|
Paul MaudDib posted:And also to be clear, Google didn't put together that chart (blog posts here and here), and that's a statement from AMD PR there, not Google. It's not clear where that image came from and it contradicts the papers themselves in several areas. Spectre cannot be patched in software and it was verified in Ryzen according to the paper. And issue 3 is the one that's driving the PTI software patches, issue 1 and 2 collectively make up the category of Specter exploits. The "software fix" for issue 1, if you can call it that, is patching thousands of individual applications. Dunno why you're calling me out, I explicitly attributed the statement to AMD PR.
|
#
?
Jan 4, 2018 05:28
|
|
|
|
| # ? May 30, 2024 08:33 |
|
|
Paul MaudDib posted:The SEC makes you schedule your sales two quarters in advance, it's kinda doubtful that BK knew the exact week that it was going to blow up. http://www.businessinsider.com/intel-ceo-krzanich-sold-shares-after-company-was-informed-of-chip-flaw-2018-1 Business Insider posted:* Intel CEO Brian Krzanich sold off $24 million worth of stock and options in the company in late November.
|
|
#
?
Jan 4, 2018 05:36
|
|
|
SwissArmyDruid posted:Dunno why you're calling me out, I explicitly attributed the statement to AMD PR. I'm not calling out you specifically, I'm calling out that image, because it's deceptive and it's being passed around as authoritative by tech journalists and others when it's not. It took me like a half hour of head-scratching trying to square away what's going on between that image and the papers before someone finally pointed out that's not actually an official source in any way, and then I realized that it's just wrong and/or coming from a very narrow/specific perspective. The "mitigation" for issue 1 is going application by application and patching the bug out, that's not actually a mitigation in any real sense. I'm pretty sure someone just mixed up 1 and 3 and didn't realize that 3 was the one there was a software mitigation for and also the one AMD was not affected by. SwissArmyDruid posted:http://www.businessinsider.com/intel-ceo-krzanich-sold-shares-after-company-was-informed-of-chip-flaw-2018-1 Haha holy poo poo, he filed in October? He gonna get suuueedddd Paul MaudDib fucked around with this message at 06:02 on Jan 4, 2018 |
|
#
?
Jan 4, 2018 05:41
|
|
|
Paul MaudDib posted:I'm not calling out you specifically, I'm calling out that image, because it's deceptive and it's being passed around as authoritative by tech journalists and others when it's not. It took me like a half hour of head-scratching trying to square away what's going on between that image and the papers before someone finally pointed out that's not actually an official source in any way, and then I realized that it's just wrong and/or coming from a very narrow/specific perspective. Going to jail my man. Martha went to jail for $45k
|
|
#
?
Jan 4, 2018 05:43
|
|
|
I'd  hard if he got caught doing something that blatantly obvious but INTC was up like 30% on October 30th after being stuck in the mid 30s for like ever so it's not unbelievable that he wanted to take a bunch of money off the table for a stock that was previously considered a door stop
|
|
#
?
Jan 4, 2018 05:57
|
|
|
For better or worse, this scene basically goes on in my head whenever I see Paul throwing AMD under the wheels of a never-ending bus, while poo-pooing anything negative about Intel:"SA AMD Thread Posters posted:Paul, do you hate AMD? SA Moderator posted:You don't have to answer that! Paul MaudDib posted:I'll answer the question. You want me to answer that? "SA AMD Thread Posters posted:This thread is entitled! Paul MaudDib posted:You want me to answer it?!?! SA AMD Thread Posters posted:We want the truth! Paul MaudDib posted:YOU CAN'T HANDLE THE TRUTH! We live in a world that needs frames, and those frames need to be provided by someone. Who's going to do it? AMD? ARM? Intel has a greater responsibility than any of you can possibly fathom You all cheer for AMD and you ridicule Intel. You have that luxury. You have the luxury of not knowing what Intel knows, that AMD's minuscule marketshare, while tragic, probably saves frames. And Intel's existence, while monopolistic and glacial in advancement, saves frames! You all don't want the truth because deep down in places you don't talk about at LAN parties and Reddit, you need Intel in that desktop. You want Intel in that desktop. Intel uses words like "Core", "threads", and "cache". Intel uses those words as the descriptor of a history spent developing something. AMD uses them as a marketing line. I have neither the time nor the inclination to respond to a group of posters who uses the very microarchitecture that Intel provides, and then questions the security flaws of the processors Intel provided! I would rather you all just thank and purchase Intel, and let AMD die away. Otherwise, I suggest you all become electrical engineers and create your own microarchitectures. Either way, I don't give a drat what any of you think you are entitled too! SA AMD Thread Posters posted:Do you hate AMD? Paul MaudDib posted:Intel is simply a superio... SA AMD Thread Posters posted:Do you hate AMD?!?! Paul MaudDib posted:YOU'RE GODDAMN RIGHT I DO! (fwiw,I don't have a preference for either company, just thought it's funny...)
