Slashdot Mirror
New Ryzen Running Stable On Linux, Threadripper Builds Kernel In 36 Seconds (phoronix.com)
An anonymous reader writes: After AMD confirmed the a "performance marginality problem" affecting some Ryzen Linux users, RMAs are being issued and replacement Ryzen processors arriving for affected opensource fans. Phoronix has been able to confirm that the new Ryzen CPUs are running stable without the segmentation fault problem that would occur under very heavy workloads. They have also been able to test now the Ryzen Threadripper 1950X. The Threadripper 1950X on Linux is unaffected by any issues unless you count the lack of a thermal reporting driver. With the 32 threads under Linux they have been able to build the Linux kernel in just about a half minute.
Apple needs this not the $700 more intel cpu!
I will love using dozens of these for my Beowulf cluster.
AMD is using CPUs from week 25+ to fullfill RMAs. They have been doing additional testing in Customer Service on those CPUs -people are getting boxes that have been opened with handwritten notes relating to this testing.
It's *not known with certainty that ALL* week 25+ CPUs are good. AMD has made no official statement on that. They sent Phoronix a testing CPU just like they have been sending to their RMA customers.
Most stores and retail sellers are still selling pre-week 25 CPUs, so those may still be impacted.
I find it really interesting that Phoronix received a bunch of Threadripper and Epyc test hardware immediatly after they published AMDs PR-speak statement. And this article which has no concrete information is being used all over the internet to say that this is "fixed"
Is it the first time that AMD will have a stable processor? And most importantly, it it spiked by the Agencies?
For those of us that have not actually built a kernel, is 36 seconds astonishingly fast? A little faster? A totally random number with no meaning whatsoever?
Maybe some of you that do build kernels every once in a while could share your times along with specs for your rig.
- AMD acknowledged that pre week 25 CPU's are still on shelves and that customers who have one have the option to RMA if they are impacted(*).
- AMD didn't send the TR and Epyc to Phoronix to buy him off... they sent them for him to test so he verify that they did not have the same issue. Phoronix has a large following in the Linux community and they're the demographic most impacted by this issue. I can't remember if Phoronix was the site that original broke the issue but they researched it thoroughly so it makes complete sense for AMD to give them equipment to validate it's fixed.
(*) And before you demand AMD replace everyone's CPU's keep in mind that both Intel and AMD have numerous erratas for corner cases for every generation of CPU and they rarely replace CPUs. The latest "major" one for Intel (Kabylake/Skylake Hyperthreading) can only be fixed through a BIOS update for Kabylake chips and if you're motherboard manufacturer doesn't make an update available you're SoL. Same is true for my Atom C2750, there's a known issue that will cause some to brick after a period of time. Intel isn't RMA'ing consumer models (but they were/are doing RMA's for vendors).
Sadly, AMD did not learn from the AGESA 1006 fixes / PCI Passthrough mess on the original Ryzen. It is completely busted on X399/Threadripper, which is ironic because it now works on their low end parts and not their expensive workstation class Threadripper. The chipset also seems to be struggling across the board with PCIe card compatibility in general, with common cards like USB expansion cards not booting on multiple motherboards.
https://www.reddit.com/r/Amd/comments/6vbe6w/
https://community.amd.com/message/2820922#comment-2820922
Come on, /. , it's not difficult to make headlines that make sense...
Linux Running Stable On New Ryzen, Threadripper Builds Kernel In 36 Seconds
FTFY
Intel has been without a viable competitor for some time now (especially on low power CPUs). But they've been careful to keep it that way by giving OEMs sweetheart deals.
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
It's also on a high quality board from MSI with good quality RAM. AMD's a lot less tolerant of cheap boards/ram than Intel. You'll regret that $40 dollar AMD mobo. It's the reason I never bothered with an 8350. It wasn't any cheaper than the faster i5 by the time I factored in the cost of the mobo.
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
I had many Athlon XPs. I used to run Gentoo on them. Compiling in a distcc cluster. I beat the hell out of those processors, all on air with stock coolers.
You're spreading disinformation.
HBI's Law: Frequency of calling others Nazis is directly correlated with the likelihood of the accuser being Communist.
