Advertisement
Guest User

Untitled

a guest
Aug 17th, 2018
775
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 21.57 KB | None | 0 0
  1. <enhanox> yeah :P
  2. <enhanox> that's still a hell of a lot of TV for the money though, compared to how much it used to cost to get like a 32" CRT or something
  3. * steve30 remembers paying £75 for a 14" CRT TV in 2004
  4. <enhanox> in 1999 (I couldn't find a 1998 Argos catalogue), a 32" widescreen Trinitron was £1250 (which is about £2060 now - or €2300/$2600), and that was the biggest model they had
  5. <steve30> I miss our old flat trinitron TV
  6. <enhanox> so to be able to get a 50"+ flat panel for £300 is mindblowingly cheap in comparison :P
  7. <steve30> I think in the early/mid 2000s, a 32" off-brand CRT was a couple of hundred.
  8. <enhanox> admittedly it's probably unfair to compare an unknown budget brand against Sony, but even a big Sony TV now doesn't cost £2000
  9. * enhanox nods
  10. <enhanox> I forget how much I paid for my Wharfedale CRT, that was a widescreen 20-something inch (or low 30s? I forget)
  11. <enhanox> but I think when I got that, LCDs were very much a thing already
  12. <enhanox> so it probably would've been "low-end" and cheap
  13. <steve30> Is Wharfedale still a thing? I don't routinely read argos catalogues any more
  14. <enhanox> dunno, this would've been mid-00s somewhen, so quite a while ago
  15. <enhanox> seems the brand was licensed to Argos until 2008, so it probably wasn't "real" Wharfedale, just the brand name glued onto some generic imported gear and sold in Argos
  16. <enhanox> wikipedia says they only made audio equipment since that point
  17. * msimpson has quit (Read error: Connection reset by peer)
  18. <enhanox> it must've been cheap though, because I wouldn't have been able to afford hundreds and hundreds on a TV, I wouldn't have thought it would've been any more than a couple of hundred
  19. <enhanox> likely closer to £100, at a guess
  20. <enhanox> but yeah, I only want monitors these days, if I had a reason to buy a larger display to hang on a living room wall or something I might possibly consider a TV (just because monitors don't tend to go that large) but I'd almost certainly just plug other stuff into it and use it as a simple screen
  21. <steve30> I've watched TV on my 14" CRT video monitor this last couple of weeks :D
  22. <enhanox> worked perfectly well in the 90s, why not now, right? :P
  23. <steve30> I think this chinese takeaway we're having for tea will nicely complement the fish and chips I had for lunch :p
  24. <Polesch> Nvidia is hyping the new RTX so much I might just have to get one.
  25. <Polesch> https://videocardz.com/newz/exclusive-msi-geforce-rtx-2080-ti-gaming-x-trio-pictured
  26. <enhanox> steve30 - both of those sound good right now :P
  27. <enhanox> Polesch - I'll be sure to start saving up my $800 or whatever they'll cost ;)
  28. <Polesch> That was optimistic :P
  29. <enhanox> it's a non-issue, I won't be getting one anyway :P
  30. <MrElendig> I might get a 20xx card next year
  31. <MrElendig> going to see what amd comes out with as a responce, if any, might skip yet another gen too though
  32. <enhanox> my 1050Ti works fine, can't do much 4K gaming with it, but any significant upgrade would be more expensive than it'd be worth for me personally, and anyway the GTX2050(Ti) isn't due out any time soon, and rumours say it won't have the advanced features of the RTX20x0 anyway
  33. <MrElendig> currently on a 970
  34. <MrElendig> but I want to move to 1440p
  35. <enhanox> for general desktop/workstation use even a mid-range card seems perfectly fine to me
  36. <enhanox> just when you start gaming at higher resolutions it all falls apart a bit
  37. <MrElendig> will probably go for 2060ti or similar
  38. <monsted> i'll probably get a 2070 quite quickly. my 970 is sick.
  39. * enhanox puts medicine in it
  40. <Polesch> The RTX sure is tempting.