|
|
#
?
Jan 4, 2018 06:02
|
|
|
Yeah, this last page or two really makes it apparent.
|
|
#
?
Jan 4, 2018 06:03
|
|
|
Sorry not sorry for baiting the hook
|
|
#
?
Jan 4, 2018 06:17
|
|
|
Risky Bisquick posted:Going to jail my man. Martha went to jail for $45k Honestly good riddance. BK has presided over an epic squandering of Intel's leads in process and uarch. Intel is finally just about ready to bring 10nm into action... for 2+0 core laptop chips. Speaking of, I wonder what this means for uarchs down the road. Intel is going to have to scramble for a fix for Meltdown, even if the impact is mostly limited to databases they won't want to needlessly yield ground. And both AMD and Intel are going to have to fix their speculative execution, I don't see how "just patch every single app, easy!  " is a viable approach at all. Not that I would complain if we went back to document-style Web 1.0 instead of needing 50,000 lines of javascript to render a news article, but...Will Intel scrap Cascade Lake and future generations and actually do a major new uarch revision? Do they even release Ice Lake at this point? Does AMD have time to ghettopatch Zen2 in time for a 2019 launch? (bearing in mind that both companies have apparently known about these exploits since last June - which just makes the BK thing even worse.) Paul MaudDib fucked around with this message at 06:27 on Jan 4, 2018 |
|
#
?
Jan 4, 2018 06:25
|
|
|
Paul MaudDib posted:That's not what the paper stated. Maybe you should read it? Or excerpt I previously linked for you? Does the language difference make any difference in the first image? They state they have empirically tested intel, but AMD has been verified to have attack applicability. Not sure why they would use different language there.
|
|
#
?
Jan 4, 2018 06:49
|
|
|
GRINDCORE MEGGIDO posted:I was referring to Paul's post about the patch. I don't think Paul is a bad person/poster. I think he/she does post some good information on plenty other topics. Just anything Intel/AMD is gonna be questionable all too often.
|
|
#
?
Jan 4, 2018 06:58
|
|
|
B-Mac posted:Does the language difference make any difference in the first image? They state they have empirically tested intel, but AMD has been verified to have attack applicability. Not sure why they would use different language there. They said "empirically verified" and "verified", I'm not sure that's much of a distinction. Also, in the third image they state "experiments were performed" and "vulnerability was observed" on Ryzen, so my interpretation here is they've run it on Ryzen. Also good news for once, AMD and Intel both think adding mfence instructions can mitigate Spectre. I guess that might have been what that image was referencing. There will definitely be some performance impact, but at least it'll help a bit when running untrusted code in your browser. That would take the attack surface way down IMO. That seems like an obvious enough fix that I'm unclear on why they didn't come up with this idea sometime in the previous 7 months, rather than the day the vuln went public. Did they just not tell those researchers about it or something? This whole release has just been a clusterfuck of conflicting information. Paul MaudDib fucked around with this message at 07:37 on Jan 4, 2018 |
|
#
?
Jan 4, 2018 07:06
|
|
|
 Welp.
|
|
#
?
Jan 4, 2018 07:19
|
|
|
I'm glad everyone is hating on Intel right now
|
|
#
?
Jan 4, 2018 08:59
|
|
|
https://twitter.com/LinusTech/status/948735713440186368
|
|
#
?
