|
Saukkis posted:No, I think this is exactly what was talked about.
|
#
?
Mar 10, 2019 09:01
|
|
|
|
| # ? May 29, 2024 01:01 |
|
|
Craptacular! posted:As a person who simply buys whatever is best for my price range and doesn't get caught up in business wars, I'll never understand the people who take it to the level of thinking that chipmakers are using security research firms to sabotage one another etc. Oh, there's targeted marketing. Remember the "Nobody ever got fired for buying Intel... until now" campaign last year? With every additional Intel shortcoming that crops up, I expect AMD to keep pushing that line. Sorry for the reddit link, but some redditor took video of ads that were being placed at SJO airport last year. https://old.reddit.com/r/Amd/comments/90ho1m/shots_fired_amd_epyc_server_processor_campaign_at/
|
|
#
?
Mar 10, 2019 11:18
|
|
|
dont be mean to me posted:e2: this AMD community thread suggests it got supported driver-side but only for discrete Polaris. And not forward-ported to Vega. What. I don't get it, I was all set to order a 2200g for a httpc. I thought support arrived last may until I read nobody could get it to work. Also weird, that link mentions Ultra HD Blueray disk playback working. I thought they couldn't without sgx.
|
|
#
?
Mar 10, 2019 11:28
|
|
|
Craptacular! posted:As a person who simply buys whatever is best for my price range and doesn't get caught up in business wars, I'll never understand the people who take it to the level of thinking that chipmakers are using security research firms to sabotage one another etc. I don't actually think that is happening. For the ones that are sufficiently sketchy that they'd put their hands out like that, there's faster ways of getting paid. Like that israeli group with the bogus Ryzenfall thing, who were connected to an investment group that had positions in AMD stock. But my proposal of Intel handing out hardware to researchers wouldn't be sabotage. It would be sponsorship. And if they gave out their own chips as well -- not just AMD, ARM, Power, and whatever else -- it wouldn't even be unethical. If non-Intel CPUs have major security flaws, they should be discovered by ethical researchers who do proper disclosure. Not random blackhats or the GRU. But the whole idea was just a thought-experiment response to the idea that AMD CPUs are not being investigated as thoroughly because said researchers can only get their hands on a single AMD chip.
|
|
#
?
Mar 10, 2019 18:38
|
|
|
Combat Pretzel posted:You bought that new or used? Been a few busy days but I figured it out - the reviewers got a super nice engineered box and they tried to replicate that for the consumer version. The threadripper chip is held in by a latch that also closes the lid, and it's "prevented" from opening during shipping with a tiny cardboard insert. Don't try to make a latching mechanism out of cardboard. Without opening the one I sent back, I'm assuming the cardboard wasn't inserted right and the latch was just flopping around. The pod is cool as hell, and I'm keeping the package because I intend to sell this one as soon as the zen2s come out. Thermals:  tdie vs tctl: Which compares to "other" chips? This is under linux 5.0 so the offset bug should be fixed by now. tdie of 65c under full load is great, 90c tctl (junction?) is pretty hot though. I'm running a thermalright silver arrow TR4 until I do a custom loop later. Overclockin: I've got 4ghz at 1.225v, with ram at 3333mhz. I don't know what "normal" is since the MB defaulted to 1.075 but that seemed a little unstable at stock. Time spy likes it as well: https://www.3dmark.com/compare/spy/6213507/spy/6534395/spy/6535256 (4903 is windows bare metal, 4547 is is the VM with 4c/8t passed through.) E: tctl is +27c exactly, so it's just a straight offset for fan curves or something. I'll take the overclocking questions to the overclocking thread. Harik fucked around with this message at 04:20 on Mar 11, 2019 |
|
#
?
Mar 11, 2019 00:33
|
|
|
spasticColon posted:So even if I upgrade from my 2500k to a 8C/16T Zen2 CPU something much better is going to be right around the corner (1-2 years) anyway? If you're still on a 2500k, try manually disabling the meltdown/spectre mitigations. Breathed life back in to my system, overhead on those old CPUs is nasty especially if you have an SSD. Something in the background starts throwing a lot of ops around and the added CPU overhead from the syscalls turns everything to a stuttering mess.
|
|
#
?