My Phenom II X2 550 BE, which had two unlockable cores in addition to the two "offical" ones, has been running rock-solid in quad-core mode since I built up the system, and does fine with the stock AMD cooler. Yeah, I lucked out with the extra cores, but it's been running 24/7 aside from occasional hardware changes and OS updates for at least six years now.
My only regret was not springing for 8 GB of ECC memory instead of 4 GB. At the time, I could only get 4 x 1GB sticks of ECC RAM at a reasonable price. But even with "only" 4 GB it's still pretty quick, especially for its age. A solid-state drive on a SATA-3 controller helped a great deal.
Oh, no! You have walked into the slavering fangs of a lurking grue!
Have an Ryzen 7 1700 that was affected by the segfault issue. Contacted AMD, they wanted a pic of my case to make sure it wasn't a thermal issue. Then asked me to try some different vcore/vsoc voltages and retest. When I still had the problem they shipped me out a brand new in box CPU, and it's been working perfectly.
AMD support is bloody stellar.
Yes they are very similiar but Threadripper is their consumer version of the upcoming Xeon competitor.
AMD admitted it did little testing on the regular Ryzen line as most consumers would be running WIndows anyway and admitted in the future they will test this out. FreeBSD is also impacted by the same bug where things get out of order and corrupted under heavy loads.
Threadripper has more cache and a different caching and memory as it supports NUMA and non-NUMA for server oriented loads and this is where the bug is here.
Unfortunately, this makes me very cautious to purchase an AMD system as it does have a reputation of being bargain grade. But, it is a brand new architecture from scratch. I maybe open to Ryzen2 or Threadripper2 after some of the bugs are worked out.
http://saveie6.com/
But as an example from yesterday, on a dual core 2.4ghz processor running a single channel of DDR3:
Around 8 hours to compile the kernel and modules. Or around an hour to compile just the kernel.
If you build a custom kernel for just your system you can have the whole thing done in under a half hour on similiar hardware. The distribution kernels having the kitchen sink compiled however are *HUGE* (6-26 megs for an xz to gz compressed kernel, and ~3 gigs for all the compiled modules.)
If you order one now, will it have the fix or not? In other words are they still selling units that were made before the fix was incorporated into the production process?
I am building a heavily threaded Windows program that does data management. One of the things it does is break SQL queries into small pieces and run them in parallel on multi-core boxes. I have queries that run more than twice as fast (or is that less than half as long?) as the same queries on PostgreSQL v9.5 running on the same box. I am concerned that this bug might affect my software.
AMD in many countries sends people who are affected by this bug off to their computer dealer if they did not buy a boxed retail CPU, but bought a whole computer with the Ryzen CPU as part of it.
Which basically means they are screwed: Computer dealers totally lack the knowledge or incentive to even understand the nature of this bug, especially given the fact that AMD has never explained which of their CPUs are affected and which are not.
So at best, your computer dealer will give you another computer in return, with a CPU in it that likely has the same defect. At worst, the computer dealer will just tell you that "Linux isn't supported" because many press articles wrongly claimed that this bug occurs only when using Linux.
Two of my friends are currently going through this service nightmare, and I have postponed my plans for Ryzen/Threadripper/Epyc computers until I know how this will turn out for them.
Ryzen is generation ONE of a new architecture and it already slaughters Intel's entire x86 range, top to bottom. So Intel, in desperation, floods forums with FUD.
This lying dribbler, guruevi- trusts that you, the Slashdot reader, are clueless. AMD encryption instructions are much faster than Intel's. Hyperthreading gen 1 on AMD is much more efficient than Intel hyperthreading gen 8, and what the hell is 'encryption' and 'hyperthreading' 'compatibility' even supposed to mean. These are things measured in performance alone, not 'compatibility'.
As for these 'adavanced' features, Intel actually disables them on most chips it builds. Ryzen suffers far less from this cynical ploy.
Oh, and BTW, the first two gens on Intel's hyperthreading were so broken you had to switch off HT in the BIOS to ensure serious software would run correctly on your computer.
It gets worse. Lying dribblers like guruevi previously stated that Intel's R+D spend, and engineering 'expertise' meant it was impossible for any competitor to ever match Intel again. Yet the first gen Ryzen chips are MORE power efficient than all current Intel parts- a true humiliation for Intel. All Intel has left is so-called AVX instructions, an almost never used set of parallel maths processing units. Only problem is that they use so much power they throttle (slow down) greatly if you try to use them.