Jan 4, 2018 09:16
|
|
|
Haha for having a desktop CPU in it its not even all that thick for a 2017 mid-high end gaming laptop. Its almost like a svelter version of the old school desknote which I haven't seen in a long time. I kinda wish more companies would do something like that for the gaming/power laptop crowd. Thin n' light is nice but if I really need/want the performance and upgradability I'm happy to sacrifice those things to get it. Especially if it can save some money.
|
|
#
?
Jan 4, 2018 09:25
|
|
|
Paul MaudDib posted:That seems like an obvious enough fix that I'm unclear on why they didn't come up with this idea sometime in the previous 7 months, rather than the day the vuln went public. Did they just not tell those researchers about it or something? Imagine what would happen if suddenly all major chip vendors had gone around telling compiler vendors "hey, sprinkle mfence ops liberally around any array traversals and don't ask why" Did you imagine people would ask why and put a bunch of effort into figuring out why? Because that's what would happen.
|
|
#
?
Jan 4, 2018 16:26
|
|
|
Anyone benched Ryzen yet with KB4056892 ?
|
|
#
?
Jan 4, 2018 16:27
|
|
|
Munkeymon posted:Imagine what would happen if suddenly all major chip vendors had gone around telling compiler vendors "hey, sprinkle mfence ops liberally around any array traversals and don't ask why" That's what NDAs are for. Plenty of people outside Intel and AMD knew about this vuln and they managed to keep it under wraps for 7 months.
|
|
#
?
Jan 4, 2018 16:48
|
|
|
GRINDCORE MEGGIDO posted:Anyone benched Ryzen yet with KB4056892 ? There's a bunch of reports on Reddit saying Window's Meltdown mitigation is not enabled on AMD systems so there should be no difference.
|
|
#
?
Jan 4, 2018 18:28
|
|
|
repiv posted:There's a bunch of reports on Reddit saying Window's Meltdown mitigation is not enabled on AMD systems so there should be no difference. Holy poo poo, good work for not being dicks Microsoft.
|
|
#
?
Jan 4, 2018 18:31
|
|
|
repiv posted:There's a bunch of reports on Reddit saying Window's Meltdown mitigation is not enabled on AMD systems so there should be no difference. https://twitter.com/GossiTheDog/status/948833769963900929 essentially don't trust any benchmarks as there's too many ways for them to gently caress up and misinform people atm
|
|
#
?
Jan 4, 2018 18:37
|
|
|
Wiggly Wayne DDS posted:yeah take all of that with a grain of salt as there's steps required to activate the mitigations: You misread his tweet, and he may have misread the article. Windows Update needs to know that the antivirus isn't going to take down the entire system with that patch. This happens because some antivirus vendors are worse actors than virus makers are. Until it can determine that it won't deploy the patch. Since Windows 8 and 10 come with Windows Defender, the average user is going to get the patch. Windows 7 and Server 2008 R2 users have to have a registry key set because Windows Update on a client that old needs a registry key to know things because it can't go examine the antivirus on its own. Microsoft Security Essentials exist but isn't the default (Windows 7 still ships without built-in antivirus), so things could get interesting. That this leaves Windows 7 users with no antivirus out in the cold isn't something I'm going to lose much sleep over.
|
|
#
?
Jan 4, 2018 19:31
|
|
|
BangersInMyKnickers posted:Apple patched for this as of 10.13.2 i was more talking about the performance impact, but yeah this at least makes them look good compared to their recent spate of bizarre security flaws dont be mean to me posted:You misread his tweet, and he may have misread the article. i continue to be amazed people are still using third party antivirus software in 2018 Generic Monk fucked around with this message at 19:39 on Jan 4, 2018 |
|
#
?
Jan 4, 2018 19:37
|
|
|
dont be mean to me posted:You misread his tweet, and he may have misread the article.
|
|
#
?
Jan 4, 2018 19:38
|
|
|
they just left the desktop headspreader on and everything? since they're that unconcerned with thickness i'm betting the thing will cut off the circulation to your legs if you actually try to use it as a laptop
|
|
#
?
Jan 4, 2018 19:42
|
|
|
If coders can modify the bios to accept chip upgrades that's legit cool.
|
|
#
?