Mar 11, 2019 17:13
|
|
|
BangersInMyKnickers posted:If you're still on a 2500k, try manually disabling the meltdown/spectre mitigations. Breathed life back in to my system, overhead on those old CPUs is nasty especially if you have an SSD. Something in the background starts throwing a lot of ops around and the added CPU overhead from the syscalls turns everything to a stuttering mess. I do have an SSD but I'm mostly concerned with gaming performance and I thought games weren't affected by the meltdown and spectre mitigations. It's only some newer games that are having performance issues but I simply chalked that up to my 2500k being 8 years old. If gaming is affected by the mitigations, is there a how-to guide on disabling them?
|
|
#
?
Mar 11, 2019 21:13
|
|
|
BangersInMyKnickers posted:If you're still on a 2500k, try manually disabling the meltdown/spectre mitigations. Breathed life back in to my system, overhead on those old CPUs is nasty especially if you have an SSD. Something in the background starts throwing a lot of ops around and the added CPU overhead from the syscalls turns everything to a stuttering mess. How do you do this?
|
|
#
?
Mar 11, 2019 21:23
|
|
|
spasticColon posted:I do have an SSD but I'm mostly concerned with gaming performance and I thought games weren't affected by the meltdown and spectre mitigations. It's only some newer games that are having performance issues but I simply chalked that up to my 2500k being 8 years old. If gaming is affected by the mitigations, is there a how-to guide on disabling them? It's little enough that it's probably not worth disabling unless you're using PCI-E or NVMe (M.2 PCI-E, and unlikely for a 2500K without a PCI-E adapter and - for booting - some very understanding firmware), in which case you should probably enable retpoline instead. It's been backported to Windows 1809 and will eventually be enabled by default in 19H1 but for now it's a registry hack, and a set of bit-flags, so read the part you get to if you find-in-page (Ctrl+F or F3) for 'bitwise OR'. What's happening is that drive access requires context changes from user to kernel space and with the mitigations that takes extra time each and every time. For NVMe it's worse because more of it is happening. Retpoline is basically diverting fruitless speculative instructions into endless loops that are promptly noticed and dropped by the CPU itself, which is faster than manually pruning them, so the context change takes less time. For other people reading this: If you're using Intel processors, you can do the same. (If it's Skylake or later retpoline doesn't work but it still enables BTI import optimization, which might help speed things up.) If you're using AMD, you'll have to figure out your own bit-flags if you still want as complete of Spectre protection as Windows offers (if you follow all the find-in-page 'full mitigation' parts in KB4073119 and that's all the overrides you need to do - which it probably is for home users - FeatureSettingsOverride is 0x448 (1096) and FeatureSettingsOverrideMask is 0x403 (1027)). v v v Oh well in that case if you're on 1809 and on any SSD you might try the retpoline registry thingy before disabling system security. An entire core, wow. dont be mean to me fucked around with this message at 21:48 on Mar 11, 2019 |
|
#
?
Mar 11, 2019 21:26
|
|
|
spasticColon posted:I do have an SSD but I'm mostly concerned with gaming performance and I thought games weren't affected by the meltdown and spectre mitigations. It's only some newer games that are having performance issues but I simply chalked that up to my 2500k being 8 years old. If gaming is affected by the mitigations, is there a how-to guide on disabling them? The newer game engines that are CPU bound and threading against all 4 core are definitely impacted. It's not the game itself incurring the overhead, its whatever background task that fires up that does a bunch of disk calls. My sata ssd doing moderate amounts of disk IO will consume almost an entire core worth of overhead with the mitigations enabled which starves whatever game I am playing and makes the frame rate stutter and drop for a couple seconds before going back to normal which is incredibly annoying. Turning them off brought that overhead down to the 5-10% range on the core it was hitting while servicing disk ios in the background which was low enough to not cause much of a noticeable impact. Newer processors (4th gen forward I think) have hardware PCID optimizations that greatly reduce the overhead and the impact is barely noticeable. e: https://support.microsoft.com/en-us/help/4072698/windows-server-speculative-execution-side-channel-vulnerabilities-prot That KB talks you through various configurations for the mitigations, but Win10 desktop has pretty much everything on by default. If you're feeling adventurous, I'd say try to retpoline route first. I didn't realize they backported that and was waiting on official support in the next Win10 build. BangersInMyKnickers fucked around with this message at 21:51 on Mar 11, 2019 |
|
#
?