  41. <Polesch> MrElendig https://www.komplett.no/kampanje/67444/nvidia-live-stream
  42. <monsted> i never had any luck with it. my first one made the system blue screen just rarely enough that they couldn't reproduce it when i tried to get it replaced. it took a year before i finally got it replaced under warranty. the new one had sticky fans and a "turns fans off when idle" feature, which results in it running one fan at 100% while the other is dead.
  43. <MrElendig> 2080 would be tempting for 4k/vr use though
  44. <Polesch> I should try selling my 1080 Ti again
  45. <monsted> i could be tempted to get a 2080 and a 1440p ultrawide hdr 100+ hz gsync monitor
  46. <monsted> i really wish they'd adopt freesync. gsync for 4k high hz monitors is like $500 :(
  47. <Polesch> Have you looked at nvidia's bank account lately?
  48. <monsted> yeah, they don't give a shit
  49. <monsted> they could be forced into it with HDMI adaptive sync
  50. <MrElendig> monsted: non-hdr is <100usd premium
  51. <MrElendig> it is the hdr one that is insane
  52. <monsted> (which is pretty much just freesync)
  53. <MrElendig> but hdr is awesome
  54. <MrElendig> sort of strange that they have not rolled an asic yet
  55. <monsted> yeah, it's rumored that the gsync module is a $2500 FPGA
  56. <monsted> (though it probably sells for more like $500 in bulk)
  57. <Polesch> Something crazy.
  58. <MrElendig> even for the non-hdr one it would probably work out, specially if they got even more market share with it
  59. <Polesch> My new monitor is beautiful, I like 4k and HDR.
  60. * MrElendig is tempted to get a new monitor due to -20% off everything at power.no
  61. <monsted> i don't really want 4k. i seriously regret it on my hackintosh
  62. <MrElendig> I want 4k, but the cards to drive it doesn't really exist yet :p
  63. <MrElendig> for newer games
  64. <Polesch> If I turn off Windows scaling, I regret it too. lol.
  65. <MrElendig> though 1080p upscaled can look quite nice too though
  66. <Polesch> I can achieve 100 fps in BF1, 4k and HDR.
  67. <Polesch> Which is far better than I had thought
  68. <Polesch> It's a very CPU-heavy game though
  69. <enhanox> this is a thing, apparently: https://sputniknews.com/viral/201808011066849989-polish-brewery-vagina-beer/
  70. <enhanox> I joked about that years ago, but... uh... somebody actually did it
  71. * enhanox chalks up a conversation murder on the scoreboard
  72. <monsted> i still think i'd enjoy 1440p/144hz more than the 4k/20hz my GPU could deliver :)
  73. <monsted> but i'll obviously have to try the 4k display for gaming if i get a better gPU
  74. <monsted> GPU
  75. <Polesch> Hmmm, now I'm confused - Intel's new HEDT line with the 28-core LGA3647 part will be on a new X599 chipset - what happens to the LGA2011 socketed X299?
  76. <Polesch> That's all becoming messy, as AMD likes to play off Intel's names
  77. <Polesch> AMD has X399, X370 and X470, and expected to release X499
  78. <monsted> i'm pretty sure that X599 thing will sell approximately zero units
  79. <Polesch> Really depends what plans they have for LGA3647 CPUs
  80. <MrElendig> I'm a bit more optimistic on intels behalf
  81. <Polesch> They already have 18 cores on X299, right?
  82. <Polesch> Will they transition all those over to the larger socket for no good reason?
  83. <MrElendig> for certain workloads and the "we have always used intel" factor
  84. <monsted> they'll have to make it ridiculously oversized and overengineered to deliver and cool the kilowatt it eats
  85. <Polesch> Yeah, it's a ridiculous CPU to sell to the consumer market
  86. <monsted> and very few people will buy that instead of a Xeon
  87. <MrElendig> it is sort of competing more against the epyc than the tr
  88. <Polesch> They want to compete with TR obviously, it's a consumer CPU
  89. <MrElendig> in some ways
  90. <Polesch> Will not have ECC memory support for sure, less memory channels, less PCIe lanes etc.
  91. <monsted> i don't see them aiming this at the epyc market
  92. <MrElendig> less issue with numa
  93. <Polesch> At least some of that.
