|
eames posted:Maybe it eventually will... That would be FANTASTIC for AMD.... ....just as soon as they completely un-gently caress their APU graphics driver stack.
|
#
?
Nov 30, 2018 21:43
|
|
|
|
| # ? May 15, 2024 09:06 |
|
|
SwissArmyDruid posted:That would be FANTASTIC for AMD.... is it the apu driver stack or the way 3rd parties override amd on drivers?
|
|
#
?
Nov 30, 2018 23:18
|
|
|
Palladium posted:I wondering just how big is that "I'm dirt poor but I also insist on not buying refurb/used" segment the 200GE is aimed at. I'm totally fine with buying used parts, but I want AM4 and DDR4 RAM which is not very common on the used market, the 200ge would lessen the blow of a platfrom upgrade and allow me to get a better CPU later. And the current fastest CPU I own is a Phenom II 840- a 10yr old 3.2GHz quad core. I think the 200ge might be faster?
|
|
#
?
Dec 1, 2018 01:50
|
|
|
wargames posted:is it the apu driver stack or the way 3rd parties override amd on drivers? I tried to write this post and started over three times before giving up. SwissArmyDruid posted:I found a Hardware Unboxed video talking about it. Just the front half is about it, the rest can be skipped. I am personally invested in AMD unfucking their poo poo, because my brother purchased one of those HP Envy x360s on my recommendation.
|
|
#
?
Dec 1, 2018 03:42
|
|
|
Crotch Fruit posted:As a kid haver devoting all my spending to their toys, I consider myself in the "I'm dirt poor" segment Were you being sarcastic? https://cpu.userbenchmark.com/Compare/AMD-Phenom-II-X4-840-vs-AMD-Athlon-200GE/m3302vsm592714
|
|
#
?
Dec 1, 2018 04:51
|
|
|
Been trying to figure out if 2700x under full load at 69C is too hot, or acceptable. I can push it a little lower if I crank the cooling system to max.
|
|
#
?
Dec 1, 2018 14:28
|
|
|
GreenBuckanneer posted:Been trying to figure out if 2700x under full load at 69C is too hot, or acceptable. I can push it a little lower if I crank the cooling system to max. Artificial full load via something like Prime95? Or an actual all core workload? For the former 69 is probably OK, for the later it's probably on the border of being OK.
|
|
#
?
Dec 1, 2018 14:40
|
|
|
69c is perfectly fine. The max temp before thermal throttling is 85c. Then if it continues to heat up it just turns itself off. It’s not the old days where CPUs cook themselves.
|
|
#
?
Dec 1, 2018 15:31
|
|
|
Stanley Pain posted:Artificial full load via something like Prime95? Or an actual all core workload? For the former 69 is probably OK, for the later it's probably on the border of being OK. Like, running folding@home on max.
|
|
#
?
Dec 1, 2018 15:33
|
|
|
GreenBuckanneer posted:Like, running folding@home on max. Pretty sure the FAH client supports AVX so that's pretty normal temperature wise.
|
|
#
?
Dec 1, 2018 17:28
|
|
|
Stanley Pain posted:Pretty sure the FAH client supports AVX so that's pretty normal temperature wise. Cool, it's been quite some time since I've had an AMD CPU so I wasn't sure if that was normal or not.
|
|
#
?
Dec 1, 2018 18:15
|
|
|
AMD roadmap leak for next year. X570 will supposedly launch at Computex (May 28 - June 1) which I imagine they would launch with the Matisse/Zen 2 parts as a way to officially put PCIe4 on the desktop. Although maybe Zen2 could launch later if AMD needed.
|
|
#
?
Dec 1, 2018 20:56
|
|
|
.
sincx fucked around with this message at 05:50 on Mar 23, 2021 |
|
#
?
Dec 1, 2018 21:31
|
|
|
Probably not until 2H2019
|
|
#
?
Dec 1, 2018 21:41
|
|
|
I can live with that. Or rather, my wallet can. If it comes with a decent speed and IPC bump, it'll be my last CPU upgrade for a while. Zen 2 will probably bring the last worthwhile performance bump for the next few years.
|
|
#
?
Dec 2, 2018 01:23
|
|
|
TR Zen2 might even be first half of 2020, god knows what the demand for Rome might be.
|
|
#
?
