atomicthumbs posted:Apparently Asus is planning to break Intel's Z97 embargo and ship its motherboards tomorrow But... why... lol. This isn't halo 6
|
|
# ? Apr 28, 2014 04:33 |
|
|
# ? May 31, 2024 02:52 |
|
atomicthumbs posted:Apparently Asus is planning to break Intel's Z97 embargo and ship its motherboards tomorrow Is there any news on this from a legitimate tech website? Wccftech is the internet equivalent of little old ladies gossiping at church over stuff they made up/heard 3rd hand.
|
# ? Apr 28, 2014 05:02 |
|
The Lord Bude posted:Is there any news on this from a legitimate tech website? Wccftech is the internet equivalent of little old ladies gossiping at church over stuff they made up/heard 3rd hand. I looked around, and most of them were sourcing wccftech, so eh.
|
# ? Apr 28, 2014 05:20 |
|
Holy poo poo they weren't full of it, Asus Z97 reviews incoming HardOCP reviews the Asus Z97 Deluxe - I'm digging the low DPC latency and that Asus integrated the ROG audio solution into their Deluxe board. If the DPC latency is this low on a board with this much integrated hardware, I'm hoping its amazing on the ROG Hero and Ranger boards. Alereon fucked around with this message at 14:09 on Apr 28, 2014 |
# ? Apr 28, 2014 07:19 |
|
Alereon posted:Holy poo poo they weren't full of it, Asus Z97 reviews incoming Well I guess sometimes little old gossiping church ladies really do know what's really going on in the community. Silly black and gold colour scheme is gone finally, and Asus seems to have finally figured out that we don't need loving PCI ports any more. The Lord Bude fucked around with this message at 07:50 on Apr 28, 2014 |
# ? Apr 28, 2014 07:45 |
|
The Lord Bude posted:Asus seems to have finally figured out that we don't need loving PCI ports any more. says you
|
# ? Apr 28, 2014 09:03 |
|
Haha why the hell would Asus do that? Is really a little publicity and a few more sales worth pissing Intel off?
|
# ? Apr 28, 2014 10:21 |
|
Interesting that it overclocks so poorly, at least on non-Refresh chips. I wonder if that's just this board in particular, or if that's going to a Z97 problem in general.
|
# ? Apr 28, 2014 10:45 |
|
Looks like MSI still hasn't figured out how to position a cpu socket on a mITX board so that you can still fit decent tower coolers
|
# ? Apr 28, 2014 13:34 |
|
Alereon posted:Holy poo poo they weren't full of it, Asus Z97 reviews incoming The first paragraph of this review is astounding.
|
# ? Apr 28, 2014 14:02 |
|
Hace posted:Interesting that it overclocks so poorly, at least on non-Refresh chips. I wonder if that's just this board in particular, or if that's going to a Z97 problem in general.
|
# ? Apr 28, 2014 14:53 |
mobby_6kl posted:Haha why the hell would Asus do that? Is really a little publicity and a few more sales worth pissing Intel off? I know, I can't get over this. It's motherboards... is it even worth the tiniest bit off ill-will? Regardless cool for us. Alereon posted:Holy poo poo they weren't full of it, Asus Z97 reviews incoming DPC latency is a big one to me nice quote:In manual mode things were quite different. On a Z87 motherboard using this CPU I need about 1.285v max to get a 4.7GHz overclock out of the system. ugh so lucky, but quote:And again it was fairly boring as all I did was play with one setting and that’s the CPU voltage. alright, so he was requiring insanely higher vcore for a chip with otherwise proven to work at 1.285, yet all he was doing was messing with vcore? That isn't proof that the board sucks for OC, it just likely means he needs to raise vrin, stabilize with cache voltage, or make sure the default uncore ratios aren't going nuts or trying to match. Also the fact that it was strangely under temperature, while might just be motherboard improvements as implied... but if the other voltages were just too low those directly reduce temperature as well. IMO not proof that the board can't OC. To be fair I'm skimming very quickly because of work, but just my initial impressions Ignoarints fucked around with this message at 15:10 on Apr 28, 2014 |
|
# ? Apr 28, 2014 15:01 |
|
What the hell is the slot in the upper left of the Maximus VII Gene, above the ports? edit: it says "MPCIE" on the board
|
# ? Apr 28, 2014 21:12 |
|
What's the use of a mobile GPU slot? Does it provide video processing under light loads?
|
# ? Apr 28, 2014 21:20 |
|
Panty Saluter posted:What's the use of a mobile GPU slot? Does it provide video processing under light loads? it's not a mobile GPU slot, it's mini PCI express. you can install all sorts of stuff in there
|
# ? Apr 28, 2014 21:40 |
|
Oh, I looked up that set of initials and the first thing that popped up was mobile GPUs.