  94. <monsted> they already have a bunch of xeons aimed at epyc
  95. <MrElendig> also xpoint ram
  96. <Polesch> If they make the 28-core compete with TR, it will require double the power.
  97. <monsted> and datacenter people will just laugh at this thing and buy one that doesn't eat a kilowatt :)
  98. <Polesch> That's the simple fact, as far as I see it.
  99. <monsted> i wouldn't automatically assume that this thing works with Xpoint NVDIMMs
  100. <Polesch> The benefits of the monolithic architecture, like memory latency, won't be obvious or very useful the consumer
  101. <Polesch> Maybe a very niche market of "deep learning" and "AI" semi-professionals
  102. <monsted> nope, they use GPUs or FPGAs
  103. <MrElendig> and 250w is not that bad compared to amd
  104. <Polesch> The 28-core Xeons that run at acceptable power usage is at 2.2 GHz base or something.
  105. <Polesch> To compete they need to push that up to 3 GHz at least
  106. <monsted> there will be a very small number of people who want high clock rates and many cores, but i wouldn't be surprised if the only people who gets this are people like linus (mr linustechtips) who gets theirs for free more or less as a joke
  107. <Polesch> https://ark.intel.com/products/120508/Intel-Xeon-Platinum-8176-Processor-38_5M-Cache-2_10-GHz
  108. <Polesch> 2.1 GHz base
  109. <monsted> it'll be much more reasonable to just get two 14 core xeons instead of one of these
  110. <Polesch> Dual-xeon machines work very well
  111. <Polesch> Whatever happens with Intel, it will be fun and interesting to experience
  112. <monsted> there's a 3 GHz 18-core xeon 6154. just get two of those.
  113. <Polesch> Damn
  114. <Polesch> Indeed
  115. <enhanox> how many limbs do they cost? :P
  116. <monsted> list price is $3543
  117. <enhanox> ouch
  118. <Polesch> 200 W Intel TDPs
  119. <Polesch> That's 300 W real ones
  120. <Polesch> Or are they not faking it on the Xeons yet?
  121. <enhanox> if I was trying to optimise a datacentre (or even a rack) I'd want to know the /real/ power consumption and heat production
  122. <monsted> i wouldn't think so, since they have to conform to the cooling properties of whatever they go into
  123. <monsted> HP increased the size of their blades (in the Synergy 12000 enclosure) to make room for 300W TDP chips
  124. <enhanox> wouldn't be very efficient to have to design enterprise computing around "well, Intel says it's 200W, but they're probably talking out of their arses, so let's guesstimate 350W worth of cooling just to be on the safe side"
  125. <monsted> well, that's how it usually goes in the real world :)
  126. <enhanox> but it shouldn't, surely?
  127. <enhanox> with specs that important, in that environment, you'd hope for /actual/ numbers, not just some marketing figure someone imagines might sound vaguely acceptable on the spec sheet
  128. <monsted> "ok, it's 42U, so it'll take 21 2U systems with two 150W CPUs. We'll double that to be safe." *hands off to the facilities people* "OK, they wanted 6 kW per rack. We'd better double that to be safe."
  129. <enhanox> sure, naturally you build in margins, but if the numbers you're working with are totally bogus you'd be massively overspeccing for no good reason
  130. <enhanox> or you could potentially be, I should say
  131. <monsted> you should've seen the IBM sales droid when he saw our new datacenters with 125ish racks. he was obviously doing the mental math of how many RS/6000 BS50's he could sell to fill it up :)
  132. <monsted> er, B50
  133. <enhanox> you might've been right the first time ;)
  134. * firewalker has quit (Ping timeout)
  135. <monsted> our last two datacenters got 21 kW per rack
  136. <monsted> when i left, they were both full. and consumed 5 kW per rack.
  137. <enhanox> I suppose it does make some sense to overspec, both for safety and in case anything more powerful goes in there in the future
  138. <enhanox> but I'd still want proper numbers :P
  139. <monsted> this got overspec'ed because my boss was an idiot
  140. <monsted> he bought three complete datacenters without talking to anyone, so everything was wrong.