Dec 2, 2018 02:28
|
|
|
They use lesser dies right? The io die looks to be what would be crippled or harvested. TR has always sort of been a yield improvement program Hmmm I really need to upgrade ol 5820k since it's falling behind in compilation tasks but I would really like more than a 2c/4t jump but I'm limited by mini itx :/
|
|
#
?
Dec 2, 2018 02:51
|
|
|
The tiny rear end core dies are likely to have great yields which is why amd can be so aggressive on price
|
|
#
?
Dec 2, 2018 02:51
|
|
|
Malcolm XML posted:They use lesser dies right? The io die looks to be what would be crippled or harvested. TR has always sort of been a yield improvement program The opposite, actually, TR is harvested from the top 5% of Zen dies. This leads to funny things like the 2950x having a higher max turbo than the 2700x, despite having twice as many cores. I guess they don't bin the 2700x as much as they did the 1800x, even when there's no 2800x.
|
|
#
?
Dec 2, 2018 04:06
|
|
|
Finally got around to ripping out the nvme drive and just putting my 850 Pro SATA drive in as the main drive and so far everything seems good. I'll troubleshoot the nvme drive later- or I'll just sell it. I'm happy as hell with my 2700x so far. It's a huge step up from my 3770k that served me well for like 6 years or so.
|
|
#
?
Dec 2, 2018 05:20
|
|
|
Combat Pretzel posted:I can live with that. Or rather, my wallet can. If it comes with a decent speed and IPC bump, it'll be my last CPU upgrade for a while. Zen 2 will probably bring the last worthwhile performance bump for the next few years. Are you kidding? It's the first product on a new process, there's miles and miles of improvement to go.
|
|
#
?
Dec 2, 2018 07:48
|
|
|
Dr. Fishopolis posted:Are you kidding? It's the first product on a new process, there's miles and miles of improvement to go. Tick tock tock on the new process will be interesting, especially since it's a quad patterning vs. true EUV, which should be coming like H2 2019. TR 2/2+/3 is gonna be REALLY interesting from a generational improvements perspective.
|
|
#
?
Dec 2, 2018 08:21
|
|
|
Malcolm XML posted:They use lesser dies right? The io die looks to be what would be crippled or harvested. TR has always sort of been a yield improvement program Nah, the highest-end Threadripper SKUs used the absolute best dies. They've always correspondingly clocked higher out of the box than the consumer Ryzen CPUs. 1950X XFR 4.2, 1800X XFR 4 2950X XFR 4.4, 2700X XFR 4.3
|
|
#
?
Dec 2, 2018 10:43
|
|
|
Malcolm XML posted:TR has always sort of been a yield improvement program Here's the story behind Threadripper, I think I posted it a while back. Very cool stuff imo. https://www.forbes.com/sites/antonyleather/2017/09/05/amd-ryzen-threadripper-the-fascinating-story-behind-the-processor-that-beat-intel/
|
|
#
?
Dec 2, 2018 12:39
|
|
|
Desuwa posted:The opposite, actually, TR is harvested from the top 5% of Zen dies. This leads to funny things like the 2950x having a higher max turbo than the 2700x, despite having twice as many cores. I guess they don't bin the 2700x as much as they did the 1800x, even when there's no 2800x. Lesser dies than epyc I mean
|
|
#
?
Dec 2, 2018 19:33
|
|
|
eames posted:Here's the story behind Threadripper, I think I posted it a while back. Very cool stuff imo. Interesting. I was under the impression they basically took epyc products that didn't meet the cut and sold em as TR like intel does with xeons and HEDT
|
|
#
?
Dec 2, 2018 19:34
|
|
|
LRADIKAL posted:Were you being sarcastic?
|
|
#
?
Dec 2, 2018 20:11
|
|
|
For just a little bit more ~30 to 40 more $ you can get a 2200G and then the comparison is comical. https://cpu.userbenchmark.com/Compare/AMD-Phenom-II-X4-840-vs-AMD-Ryzen-3-2200G/m3302vsm441832
|
|
#
?
Dec 2, 2018 20:43
|
|
|
Crotch Fruit posted:I had no idea my CPU was that new, but I am genuinely surprised a dual core 3.2GHz could perform as fast or faster than my 3.2GHz quad core. Is the dual core faster at core intensive take like Pretty sure the Phenom II X4 840 came out circa 2011-2012. Because that's most certainly when I got mine, paired with an 880GM motherboard. It transitioned to full time Plex server duty in late 2013, and I finally got around to replacing it this year with a handful of server upgrades.
|
|
#
?
