|
It wouldn't have happened if the processor had a gun.
|
#
?
Feb 18, 2018 12:56
|
|
|
|
| # ? Apr 25, 2024 18:53 |
|
|
Wouldn't have happened if it was ARMed
|
|
#
?
Feb 18, 2018 13:28
|
|
|
Some people have too much money.
|
|
#
?
Feb 18, 2018 14:28
|
|
|
Potato Salad posted:Wouldn't have happened if it was ARMed Yes. Yessss
|
|
#
?
Feb 18, 2018 17:00
|
|
|
Potato Salad posted:Wouldn't have happened if it was ARMed 
|
|
#
?
Feb 18, 2018 18:18
|
|
|
Has there been any actual benchmarks or at least informed discussion on the relative performance between the Intel w/Radeon and HBM2 and EMIB vs the AMD 2400G w/ Vega? The Intel thing seems like it's going to be way quicker with the HBM2 in there giving it mad bandwidth. Maybe we won't know until the Intel poo poo launches? https://en.wikichip.org/wiki/amd/ryzen_5/2400g https://en.wikichip.org/wiki/intel/core_i7/i7-8809g
|
|
#
?
Feb 18, 2018 21:08
|
|
|
I think it's going to get utterly murdered by the Intel hbm variant. Which annoys me as AMD could have launched that with ryzen cores and it'd own in a nuc or tiny steam box, and it's their tech.
|
|
#
?
Feb 18, 2018 21:33
|
|
|
The bus between the things was Intel’s tech. Regardless, I don’t think it matters, because it’s designed for pre-builts and you’re not going to see desktop motherboards with that socket. The whole point of the enterprise seems to be halting the advance of ARM in laptops, since the latest iPads have single core performance that rival MacBooks and have had Apple evangelism outposts like Gruber‘s blog and ATP mulling the necessity of Intel in MacBooks. System builders are a smaller segment than the people buying laptops like disposable objects, and both Intel and AMD suffer as X86 stakeholders if that market even begins a shift off the platform. Craptacular! fucked around with this message at 21:43 on Feb 18, 2018 |
|
#
?
Feb 18, 2018 21:35
|
|
|
Potato Salad posted:Wouldn't have happened if it was ARMed
|
|
#
?
Feb 18, 2018 21:44
|
|
|
Jago posted:Has there been any actual benchmarks or at least informed discussion on the relative performance between the Intel w/Radeon and HBM2 and EMIB vs the AMD 2400G w/ Vega? There's not much point comparing those two because they're made to serve completely different market segments. That Intel chip is going to cost three times as much.
|
|
#
?
Feb 18, 2018 21:54
|
|
|
Sure there's a point. We can compare the interconnect technology, bandwidth v cu's v frequency and cost is an interesting metric as well. There is crossover too... Do you buy a nuc, or build a probably cheaper alternative from an AMD apu?
|
|
#
?
Feb 19, 2018 00:45
|
|
|
Craptacular! posted:The bus between the things was Intel’s tech. Regardless, I don’t think it matters, because it’s designed for pre-builts and you’re not going to see desktop motherboards with that socket. It's telling that the best the "ARM ANY DAY NOW" crew can come up with is that some of Apple's most neglected models of laptop are vaguely comparable to ARM on certain single core only benchmark. For instance the MacBook Air is still running ultra low power Broadwell and the only change they've made for the CPU since 2015 has been to bump the base model up to 1.8 GHz base clocks.
|
|
#
?
Feb 19, 2018 00:55
|
|
|
I would be curious if AMD made a competing product using infinity fabric instead of EMIB and whether the low clocks on HBM would kill performance (if IF throughput is necessarily tied to memory clocks). It would probably need an enormous socket like TR4.
|
|
#
?
Feb 19, 2018 00:56
|
|
|
fishmech posted:It's telling that the best the "ARM ANY DAY NOW" crew can come up with is that some of Apple's most neglected models of laptop are vaguely comparable to ARM on certain single core only benchmark. For instance the MacBook Air is still running ultra low power Broadwell and the only change they've made for the CPU since 2015 has been to bump the base model up to 1.8 GHz base clocks. So long as "powerful laptop" = big and chunky with a heavier battery than other models, it's going to be the case. Please note I actually don't own any, but have seen other people try to use them on buses and airplanes.
|
|
#
?
Feb 19, 2018 03:55
|
|
|
Craptacular! posted:So long as "powerful laptop" = big and chunky with a heavier battery than other models, it's going to be the case. What are you smoking? There's no need to have a "big and chunky" laptop to use a CPU newer than 2015, or even to go up to 4 real cores.
|
|
#
?
Feb 19, 2018 04:03
|
|
|
NewFatMike posted:I would be curious if AMD made a competing product using infinity fabric instead of EMIB and whether the low clocks on HBM would kill performance (if IF throughput is necessarily tied to memory clocks). AMD is doing that though, and thier version will use the full die (28CU instead of 24). I'd love to see it on TR4, but that'd be a huge investment for something no one might invest in.
|
|
#
?