Mar 11, 2019 21:43
|
|
|
Gah, that pretty much convinced me to go ahead and build an AMD rig this spring, probably in April or May. Either that or play older games and indie titles until Zen 2 becomes widely available meaning no price-gouging. And somewhat newer games don't "stutter" on my rig, I just get lower fps and some input lag. Then again I'm not playing every single new AAA game on my system.
spasticColon fucked around with this message at 00:09 on Mar 12, 2019 |
|
#
?
Mar 11, 2019 23:57
|
|
|
Why do there have to be shortages and price gouging? This isn't a game console, and it not something crypto miners are going to be interested in it. I suspect there will be a reasonable supply. I heard nothing about shortages with any Zen parts, and Amazon price history shows MSRP was the highest ever listed price for the 4 CPUs I checked.
|
|
#
?
Mar 12, 2019 00:18
|
|
|
SwissArmyDruid posted:Oh, there's targeted marketing. Remember the "Nobody ever got fired for buying Intel... until now" campaign last year? Except that this exploit is more along the lines of the Ryzen “exploits” of a year ago. You need physical access. At least, as long as browsers have been patched, which I’d think the big ones would be since Intel had three months to ask Google, Apple, and Mozilla to patch JavaScript.
|
|
#
?
Mar 12, 2019 00:37
|
|
|
There are still legions of IE6 out there in the corporate world.. probably all running on Intel processors from the era when Intel illegally cock blocked AMD from getting their processors into OEMs. :/
|
|
#
?
Mar 12, 2019 14:43
|
|
|
pixaal posted:Why do there have to be shortages and price gouging? This isn't a game console, and it not something crypto miners are going to be interested in it. Still, there is fairly large demand for zen2 and it's likely to sell out fast.
|
|
#
?
Mar 12, 2019 14:48
|
|
|
Khorne posted:I suspect supply may be fairly beefy. Supposedly the CPUs were ready to launch already but x570 is holding things up. that seems to be a thing with amd, cpus are available at lunch ie no paper launch but the boards partners are caught off guard by demand.
|
|
#
?
Mar 12, 2019 18:56
|
|
|
wargames posted:that seems to be a thing with amd, cpus are available at lunch ie no paper launch but the boards partners are caught off guard by demand. I think they've suddenly woken up to the fact that AMD CPUs no longer suck... the dark days of Bulldozer must have suppressed demand something fierce
|
|
#
?
Mar 12, 2019 19:03
|
|
|
 From buildzoid's Discord. Pretty obvious stuff, but still.
|
|
#
?
Mar 12, 2019 20:04
|
|
|
Bloody Antlers posted:There are still legions of IE6 out there in the corporate world.. probably all running on Intel processors from the era when Intel illegally cock blocked AMD from getting their processors into OEMs. :/ Windows 7 was shipped with IE8, so anyone on IE6 as their out-of-the-box browser would be running XP and owned several times over by this point. And then there's everything that happened in 2016. Craptacular! fucked around with this message at 20:10 on Mar 12, 2019 |
|
#
?
Mar 12, 2019 20:08
|
|
|
If anybody out there still trusts Newegg, you can get a 2700 for $230, or with some other bundle deals. Or wait, because if newegg's cutting prices to shift back stock before 3000-series comes out, other cuts elsewhere will be coming soon. edit: good to see DDR4 prices finally dropping, too. Even if it took this long to happen.  edit edit: Ebay has a 15% off your cart code today, in app only, PSAVETODAY SwissArmyDruid fucked around with this message at 23:56 on Mar 12, 2019 |
|
#
?