  141. <enhanox> better that he overspec'd rather than underspec'd everything I guess? :P
  142. <monsted> sure. but i could've used the million dollars he overspent for a lot of other things.
  143. * enhanox nods
  144. <monsted> he put in like $200k worth of fiber trunks. i used two of them.
  145. <enhanox> yay future expansion? :P
  146. <monsted> (and had to install extra of the ones i actually needed, while ignoring the mountain of stuff i didn't)
  147. <monsted> not helped by the fact that the product he bought was horrendously shitty
  148. * Okie2 (OK@29767D:974F93:88CD3B:7F08C7:IP) has joined
  149. <monsted> we'd decided not to use multimode fiber in the new datacenters. so obviously he bought 76 multimode fiber trunks
  150. <enhanox> oh dear
  151. <enhanox> he does sound a bit... unhelpful :P
  152. <monsted> (and while my back was turned, my manager decided to call in people from our other location to finish my work - so they took my unfinished cable plans and installed the wrong fiber type all around)
  153. <monsted> i took a day off and when i came back, they'd stuck multimode fiber in all of my singlemode optics... which does work for short distances, but...
  154. * Okie3 has quit (Ping timeout)
  155. * Okie3 (OK@93A78:DF276F:88CD3B:7F08C7:IP) has joined
  156. <monsted> i found it hilarious that i had swapped two firewalls, so we had one rack full of devices named *A and a single *B firewall, another rack full of *B stuff and a single *A firewall. and they then cable everything correctly from the *B firewall into the gear in the other rack, instead of swapping the firewalls.
  157. * Okie2 has quit (Ping timeout)
  158. * Okie (OK@E7F570:F89065:88CD3B:7F08C7:IP) has joined
  159. <monsted> then i called in sick with stress and when i came back 15 months later as a consultant, an old colleague greeted me with "oh, you're back? then we won't have to wait six months to have stuff done wrong!"
  160. * Okie3 has quit (Ping timeout)
  161. * SopaXorzTaker has quit (Success)
  162. * X3 sets channel limit to 116
  163. <mattinx> lol
  164. <mattinx> sounds about how it goes
  165. <Polesch> TR2 reviews are painful to watch and read somtimes.
  166. <enhanox> "I mean, it's ok, I guess, if you like that sort of thing, but why the hell would you buy an AMD chip when Intel runs games 3fps faster? what a waste of money!"
  167. <Polesch> "Here are 30 benchmarks we ran with it, does good in some and not others. Take from that what you want."
  168. <enhanox> oh, yeah, those
  169. <enhanox> realistically I dunno what else they can say though
  170. <Polesch> They don't use benchmarks to show or prove anything.
  171. <enhanox> hm, maybe it gives readers/viewers a chance to exercise their analytical skills? :P
  172. <enhanox> I do get a bit irritated by the videos though, "hi, we benchmarked a chip, here's a graph, here's another graph, here's another graph, here's another graph - so as you can see, it's kinda like other chips, bye!"
  173. <enhanox> what's the actual point? I could read those graphs on a page that doesn't go away 3 seconds later
  174. <Polesch> If you don't have a purpose with the benchmark, all it shows is how good the CPU is at running that benchmark. Nobody uses or buys CPUs to run that benchmark.
  175. <enhanox> I dunno, a few people do :P
  176. <enhanox> "my computer has more TesticleMarks than yours!"
  177. <enhanox> though game benchmarks, productivity macro benchmarks, and stuff like Cinebench might be genuinely useful
  178. <Polesch> Yeah, they are - if you as a reviewer explains what it means or shows
  179. <enhanox> true
  180. <Polesch> A lot of the benchmarks I see used on the 32-core chip has a thread limit, one is 10 threads max, another is 12. So the results are worse than other CPUs, they still just throw the numbers at you
  181. <enhanox> yeah, that's definitely something a reviewer should mention
  182. <enhanox> pretty important detail that the numbers alone don't show
  183. <Polesch> And in several benchmarks the actualy CPU usage is very sporadic throughout, so it's impossible to tell what it means
  184. <enhanox> (well, they might, but you'd have to analyse them)
  185. <enhanox> really depends who your target audience are though
  186. <enhanox> like, if you're talking to a bunch of gamers, they don't really care that 75% of the cores aren't going to do anything in a game, they only care that the Intel chip is faster because it does more frames
  187. <enhanox> "does this chip make Tomb Raider run faster?" "yes" "then I'll buy that one"
  188. <enhanox> LTT, for example, they do some stuff that's related to enterprise-level stuff sometimes, but loads of their reviews are "ok, but how does it play games?"