AMD makes its CPUs on Global Foundaries. Yet TSMC, the giant Taiwan fab company, is twice as power efficient that GF. This is how far Intel's chip production has fallen behind in state-of-the-art engineering. Behind GF and far far far behind TSMC. No, Intel is finally going the way of DEC and Sun and other over-rated dinosaurs in this biz. Once the margins collapse, Intel won't be able to afford to stay in the game.
I am more concerned about the long term implications. Two years from now will it be iffy at best wether you should even bother trying to install Linux on that old AMD box that was retired from desktop use? Which boxes are the good ones? Which will be a problem? It's really the kind of nightmare nobody needs.
If only they would compile their proprietary driver for the latest XOrg. In the interim, I have switched to NVidia.
I excitedly put together a Ryzen 1600 box with 16G of ram and an nVME SSD (512G).. running the latest Ubuntu Mate, the thing is a pile of horse shit. Once every few days it just hangs.. no warning, no logs, no nothing, and it requires a HARD POWER OFF to come back. Reset button and three-finger-salute don't do it.
It also stutters and gets sluggish for no damn reason at all... I've tried replacing the SSD thinking that was it, but no, it still happens. Memtest tests good, but I even replaced the RAM just to see and no improvement.
I called AMD and they gave me the stiff arm. They said unless I could prove it was the CPU with Kernel Logs they would NOT replace the processor. It's a catch 22. I can't give them kernel logs because it's a HARD FUCKING LOCKUP.
Today I ordered an Intel board and I'm going to say fuckall to this AMD bullshit once and for all.
I'm shocked, just shocked.
Why is Snark Required?
64bit 4 core CPU will be eternally enough for all kinds of computing! (Sarcasm)
Apple switched to Intel because the PowerPC consortium wasn't delivering on their commitments for R & D sufficient to stay competitive with the power / performance ratio of Intel. Apple hardware was falling behind PC hardware. Part of why Steve Jobs was able to convince the Apple BOD to buy NeXT was because their OS was already able to deploy on either architecture.
Intel's R & D investments were justified by the guaranteed volume. PowerPC was a niche server (IBM) and desktop (Apple) player, in contrast.
If Apple buys AMD, then they're taking on the enormous R & D expense again to outperform Intel. To defray that expense, they'd have to maintain channels with other platforms they might eventually want to compete with like PlayStation and XboX. It works better for Apple to play both CPU vendors against each other for negotiating the best vendor contracts. Don't get me wrong, I'd like to think one of the reasons AMD investors were convinced to fund Threadripper R & D was because Apple was guaranteeing a bulk purchase for the forthcoming MacPRO pending AMD's ability to deliver a compelling power / performance ration. I'd love to build my next hackintosh on an AMD platform.
$5 / month hosted VPS on linux = awesome!
Opensource fans! No more proprietary fans, one more step towards freedom!
I've been avoiding AMD based cpu / chipsets for a while now, Mostly down to the fact that its just a mistake to
use AMD for running Linux. (Terrible GPU drivers) . Any Linux users out there care to comment on how AMD fare to Intel these days ?
Thanks
Electronic Music Made Using Linux http://soundcloud.com/polyp
There's a manufacturing problem. To solve this, they've come up with a stress test that gets them enough CPUs known to run the provocation test without failing, that they can ship those to people complaining about that problem. CPUs being sold continue to be as buggy as before, since in Windows such bugs get excused as video card shittiness or w/e.
I can install Gentoo in under a day.
How long does a chromium build take?
https://news.ycombinator.com/item?id=14733829
Phronix reported that AMD has debugged the problem and that all CPUs (1600-1800x for example) manufactured after week 30 (30 July) will not have this problem.
There is a date of manufacture engraved on the cpu cover.
Leslie Satenstein Montreal Quebec Canada
Phronix reported that AMD has debugged the problem and that all CPUs (1600-1800x for example) manufactured after week 30 (30 July) will not have this problem.
Can you provide a link? I am a regular Phoronix reader, but did not spot such a report. (And AMD returned CPUs manufactured earlier to some that RMA'd, and some of these _still_ had the defect.)