Jan 4, 2018 20:31
|
|
|
Generic Monk posted:they just left the desktop headspreader on and everything? since they're that unconcerned with thickness i'm betting the thing will cut off the circulation to your legs if you actually try to use it as a laptop never mind the heatspreader - lol the ZIF socket!
|
|
#
?
Jan 4, 2018 23:15
|
|
|
Seamonster posted:never mind the heatspreader - lol the ZIF socket! with the lever arm and everything, not even the twist type. hell yeah.
|
|
#
?
Jan 4, 2018 23:27
|
|
|
Cygni posted:with the lever arm and everything, not even the twist type. hell yeah. What is a "twist type" ZIF socket? I don't know that I've heard of that.
|
|
#
?
Jan 4, 2018 23:31
|
|
|
Paul MaudDib posted:What is a "twist type" ZIF socket? I don't know that I've heard of that. i dunno what the real name is, but i was talkin about the ZIF sockets in mostly older laptops (back when laptops had sockets) that used like a coin/flathead operated twist mechanism to tension it like-a-so: 
|
|
#
?
Jan 4, 2018 23:40
|
|
|
Intel still uses a ZIF socket for their mobile products in configurable SKUs as late as Haswell.  edit: Man, Haswell is older than I remember it being. edit edit: Still being used in Skylake, according to the Precision M6600 service manual. Current model Precision laptops (7500-series) seem to be soldered, though. SwissArmyDruid fucked around with this message at 00:25 on Jan 5, 2018 |
|
#
?
Jan 5, 2018 00:09
|
|
|
Paul MaudDib posted:That's what NDAs are for. Plenty of people outside Intel and AMD knew about this vuln and they managed to keep it under wraps for 7 months. Hmm, fair. You'd still want to update the compilers ASAP (meaning 6 months ago) because of the enormous lag time between updating a compiler and seeing the results out in the world and I bet someone would notice either the additional code in open sourced compilers or the different instruction output and, presumably, performance.
|
|
#
?
Jan 5, 2018 15:34
|
|
|
A SWEATY FATBEARD posted:
You have the exact same case as me, one I purchased in 2011? Based purely on how cheap it was. Same tangle of cables, same cpu cooler. What cpu is in that? I seriously thought I posted this picture until I looked at the fans. SSJ_naruto_2003 fucked around with this message at 17:28 on Jan 5, 2018 |
|
#
?
Jan 5, 2018 17:25
|
|
|
When is Ryzen+ expected to drop?
|
|
#
?
Jan 5, 2018 19:36
|
|
|
Initial report blurbs say end of March. Which parts of the lineup get released first can vary greatly between now and then
|
|
#
?
Jan 5, 2018 20:00
|
|
|
Reserved predictions for Ryzen+ are that it's basically the same CPU, right?
|
|
#
?
Jan 5, 2018 20:08
|
|
|
PerrineClostermann posted:Reserved predictions for Ryzen+ are that it's basically the same CPU, right? Essentially, yeah. It's on a slightly improved process, and they will be launching new chipsets (rebadged? i cant think of any new features to add). But they have explicitly said its the same architecture, so maybe a few hundred more mhz? They could have something up their sleeve, but I'm personally reserved in my expectations too. If yields are good, maybe we get 6 cores all the way down the stack or more aggressive pricing, both of which would be cool for gamers targeting the R5 1600/ i5 8400 bang-for-the-buck zone.
|
|
#
?
Jan 5, 2018 20:21
|
|
|
Cygni posted:They could have something up their sleeve, but I'm personally reserved in my expectations too. If yields are good, maybe we get 6 cores all the way down the stack or more aggressive pricing, both of which would be cool for gamers targeting the R5 1600/ i5 8400 bang-for-the-buck zone. You mean 8 cores? 1600 (and i5-8400) already have 6 cores.
|
|
#
?
Jan 5, 2018 20:37
|
|
|
|
| # ? May 30, 2024 08:33 |
|
|
Malcolm XML posted:Gpu driver bugs don't cause privilege escalation vulnerabilities across vms, yet You can do lots of other things, especially in non-vm environments. Khorne fucked around with this message at 20:44 on Jan 5, 2018 |
|
#
?
Jan 5, 2018 20:42
|
|