  189. <enhanox> they could do more useful stuff though, like all the projects they've done with virtualisation, they don't really talk much about how having all those cores really helps with that, they just sort of gloss over it and try to get 10 people playing games on the same server
  190. <mattinx> http://pkl.net/~matt/secondpersoninthequeue.wav
  191. <mattinx> WTF is with the inflection at the end of "queue"?
  192. <enhanox> in the /queue?!/
  193. <Polesch> I have to stop reading and getting agitated.
  194. <Polesch> Time for bed.
  195. <enhanox> later
  196. <mattinx> enhanox: quite
  197. * ImQ009 has quit (Quit: Leaving)
  198. * X3 sets channel limit to 115
  199. <Jobe> https://www.bbc.co.uk/news/uk-england-merseyside-45225045
  200. <Jobe> well the council aren't wrong
  201. <Jobe> driving on the pavement is an offence under the highways act
  202. <T3sl4co1l> oh my god, sending huge attachments through outlook is incredibly slow
  203. <T3sl4co1l> like 10 minutes to send 15 megs
  204. <T3sl4co1l> but, doing it via outlook web app
  205. <T3sl4co1l> mere seconds
  206. <ServalSpots> The second person in any queue garners at least some incredulity. After all, if it weren't for them there wouldn't a queue to begin with
  207. * SparkGap (sparkgap@AN-3C14A070.static.zebra.lt) has joined
  208. * arch (arch@AN-23E7680B.dyn.optonline.net) has joined
  209. * SparkGap has quit (EOF from client)
  210. * SparkGap (sparkgap@AN-3C14A070.static.zebra.lt) has joined
  211. * bozi has quit (Quit)
  212. * SparkGap has quit (EOF from client)
  213. <T3sl4co1l> ten minutes to pizza :3
  214. * jspiros_ has quit (Ping timeout)
  215. * X3 sets channel limit to 114
  216. <Okie> Is it delivery?
  217. <T3sl4co1l> it's diT3sl4coilo
  218. <Okie> The transistors might be a bit crunchy if you leave it in there too long
  219. * DoctorMirabilis has quit (Quit)
  220. * X3 has changed the topic to: Welcome to the #eevgob! FOTD: pepperoni pizza croissant roll <https://i.imgur.com/75PjuWk.jpg> | Stats: http://goo.gl/7GtYOx | Official soundtrack: https://youtu.be/3NpEqnBkF0w (jspiros)
  221. <T3sl4co1l> D:
  222. * X3 sets channel limit to 113
  223. <T3sl4co1l> my brain says proceed, but my heart, and my arteries, say no!
  224. <jspiros> flipping the script, I see
  225. <T3sl4co1l> ¿ʇdᴉɹɔs ǝɥʇ ƃuᴉddᴉlɟ
  226. <T3sl4co1l> (apologies to anyones's chat client that's not unicode supporting)
  227. <Jobe> T3sl4co1l: dont apologise to those heathens
  228. <Jobe> they need to fix their shit
  229. * T3sl4co1l nods
  230. <Jobe> hell even irssi in GNU screen via ssh in PuTTY can see it fine :D
  231. <T3sl4co1l> hmm, yes, i recognize these letters
  232. <ServalSpots> I hope I will not be attacked with bludgeons, but I have some reservations about the pie-ssant
  233. <T3sl4co1l> don't get croiss with me
  234. * SparkGap (sparkgap@AN-3C14A070.static.zebra.lt) has joined
  235. <jspiros> ServalSpots: I do too, to be fair
  236. <jspiros> for me, sauce and crust /are/ the pizza
  237. <jspiros> so, as a pizza, it's a travesty
  238. <jspiros> and, as a croissant, it's also a travesty
  239. <jspiros> but as a random frankenstein of pepperoni bread, cheesy bread, and a croissant...