|
# ? Apr 28, 2014 22:01 |
|
Bare die is so mainstream now http://www.techpowerup.com/199956/next-gen-msi-oc-series-motherboard-to-feature-delid-die-guard.html
|
# ? Apr 29, 2014 00:00 |
|
Any news in the low-power world of the Atom? AMD just released some info on chips designed to compete with the current set of Atoms and seem to be doing so favorably. With the world of high end being pretty much stagnant, I'm finding myself more and more interested in how powerful and powersipping my Windows tablets can be.
|
# ? Apr 29, 2014 06:56 |
|
Just a wierd thought I had but the QX6700 extreme edition was released right around the time of Vistas ship date so you could say quite a few systems shipped and/or built ran XP. That's CPU would be fine for a lot of things today and yet it shipped during the prime of XP.
|
# ? Apr 30, 2014 22:42 |
|
OEM systems shipped that late in XP's lifespan came with Vista upgrade coupons, and anyone who spent $1000 on a CPU wanted an upgrade to 64-bit Windows as soon as it was available, so I don't think there's too many quad-cores running XP out there.
|
# ? May 1, 2014 18:34 |
|
I guess its just a new thing to have an almost 8 year old CPU be still OK for stuff today. I mean you could probably play BF4 at 1080p with an appropriately up to date GPU on that QX6700 or even a Q6600. Thats quite the upgrade path.
|
# ? May 2, 2014 00:41 |
|
Gaming has been bound by the Xbox 360/PS3 for 8 years
Proud Christian Mom fucked around with this message at 02:52 on May 2, 2014 |
# ? May 2, 2014 01:39 |
|
go3 posted:Gaming has been bound by the Xbox 360/PS2 for 8 years The PS2 is 14 years old
|
# ? May 2, 2014 02:14 |
|
go3 posted:Gaming has been bound by the Xbox 360/PS3 for 8 years I'm not convinced this is entirely a bad thing considering the longevity of any one PC build these days. And I'm sure there have been some machines with quad cores sold with XP, but I'd bet they skew towards workstations, especially since most non home users would have been specifically installing XP for most of Vista's lifetime, if not into Windows 7.
|
# ? May 2, 2014 03:56 |
|
carry on then posted:I'm not convinced this is entirely a bad thing considering the longevity of any one PC build these days. The reason for that, regarding gaming at least, is that it's been tied to consoles for ever.
|
# ? May 2, 2014 04:38 |
|
I remember my old Q6700 computer. Bought it when I knew nothing about computers way back when, paid like 2300. The thing also had a 9800GTX+ and an explosive Huntkey V500 power supply that likely ended up being the death of my motherboard. Had a 1tb HDD which BLEW my mind at the time. Ran XP on the bastard until 2011.
|
# ? May 2, 2014 13:10 |
|
hey jawn can you 'splain us what linus was on about the other day https://plus.google.com/+LinusTorvalds/posts/YDKRFDwHwr6
|
# ? May 2, 2014 15:32 |
|
StabbinHobo posted:hey jawn can you 'splain us what linus was on about the other day page fault handling apparently takes more cycles in haswell than on core duo not really a huge thing since paging is a sign you have bigger problems
|
# ? May 2, 2014 17:23 |
|
Isn't the number of cycles variable given the memory/storage used. Its not like the CPU is actually addressing (pun intended!) the page fault during those ~1000 cycles. Its simply waiting on spinning disks or SSDs or whatever to load the bits that aren't in memory. Or maybe what Linus is saying is that the page -is- in memory and the CPU just takes ~1000 cycles to realize it? edit: Refreshing my memory a bit(god drat puns) but shouldn't page fault penalties be measured in absolute time and not CPU cycles? And shouldn't it be measured against the ideal latency of the non volatile storage system? You can't have a page fault penalty that's smaller than the latency of your storage system. Shaocaholica fucked around with this message at 18:15 on May 2, 2014 |
# ? May 2, 2014 18:05 |
|
Malcolm XML posted:page fault handling apparently takes more cycles in haswell than on core duo This will make a huge drop in Linux performance aggregately for high performance compute overall.
|
# ? May 2, 2014 18:28 |
|
Shaocaholica posted:Isn't the number of cycles variable given the memory/storage used. Its not like the CPU is actually addressing (pun intended!) the page fault during those ~1000 cycles. Its simply waiting on spinning disks or SSDs or whatever to load the bits that aren't in memory. Or maybe what Linus is saying is that the page -is- in memory and the CPU just takes ~1000 cycles to realize it? Unless I'm misunderstanding what he's writing, Linus is just measuring the time the CPU takes to trap the page fault, enter the interrupt handler, and return from it to user code. Any processing by the OS kernel to handle the page fault would be on top of that cost.