Mar 12, 2019 23:48
|
|
|
SwissArmyDruid posted:If anybody out there still trusts Newegg, you can get a 2700 for $230, or with some other bundle deals. Coupon excludes tech, duder.
|
|
#
?
Mar 13, 2019 00:29
|
|
|
There's enough crap on ebay that's tech-but-not-really, or falls through the cracks that I'd give it a shot anyways just to see if the discount shows up or not.
|
|
#
?
Mar 13, 2019 00:44
|
|
|
I'm seeing a lot of goons in the building thread opt for the B450 over the X470. What are the differences? Just some minor stuff?
|
|
#
?
Mar 13, 2019 16:41
|
|
|
Someone on Reddit managed to snag an open-box Ryzen 1600 for $60 at their local Microcenter.
|
|
#
?
Mar 13, 2019 17:03
|
|
|
OhFunny posted:I'm seeing a lot of goons in the building thread opt for the B450 over the X470. Relatively minor, like less USB 3.1 ports, less SATA ports, less PCIe lanes ... nothing that would really matter to someone who just plays games or does basic stuff on his PC with a single GPU and a couple of hard drives at the most. The big difference is in the VRMs (e: as said below that's more a factor of cost, rather than the chipset itself!), affecting the capability to overclock and support beefier CPUs later, so if you go for a B450 it's usually best to go for the "best" ones (MSI Tomahawk or Gaming Carbon Pro AC, or the equivalents from other manufacturers - a table was posted not long ago with a "ranking" of the various manufacturers' B450 boards) The savings can be significant, I got my MSI Carbon Gaming Pro AC for 120€ (a top tier B450 board if not the best one - it has wifi, bluetooth, and good VRM... probably enough to even go for the beefiest Zen2 CPUs once they release) and any X470, even the worst ones, would've been at least 150€ but actually closer to 200€ ... and those 50/80€ savings meant getting a 1080 instead of a 1070 
TorakFade fucked around with this message at 17:14 on Mar 13, 2019 |
|
#
?
Mar 13, 2019 17:07
|
|
|
OhFunny posted:I'm seeing a lot of goons in the building thread opt for the B450 over the X470. X470 has more PCIe lanes and supports splitting it's lanes 8x/8x. Generally they'll have better power delivery as they're more expensive, but it's not inherent to the chipset. The MSI B450 Tomahawk getting recommended in the thread was rated by buildzoid as have sufficient power delivery for a heavy overclock on a 2700X, so it's not missing out on much.
|
|
#
?
Mar 13, 2019 17:11
|
|
|
Cool. Thanks for info goons.
|
|
#
?
Mar 13, 2019 17:45
|
|
|
I recently purchased parts to build a new system and seem to be encountering boot issues. The system won't post fully (get to BIOS announcement and won't respond to DEL or F2) and I'm getting a number of Q code errors. It seems to have more issues when either 1 or both sticks of RAM are in and in wondering if 1) is this a RAM Ryzen issue and/or 2) is there anything I can do to resolve? I had read on Reddit of someone using this exact RAM after changing settings...but I dont know how to without getting into BIOS. Without RAM I get Q code error: 46 With RAM I get Q code error: 15, followed by a number of other errors. GIF of Q Code LED Mobo: Crosshair VII Hero WiFi CPU: Ryzen 5 1600 RAM: gskill ripjaws v (F4-3200C16D-32GVK) PSU: EVGA 650w G+ GPU: XFX RX 570
|
|
#
?
Mar 14, 2019 13:42
|
|
|
Are you putting the RAM in the DIMMB2 and DIMMA2 slots Edit: if you are using a non-stock cooler, has the thin plastic covering on the CPU cooler been removed yet Sidesaddle Cavalry fucked around with this message at 18:22 on Mar 14, 2019 |
|
#
?
Mar 14, 2019 18:15
|
|
|
I'm using A2 and B2 DIMM slots. I'm using the stock cooler. Someone suggested trying new memory to see if it's an issue with it being Ryzen compatible. Might not play nice until I can get timing adjusted.
|
|
#
?
Mar 14, 2019 18:41
|
|
|
It might also just be bad ram to begin with.
|
|
#
?