  240. <T3sl4co1l> you're a travestite
  241. <jspiros> I'd surely eat a wedge
  242. <jspiros> not sure if it'd be very good or not, though
  243. <jspiros> I would be much chuffed if it came with a bowl of pizza-type sauce
  244. * T3sl4co1l marinarates jspiros
  245. <T3sl4co1l> (a traditional weekend bath for him, i think)
  246. <jspiros> well, you know, people call that marinara, but often pizza places have multiple types of it
  247. <jspiros> so I thought "pizza-type sauce" would more clearly define what I wanted
  248. <T3sl4co1l> marinara per se isn't very interesting to me, annoying even
  249. * SparkGap has quit (Ping timeout)
  250. <T3sl4co1l> but "pizza-type-sauceinates" doesn't pun
  251. <jspiros> indeed
  252. <mattinx> you need additional things
  253. * sku1d has quit (Quit: -)
  254. * X3 sets channel limit to 112
  255. * sku1d (sku1d@D412DC.8680E0.44C7B1.1F4DFD.IP) has joined
  256. * jspiros sprinkles mattinx with some basil and oregano
  257. <jspiros> operating on the principle of "if you herb it, tomatoes will come"
  258. <T3sl4co1l> mattinx: you must build additional pizzalons
  259. <jspiros> construct!
  260. <T3sl4co1l> *
  261. <ServalSpots> jspiros: It's one of those things that I can't imagine I would ever get, but that I would certainly try a bite of. I am assuming the dough is something more sensible than regular croissant pastry, and that they've foregone the butter layers completely
  262. <T3sl4co1l> idunno, it looks pretty fluffy
  263. <ServalSpots> Well rested and lots of gluton?
  264. <T3sl4co1l> you're a gluten for PUNishment
  265. <ServalSpots> Yes, that's it... I totally know how to spell, though >.>
  266. <jspiros> it actually looks like phyllo to me
  267. <ServalSpots> Plausibly so, and the olive oil would make sense in this application (over something like butter, anyway)
  268. <mattinx> jspiros: there's fresh basil and oregano in the planter out back
  269. <mattinx> and tomatoes on the counter
  270. <jspiros> \o/
  271. <T3sl4co1l> tomatoes are a firm counter to hunger and lack of flavor
  272. <jspiros> at some point, I fully expect to have a greenhouse in my life
  273. <mattinx> there's also a log of goat cheese in the fridge
  274. <jspiros> "An issue affecting the physical host this Linode resides on has been detected." :(
  275. <mattinx> time to replace it with a small piece of cheese
  276. <T3sl4co1l> examine the goat log
  277. <mattinx> You peer at the cheese, it appears to be sealed in plastic and dated late september
  278. <T3sl4co1l> use knife on log
  279. <mattinx> You try, but the knife just presses into the plastic, squashing the cheese
  280. <T3sl4co1l> sharpen knife
  281. <mattinx> Sorry?
  282. <T3sl4co1l> use sharpening stone on knife
  283. <mattinx> You don't have that
  284. <T3sl4co1l> >_>
  285. <T3sl4co1l> open fridge door
  286. <mattinx> It's already open
  287. <T3sl4co1l> leave kitchen in frustration
  288. <T3sl4co1l> ha, now your electricity bill goes up! :P
  289. <mattinx> As you leave the kitchen, you hear a beeping coming from the fridge. You are now in the hallway. It is dark.
  290. <T3sl4co1l> walk east
  291. <mattinx> You have been eaten by a grue
  292. <T3sl4co1l> :D
  293. * Laogeobunny nibbles on the grue
  294. <Jobe> dont eat glue
  295. <T3sl4co1l> it's a japanese grue?
  296. * jspiros_ (jspiros@jspiros.Users.AfterNET.Org) has joined
  297. * jspiros_ has quit (Ping timeout)
  298. * arch has quit (Success)
  299. * jspiros has quit (Ping timeout)
  300. * X3 sets channel limit to 111
  301. * jspiros (jspiros@jspiros.Users.AfterNET.Org) has joined
  302. <jspiros> fuck, the power just went out
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement