2002 Posts: 161
|
Posted: Tue Sep 19, 2017 5:31 pm Post subject: |
|
|
NeddySeagoon wrote: | In the beginning was the plan, and then the specification.
... (Google if you need to)
The product will have been tested to the specification, not real world test cases.
That won't change. There are just too many real world test cases. |
If the CPU met the specification, it would pass the real world test cases. It isn't like the CPU says "hey, look, somebody is compiling something, let's behave randomly." We don't know exactly what is happening, but most likely AMD does since they've fixed the problem. There is some set of circumstances when the CPU decodes some opcode and does something other than what the opcode tells it to do.
My suspicion is that they ended up rushing it to market. It would have been obvious by late development that this product was going to have a huge impact on their bottom line and every day's delay probably cost them millions.
I won't even say it was the wrong decision. If they just followed up with a microcode patch or were a bit nicer about replacing CPUs for people who complained they'd get to keep their profits and make everybody happy. I'd have been happy if they simply had offered advance replacement without making me take 47 pictures of my case and change my voltages. If your CPU doesn't work with the firmware set to Auto on everything, then your CPU is broken. I'm not going to browse overclocking forums just to get the CPU to run stock. |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54028 Location: 56N 3W
|
Posted: Tue Sep 19, 2017 5:47 pm Post subject: |
|
|
rich0 wrote: | ... if they simply had offered advance replacement without making me take 47 pictures of my case and change my voltages. If your CPU doesn't work with the firmware set to Auto on everything, then your CPU is broken. I'm not going to browse overclocking forums just to get the CPU to run stock. |
The CPU is a part of the system. You need to do some basic system checks before swapping out the CPU.
There would be a lot more bad PR for AMD if they swapped out CPUs and it didn't fix fix the problem because of some other issue.
After 4 or 5 CPU swaps an the system is still faulty what happens? _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
rich0 Developer
Joined: 15 Sep 2002 Posts: 161
|
Posted: Tue Sep 19, 2017 5:58 pm Post subject: |
|
|
NeddySeagoon wrote: | rich0 wrote: | ... if they simply had offered advance replacement without making me take 47 pictures of my case and change my voltages. If your CPU doesn't work with the firmware set to Auto on everything, then your CPU is broken. I'm not going to browse overclocking forums just to get the CPU to run stock. |
The CPU is a part of the system. You need to do some basic system checks before swapping out the CPU.
There would be a lot more bad PR for AMD if they swapped out CPUs and it didn't fix fix the problem because of some other issue.
After 4 or 5 CPU swaps an the system is still faulty what happens? |
If this wasn't a known product defect I'd be more willing to buy that argument.
Every single CPU AMD manufactured prior to a certain date contained the flaw. From the serial number alone it could be verified that the customer had a problem. It wouldn't be necessary to swap it 4 or 5 times to fix the known flaw.
If they had fixed it with a microcode update I'd be willing to accept that (assuming it didn't degrade performance). Otherwise they should be willing to fix the CPU for no other reason than the fact that the CPU is known to be defective.
By all means have some troubleshooting tips on the website, but they shouldn't be doing RMA questions for a week before shipping a replacement, and they definitely shouldn't be requiring cross-shipment. |
|
Back to top |
|
|
Tony0945 Watchman
Joined: 25 Jul 2006 Posts: 5127 Location: Illinois, USA
|
Posted: Tue Sep 19, 2017 7:50 pm Post subject: |
|
|
rich0 wrote: | If your CPU doesn't work with the firmware set to Auto on everything, then your CPU is broken. I'm not going to browse overclocking forums just to get the CPU to run stock. | Exactly! |
|
Back to top |
|
|
pjp Administrator
Joined: 16 Apr 2002 Posts: 20048
|
Posted: Tue Sep 19, 2017 10:12 pm Post subject: |
|
|
NeddySeagoon wrote: | In the beginning was the plan, and then the specification.
... (Google if you need to)
The product will have been tested to the specification, not real world test cases.
That won't change. There are just too many real world test cases. | I have no problems with needing to RTM and addressing flaws later. That said, this release seems to demonstrate the basic need for at least a few AMD Dogfooders to install Gentoo, at least a few times. _________________ Quis separabit? Quo animo? |
|
Back to top |
|
|
Tony0945 Watchman
Joined: 25 Jul 2006 Posts: 5127 Location: Illinois, USA
|
|
Back to top |
|
|
|