Dec 2, 2018 22:30
|
|
|
Dr. Fishopolis posted:Are you kidding? It's the first product on a new process, there's miles and miles of improvement to go. I think trying to predict several years ahead in this game is just foolish in general, no way would I have predicted $130 8-core CPUs just a couple years back.
|
|
#
?
Dec 2, 2018 23:33
|
|
|
teagone posted:Pretty sure the Phenom II X4 840 came out circa 2011-2012. Because that's most certainly when I got mine, paired with an 880GM motherboard. It transitioned to full time Plex server duty in late 2013, and I finally got around to replacing it this year with a handful of server upgrades.
|
|
#
?
Dec 3, 2018 02:30
|
|
|
Crotch Fruit posted:I had no idea my CPU was that new, but I am genuinely surprised a dual core 3.2GHz could perform as fast or faster than my 3.2GHz quad core. Is the dual core faster at core intensive take like The page says </=2014, so I think they don't do any dates before that.
|
|
#
?
Dec 3, 2018 03:15
|
|
|
Crotch Fruit posted:CPU world says it was released January 2011, it's kinda amazing and depressing that my CPU is so old. It was a great budget gaming CPU option back then imo, mine lasted a good long while in my gaming PC before I jumped to Intel when Haswell was released. Looking back at my email receipts, I picked up my Phenom II X4 840 back in March of 2011 from Micro Center for $100. ~7 years later I got a Ryzen 3 2200G for $80 from the same Micro Center to replace it. It had a good run 
teagone fucked around with this message at 05:56 on Dec 3, 2018 |
|
#
?
Dec 3, 2018 05:53
|
|
|
If this is real   That 20CU 8C/16T looks loving bonkers for some VFIO stuff. I desperately want that 3920X. 32C/64T at 5GHz inspires some deeply horny thoughts.
|
|
#
?
Dec 3, 2018 06:03
|
|
|
TIL AMD still has products named Duron in their stack. Duron. loving DURON. Also, that 3600GX, tho. 8c/16t AND an RX 565 stapled to it? SwissArmyDruid fucked around with this message at 06:09 on Dec 3, 2018 |
|
#
?
Dec 3, 2018 06:06
|
|
|
Yeah, Deneb rocked. I had a three core part with the 4th core BIOS unlocked (which is hilarious that was even a thing) that was my primary from ‘09 to 2016. Def up there with the orig 486 DX for the longest run I’ve gotten out of a CPU. E: ^^^ I prolly built 50+ systems using Socket A Dirt-rons. Great budget chia with cheap DDR mobos. The later Applebred Durons could be turned into full T-bred Athlons with a pencil, lol Cygni fucked around with this message at 06:11 on Dec 3, 2018 |
|
#
?
Dec 3, 2018 06:08
|
|
|
SwissArmyDruid posted:TIL AMD still has products named Duron in their stack. Yeah, I don't even know what to think of the use case for that besides entry level gaming and awesome productivity. I'm fully into it.
|
|
#
?
Dec 3, 2018 06:11
|
|
|
Gonna rock that 3970wx Edit: For halarity sake I think they should make a TR with 64 cores and a Vega64 on it as well. Call it the TR64^2 or some poo poo. FlapYoJacks fucked around with this message at 06:34 on Dec 3, 2018 |
|
#
?
Dec 3, 2018 06:24
|
|
|
NewFatMike posted:If this is real calling bullshit because it's overall crazy overhype and also that 8c / 20CU is both dumb (from a market standpoint*) and improbable. you realize that if they do zen2 with integrated GPU, it would either need to be a custom respin or put the iGPU on the I/O chip? If it is a respin I doubt it would be on their list yet. If it were on the I/O that means 14nm, so cramming a RX56½ and the zen2 chip in one package seems difficult. *exceptions: it's related work to the chip that goes in the next consoles, or apple asked AMD to make a chip for their next imac. edit: also, "estimated non-exotic overclock" yeah right that seems like a completely normal heading on a OEM product list Klyith fucked around with this message at 06:35 on Dec 3, 2018 |
|
#
?
Dec 3, 2018 06:33
|
|
|
|
| # ? May 15, 2024 09:06 |
|
|
NewFatMike posted:If this is real Icept posted:Anyone think AMD is going to follow Intels footsteps and do a Ryzen 9 this next generation? Pending actual confirmation...
|
|
#
?
Dec 3, 2018 08:01
|
|