|
Sinestro posted:The issue with IF as implemented in Zen 1 is that the memory controller also handles generating the clock for the bus, and that clock is at the same rate as the one that goes out to the DDR4 PHY. So in the most pessimistic case of DDR4-2133, it's limited to just 68GB/s, and in the perfect case of DDR4-3200 you're still only getting 102GB/s. Hopefully future revisions will fix that... frankly inexplicable design choice. Probably was this works and we can put it out without delays or costing us too much just make it happen.
|
#
?
Jul 10, 2017 20:10
|
|
|
|
| # ? Apr 20, 2024 02:21 |
|
|
wargames posted:Probably was this works and we can put it out without delays or costing us too much just make it happen. That and an async clockgen for it would complicate the poo poo out of a lot of things, complications they desperately were trying to avoid at all costs. Now that Zen1 is the poo poo, they can work on teasing out all that performance they left on the table, you know, the stuff in the big box labeled 'untested, do not touch'.
|
|
#
?
Jul 10, 2017 21:14
|
|
|
EDIT: Wrong thread, sorry
L33t_Kefka fucked around with this message at 01:23 on Jul 11, 2017 |
|
#
?
Jul 11, 2017 01:07
|
|
|
Yeah, I greatly overstated it, because there is an obvious explanation that, indeed, it does work well enough and doesn't involve anything with asynchronous clock domains or any such dark wizardry that would be hell to deal with. I'm not an expert on actual hardware implementation by any means but I know enough that it'd be a nightmare. I don't think that having it run at the full 8 GT/s would be feasible at all and definitely wouldn't be worth the effort when they're trying to get Zen 1 out the door, but it wouldn't have been that hard with a multi-master bus to run it at double the RAM speed, kept in phase, which would double the inter-core bandwidth. It wouldn't be an unimaginable surprise if something like that happened even with the same silicon on the MCM parts, because trying to have 32 cores squabbling over what in the case of a server is almost certainly 68GB/s bandwidth doesn't seem like it's a recipe for success.
|
|
#
?
Jul 11, 2017 01:59
|
|
|
I'm assuming that a R7 1700 for $269.99 is a good deal?
|
|
#
?
Jul 11, 2017 18:26
|
|
|
SourKraut posted:I'm assuming that a R7 1700 for $269.99 is a good deal? I mean, it's pretty loving unbeatable TBH, it's only like 10% slower to a 7700k max clock to max clock in gaming but it'll curbstomp it in productivity, and for productivity the R5's are a noticeable step down. If you are interested exclusively in gaming, 7700k still has better performance, R5 1600 is the better price/perf buy.
|
|
#
?
Jul 11, 2017 18:42
|
|
|
FaustianQ posted:I mean, it's pretty loving unbeatable TBH, it's only like 10% slower to a 7700k max clock to max clock in gaming but it'll curbstomp it in productivity, and for productivity the R5's are a noticeable step down. 1600 is not better better price/perf if you can snag a 1700 for 269.
|
|
#
?
Jul 11, 2017 18:57
|
|
|
wargames posted:1600 is not better better price/perf if you can snag a 1700 for 269. From the benchmarks I've seen for gaming, the 6 core and 8 core Ryzens exist within margin of error of each other and the 1600 currently is going for 199$, that's a 70$ difference you could throw around elsewhere.
|
|
#
?
Jul 11, 2017 19:01
|
|
|
AMD Ryzen 7 1700X Processor - 299 on amazon prime day. It looks to be the lowest ever.
|
|
#
?
Jul 11, 2017 21:03
|
|
|
incoherent posted:AMD Ryzen 7 1700X Processor - 299 on amazon prime day. It looks to be the lowest ever. Fuuuuuck. That's really good. It's too bad I just bought a new car 
|
|
#
?
Jul 11, 2017 21:18
|
|
|
edit, wrong thread
inkwell fucked around with this message at 23:02 on Jul 11, 2017 |
|
#
?
Jul 11, 2017 22:45
|
|
|
inkwell posted:look what i found in my local craigslist. 200 dollar 8 gig rx480 rx480s are infamous mining cards fwiw.
|
|
#
?