|
# ? May 2, 2014 19:00 |
|
Mr.Radar posted:Unless I'm misunderstanding what he's writing, Linus is just measuring the time the CPU takes to trap the page fault, enter the interrupt handler, and return from it to user code. Any processing by the OS kernel to handle the page fault would be on top of that cost. In that case it would be less efficient than the Core arch he tested against. Is there anyway to see(for a non Intel engineer) what exactly the CPU is doing those cycles?
|
# ? May 2, 2014 19:41 |
|
necrobobsledder posted:It's kind of a big deal because you get page faults constantly as you use a machine. When a program loads, you get compulsory faults as various pages of a program are loaded into the correct segments and/or pages. A page fault is a major part of how the Linux kernel gets performance via the mmap call as well. It doesn't explicitly load anything, it marks pages to fault and the kernel handles the load into the page and can use neat tricks like zerocopy and predictive fault handling to queue up more fetches on I/O. "Huge"? No, I don't think so. Growth from 940 to 1045 cycles isn't that awful given that clock speeds have improved since Core 2, by enough that wall clock time should usually be lower anyways. Torvalds mentions that in the real world load he was profiling, one which hammers on the page fault handler a lot without actually swapping, CPU page fault overhead is 5% of all CPU time. If you had the power to do ridiculously impossible things, you could reduce the CPU's page fault overhead to 0 and his kernel compiles would only improve 5%. If you improved that overhead by the ratio of 1045 to 940, the improvement in wall clock time would be sub-1%. Which is why he isn't ranting and raving, the way he's wont to do when he discovers something outrageous. This is "hey Intel, you used to be able to do this in N cycles, now you're doing it in 1.11*N, so even though you're faster than you were before it looks like you could be even faster than that, what gives?"
|
# ? May 2, 2014 21:25 |
|
necrobobsledder posted:It's kind of a big deal because you get page faults constantly as you use a machine. When a program loads, you get compulsory faults as various pages of a program are loaded into the correct segments and/or pages. A page fault is a major part of how the Linux kernel gets performance via the mmap call as well. It doesn't explicitly load anything, it marks pages to fault and the kernel handles the load into the page and can use neat tricks like zerocopy and predictive fault handling to queue up more fetches on I/O. Yeah but again only like 5% of his time was taken up by page faulting and a lot of the commenters basically told him to use a better make system. I think one guy said that chromium's gyp system stats the same number of files, more or less, and takes 1sec compared to the 30 sec that linux's make does. When page fault times matter its because your app is page fault bound and if its that you really need to step back and check why your app is faulting so much.
|
# ? May 2, 2014 22:02 |
|
A researcher in HPC did some benchmarks with different RAM timings and showed a surprisingly large difference given how much data kept getting evicted from CPU cache and because people tend to emphasize slower, safer RDIMMs typically. I say it's important for HPC in the old joke / truism that supercomputing turns CPU bound problems into I/O bound problems. A performance regression like this is not exactly historically precedented beyond ones that are obviated from clock speed and architectural reasons like cycles for instructions across arch generations. Even then, I can't remember a case where the wall clock time could have even theoretically been slower without some sort of compensatory improvement. But the page fault handling latency is one thing but it's certainly within the context of other factors like page fault rate in general (the TLB should be better too after all).
|
# ? May 3, 2014 01:54 |
|
edit: Editing out this post because it's too uncomfortably close to revealing info I probably shouldn't; but the time page faults take in HPC is hugely relevant these days.
Chuu fucked around with this message at 04:55 on May 5, 2014 |
# ? May 5, 2014 04:52 |
|
Alereon posted:Holy poo poo they weren't full of it, Asus Z97 reviews incoming Hell yes, low DPC latencies. I haven't seen stuff this good for like a good 2 years.
|
# ? May 5, 2014 06:58 |
|
Most Intel 9-series boards seem to be out, here's Asus's Z97 lineup. I don't see the ROG Maximus VII Ranger, but I suspect its on its way. It looks like all but the cheapest boards also include Intel Ethernet, a significant step up from the crappy Realtek NICs.
|
# ? May 5, 2014 17:05 |
|
So is there any word on whether or not the new CPUs will work on the previous generation of motherboard?
|
# ? May 5, 2014 17:13 |
|
|
# ? May 31, 2024 02:52 |
|
Alereon posted:Most Intel 9-series boards seem to be out, here's Asus's Z97 lineup. I don't see the ROG Maximus VII Ranger, but I suspect its on its way. It looks like all but the cheapest boards also include Intel Ethernet, a significant step up from the crappy Realtek NICs. MSI is now using Killer garbage in just about everything. It's a shame, their motherboards look so nice.
|
# ? May 5, 2014 17:13 |