Feb 19, 2018 04:31
|
|
|
FaustianQ posted:AMD is doing that though, and thier version will use the full die (28CU instead of 24). I'd love to see it on TR4, but that'd be a huge investment for something no one might invest in. Oh wow I had totally forgotten. Do we have anything on that after the SiSoft posting from December?
|
|
#
?
Feb 19, 2018 07:38
|
|
|
OP finally replied with backstory on the Bulletproof Threadripperquote:The PC is not mine and everyone involved was ok. The owner of the PC was related in drugs and lived in Central America. They lost a lot of other expensive property and it looks like property that was the main target of the attack.
|
|
#
?
Feb 19, 2018 08:28
|
|
|
lmao my spare PC somehow had AMD services running even though it had never used any AMD hardware since a fresh install
|
|
#
?
Feb 19, 2018 11:04
|
|
|
GRINDCORE MEGGIDO posted:Okay that's pretty loving nice of them TBH THREADRIPHARDER obviously Palladium posted:lmao my spare PC somehow had AMD services running even though it had never used any AMD hardware since a fresh install Cool story bro
|
|
#
?
Feb 19, 2018 17:25
|
|
|
Palladium posted:lmao my spare PC somehow had AMD services running even though it had never used any AMD hardware since a fresh install x64 is AMD technology. To run x64 you need AMD services.
|
|
#
?
Feb 19, 2018 19:35
|
|
|
SlayVus posted:x64 is AMD technology. To run x64 you need AMD services. 32 bit is enough for that man, because lmao like AMD could ever out do intel in 64 bit, see how well itanium will take off.
|
|
#
?
Feb 19, 2018 19:38
|
|
|
Palladium posted:lmao my spare PC somehow had AMD services running even though it had never used any AMD hardware since a fresh install reformat
|
|
#
?
Feb 19, 2018 19:46
|
|
|
I'm a PAE man
|
|
#
?
Feb 19, 2018 19:46
|
|
|
SlayVus posted:x64 is AMD technology. To run x64 you need AMD services. I don't believe you re: the second sentence. I have seen embedded operating systems running on x64 that most definitely do not have any "AMD services" running
|
|
#
?
Feb 19, 2018 19:51
|
|
|
Threadtearer
|
|
#
?
Feb 19, 2018 20:10
|
|
|
Threadviscerator
|
|
#
?
Feb 19, 2018 20:19
|
|
|
Threadripandtear,ripandtearyourthreader
|
|
#
?
Feb 19, 2018 20:20
|
|
|
🅱️readripper
|
|
#
?
Feb 19, 2018 20:25
|
|
|
Thread 2: Ripper Boogaloo
|
|
#
?
Feb 19, 2018 20:33
|
|
|
PerrineClostermann posted:Thread 2: Ripper Boogaloo Thread's not even a year old. You AMD people and your refreshes 
|
|
#
?
Feb 19, 2018 20:41
|
|
|
....
|
|
#
?
Feb 19, 2018 21:00
|
|
|
PerrineClostermann posted:Thread 2: Ripper Boogaloo Threadripper 3: Rip Threads with a Vengeance
|
|
#
?
Feb 19, 2018 21:15
|
|
|
https://browser.geekbench.com/v4/cpu/6690641 Pinnacle Ridge lookin close. Ryzen 5 2600 sporting what looks to be a non-enginering sample ID string on Geekbench. Still in the reference AMD motherboard though.
|
|
#
?
Feb 19, 2018 21:20
|
|
|
Cygni posted:https://browser.geekbench.com/v4/cpu/6690641 I want to see the OC benchmarks. The shrink should give it some serious OC headroom (10%-15%) compared to before. If they can hit 4.4-4.6 reliably I'll upgrade from my i7 3770k and then probably upgrade again in 2019. 2019 is also going to be a big year for AMD on the server if they deliver with zen2.
|
|
#
?
Feb 20, 2018 21:06
|
|
|
Khorne posted:It should be out in april. The licensing savings for SMB moving to single die, 32 core Threaddenwerffen are freaking great
|
|
#
?
Feb 20, 2018 21:10
|
|
|
Potato Salad posted:The licensing savings for SMB moving to single die, 32 core Threaddenwerffen are freaking great All the big vendors have been moving to per-core licensing models around the time Intel started hitting 12/16c per socket. Expect more of that.
|
|
#
?
Feb 21, 2018 00:54
|
|
|
Khorne posted:It should be out in april. Is it the same process as the Apus? Not seeing any worthwhile overclocks on those anywhere.
|
|
#
?
Feb 21, 2018 01:03
|
|
|
BangersInMyKnickers posted:All the big vendors have been moving to per-core licensing models around the time Intel started hitting 12/16c per socket. Expect more of that. bring back CMT (for muh licenses).
|
|
#
?
Feb 21, 2018 01:11
|
|
|
|
| # ? Apr 25, 2024 18:53 |
|
|
GRINDCORE MEGGIDO posted:Is it the same process as the Apus? Not seeing any worthwhile overclocks on those anywhere. the ones on
|
|
#
?
Feb 21, 2018 01:12
|
|