Jul 11, 2017 22:50
|
|
|
https://www.reddit.com/r/Amd/comments/6mndgu/sizing_up_servers_intels_skylakesp_xeon_versus/ http://www.anandtech.com/show/11544/intel-skylake-ep-vs-amd-epyc-7000-cpu-battle-of-the-decade TL;DR AMD's EPYC can struggle with very small database functions compared to Skylake-SP, and lags behind in a few instances while taking the lead in others, costs half as much and runs under it's listed TDP. This test suite was comparing the 7601 from AMD and the 8176 from Intel. I'm going to go with "it'd be really hard for me to be convinced to buy Intel" for 4200$ Alex. quote:However AMD's newest core is a formidable opponent. Scalar floating point operations are clearly faster on the AMD core, and integer performance is – at the same clock – on par with Intel's best. EmpyreanFlux fucked around with this message at 00:41 on Jul 12, 2017 |
|
#
?
Jul 12, 2017 00:37
|
|
|
FaustianQ posted:https://www.reddit.com/r/Amd/comments/6mndgu/sizing_up_servers_intels_skylakesp_xeon_versus/ "The poor showing in the database test is actually due to NUMA latency, because AnandTech conducted this test in a manner that specifically presents the worst possible case for Epyc and is not even remotely representative of an actual database server"
|
|
#
?
Jul 12, 2017 00:42
|
|
|
Scarecow posted:"The poor showing in the database test is actually due to NUMA latency, because AnandTech conducted this test in a manner that specifically presents the worst possible case for Epyc and is not even remotely representative of an actual database server" Yeah, apparently the database test was designed to fit in L3? It completely eliminated EPYC's memory bandwidth advantage, but it should be representative of peak performance, no?
|
|
#
?
Jul 12, 2017 00:48
|
|
|
No as soon as it gets over a few meg it switches back to epycs favor (and what data base these days is only a few meg and can dit into l3 cache?
|
|
#
?
Jul 12, 2017 00:56
|
|
|
incoherent posted:AMD Ryzen 7 1700X Processor - 299 on amazon prime day. It looks to be the lowest ever. This is the same price its been at Microcenter for a while. Plus they kick in a $50 motherboard credit. Thats if you have a local microcenter though.
|
|
#
?
Jul 12, 2017 01:27
|
|
|
Cygni posted:This is the same price its been at Microcenter for a while. Plus they kick in a $50 motherboard credit. Thats if you have a local microcenter though. Do they still? It's not showing on my local Microcenter anymore. I thought they rolled back some (perhaps all?) of the discounts. I'm actually really tempted to do a Ryzen build for a work/occasional gaming PC for my fiance and I was disappointed to see it gone.
|
|
#
?
Jul 12, 2017 02:02
|
|
|
Paul MaudDib posted:Do they still? It's not showing on my local Microcenter anymore. I thought they rolled back some (perhaps all?) of the discounts. Still showing as available at $299 with the $50 motherboard kickback at the Microcenter in the LA area, with a couple board choices in stock. Dunno if thats around anywhere else though. Kinda kicking myself for not getting one when it was $299+$100 kickback a few weeks ago.
|
|
#
?
Jul 12, 2017 02:12
|
|
|
Cygni posted:Still showing as available at $299 with the $50 motherboard kickback at the Microcenter in the LA area, with a couple board choices in stock. Dunno if thats around anywhere else though. Kinda kicking myself for not getting one when it was $299+$100 kickback a few weeks ago. It doesn't show but it works when I put it in the cart. Thanks.
|
|
#
?
Jul 12, 2017 02:14
|
|
|
The microcenter near me still shows the $50 off on the whole line. I think my base build set me back about $350. Edit: thats what I get for not refreshing.
|
|
#
?
Jul 12, 2017 02:23
|
|
|
FWIW some people were able to sell the 1700 boxed cooler for $40 because it works well on the 1600X and some people like LED backlit AMD logos. 
|
|
#
?
Jul 12, 2017 12:11
|
|
|
I'm still perplexed on why Cooler Master doesn't sell thier own versions of the AMD stock coolers, each one is pretty drat fantastic and the Wraith Max is not only low profile but pretty drat competitive with good tower coolers. Don't get me wrong, I love my Noctua NH-U14S but a Wraith Max or Spire is going to fit into way more form factors.
|
|
#
?