Mar 14, 2019 18:43
|
|
|
If I can't post or get into BIOS, is it even possible for me to load memtest86 to see if it's good?
|
|
#
?
Mar 14, 2019 18:44
|
|
|
No? Either way you're going to need new ram to test, but I haven't heard of ram that won't post b/c it didn't like the processor, vs plenty of stories of DoA ram.
|
|
#
?
Mar 14, 2019 18:48
|
|
|
gepgepgep posted:If I can't post or get into BIOS, is it even possible for me to load memtest86 to see if it's good? No. You could put the ram into a known-good system to test it if you know anyone with a DDR4-using PC. Are these parts used? If so you should make sure to clear the bios, it may be trying to use memory settings that your ram won't work with. Looking at that mobo manual, there's a "safe boot" button that boots the system with all default settings. Also there's some weird "LN2 mode" switches and jumpers, make sure those haven't been fiddled with.
|
|
#
?
Mar 14, 2019 18:55
|
|
|
Email the vendor under support (just list the facts I have X motherboard X processor and X RAM everything else is unplugged wont post) List any part swaps you did for testing, and let them know you don't have another set of equipment to test with. If not they will say "have you tried a different processor?" and when you say "I guess I could" they'll wait for that, it's fine to say "that's not feasible I'm an individual". That line is often part of their script. I have good luck getting RMAs in the first reply with that, sometimes you still need to answer some other bullshit and it takes 2-3 but it does cut down on a lot of the bullshit emails. If you have known goods that you can borrow you should do that so you can cut down on wasting your own time since this could really be any of the 3 parts. Do you get the correct error codes (beep / light /whatever) when you don't have RAM installed?
|
|
#
?
Mar 14, 2019 19:02
|
|
|
gepgepgep posted:If I can't post or get into BIOS, is it even possible for me to load memtest86 to see if it's good? If you have the 2 digit LED display thing on the motherboard you can google what code you get with "POST CODE" and figure out what the heck it thinks is wrong. If not that, then you have to rely on beep codes.
|
|
#
?
Mar 14, 2019 19:13
|
|
|
gepgepgep posted:I recently purchased parts to build a new system and seem to be encountering boot issues. The system won't post fully (get to BIOS announcement and won't respond to DEL or F2) and I'm getting a number of Q code errors. It seems to have more issues when either 1 or both sticks of RAM are in and in wondering if 1) is this a RAM Ryzen issue and/or 2) is there anything I can do to resolve? I had read on Reddit of someone using this exact RAM after changing settings...but I dont know how to without getting into BIOS. Have you tried the SafeBoot and Retry buttons at the far end (after the expansion slots) of the motherboard? I know back on my 2500K my Asus-based system wouldn't boot with the RAM it had until I hit the MemOK! button and after that it worked fine for six and a half years and counting, but that isn't a thing on those boards.
|
|
#
?
Mar 14, 2019 22:28
|
|
|
If he didn't see my reply in the hardware catch-all thread, if you watch the GIF it displays 54 - which according to the manual is "unspecified memory initialization error."
|
|
#
?
Mar 14, 2019 23:40
|
|
|
Geoj posted:If he didn't see my reply in the hardware catch-all thread, if you watch the GIF it displays 54 - which according to the manual is "unspecified memory initialization error." The code is kinda meaningless until it stops flashing through numbers like a slot machine. The important code is the one it ends on. A gif of flashing numbers is actually not very helpful. If you can't tell where it stops because it recycles into a new boot immediately, I don't think the post code is totally reliable. And that also means that something is causing the boot process to fail and reset -- which could be a minor issue like my suggestion of needing to clear the bios or don't be mean's of safeboot. Alternately, a minor cock up like not fully seating a card or ram, which I've done before. I suspect those are the source of a lot of "well I took it apart and put it back together again, now it works perfectly" type events.
|
|
#
?
Mar 15, 2019 02:29
|
|
|
|
| # ? May 29, 2024 01:01 |
|
|
Is there even a reliable unified list of POST codes?
|
|
#
?
Mar 15, 2019 15:19
|
|