Jul 12, 2017 12:42
|
|
|
Paul MaudDib posted:It doesn't show but it works when I put it in the cart. Thanks. What does http://www.microcenter.com/site/products/amd_bundles.aspx show you?
|
|
#
?
Jul 12, 2017 14:49
|
|
|
Welcome to EPOXYC
|
|
#
?
Jul 12, 2017 16:28
|
|
|
That's the most childish bullshit. Really hope the big data centres all tell intel to go gently caress themselves.
|
|
#
?
Jul 12, 2017 16:38
|
|
|
to be honest i think Intel is becoming really worried because they have no response to Zen+ except for their current CPUs +10% clockspeed +25% power consumption 
|
|
#
?
Jul 12, 2017 16:38
|
|
|
eames posted:to be honest i think Intel is becoming really worried because they have no response to Zen+ except for their current CPUs +10% clockspeed +25% power consumption Well, or moderate discounting. That would also solve the problem completely.
|
|
#
?
Jul 12, 2017 16:41
|
|
|
Second Sun posted:That's the most childish bullshit. Really hope the big data centres all tell intel to go gently caress themselves. lmao gettin worked up about a tiny rear end joke in marketing slides
|
|
#
?
Jul 12, 2017 16:50
|
|
|
Twerk from Home posted:Well, or moderate discounting. That would also solve the problem completely. Gotta make back all that wasted mobile money
|
|
#
?
Jul 12, 2017 16:55
|
|
|
That's a lot of notmad TPU. I think X399 shot their dog.
|
|
#
?
Jul 12, 2017 17:02
|
|
|
Second Sun posted:That's the most childish bullshit. Welcome to sales and marketing.
|
|
#
?
Jul 12, 2017 17:02
|
|
|
Twerk from Home posted:Well, or moderate discounting. That would also solve the problem completely. Then they'd have to stick to those prices, at least for a time. Intel loves their giant margins. GRINDCORE MEGGIDO fucked around with this message at 17:11 on Jul 12, 2017 |
|
#
?
Jul 12, 2017 17:07
|
|
|
PerrineClostermann posted:Gotta make back all that wasted *every attempt at diversifying the company that isn't storage* money
|
|
#
?
Jul 13, 2017 00:43
|
|
|
Cygni posted:lmao gettin worked up about a tiny rear end joke in marketing slides Poor Volta.
|
|
#
?
Jul 13, 2017 00:44
|
|
|
new price and specs for the glued together HEDT CPUs Threadripper 1950X 16C/32T 3.4/4.0 Ghz 999 USD Threadripper 1920X 12C/24T 3.5/4.0 Ghz 799 USD https://videocardz.com/newz/amd-ryzen-threadripper-1950x-to-cost-999-usd I'd say 4.0 Ghz on all 16 cores should be possible with good watercooling and decent boards/vrms/heatsink. It's a huge chip and soldered so cooling it should be a lot easier than with SL-X. Intel's 18 core will probably cost $2000 and won't clock a whole lot higher unless they fix the X299 issues (VRMs, TIM, etc). Man how I wish KVM would fix that NPT issue. eames fucked around with this message at 10:24 on Jul 13, 2017 |
|
#
?
Jul 13, 2017 10:18
|
|
|
so when they say turbo to 4ghz would that be XFR? and god drat I want one of them with a good mobo so I can crank all 16 cores to 4ghz 
|
|
#
?
Jul 13, 2017 10:27
|
|
|
If my little 1700 is any indicator then an underclocked/-volted TR 1950X should run 16 cores at 3.0 Ghz at less than 100W.  It's a shame games don't scale well with threads (yet) or that'd be my go-to platform.
|
|
#
?
Jul 13, 2017 10:40
|
|
|
Intel's not wrong, all CPUs are just silicon and metal and wood and concrete and plastic and glass that's been glued together, soldered together, bolted together, welded together, screwed together.
|
|
#
?
Jul 13, 2017 10:58
|
|
|
|
| # ? Apr 20, 2024 02:21 |
|
|
https://twitter.com/AMD_UK/status/877811134593736704
|
|
#
?
Jul 13, 2017 11:04
|
|