|
Pawn 17 posted:At least read up on the basics of microeconomics before you postulate something like that. Competition does affect luxury good pricing (competition shifts the supply curve). There's nothing in that link about pricing, specifically, and the lawsuit you're referencing was centered around intel bullying oems into ignoring amd, and had nothing to do with predatory pricing in the high-end consumer market. And as fishmech has already noted, amd and intel have both repeatedly offered gently caress off pricing for top end processors for a long time. Here's an inferior pentium 4 extreme edition launching at $900+: http://www.tomshardware.com/reviews/intel,751-8.html
|
# ? Mar 3, 2017 09:01 |
|
|
# ? May 21, 2024 14:03 |
|
People don't buy those gently caress off top of the range processors in spite of them being expensive, they buy them because they're expensive. High end stuff isn't rational and value for money is unimportant compared to exclusivity.
|
# ? Mar 3, 2017 10:05 |
|
|
# ? Mar 3, 2017 10:09 |
|
Over at Semi Accurate there is some juicy tidbits about what is going on with the benchmarks being all over the place... Apparently, some reviewers were not using a fresh Windows 10 build (they were taking an Intel build drive and letting it "see" the new AM4 Platform - WHAAAA??? WHY???) in some attempt to save time I guess? This lead to Boost being BROKEN, which apparently escaped several reviewers. They also had to use non-public beta BIOSes which were sent out a few days ago, guess they didn't get the memo? Also they needed to disable high precision event timers and set the windows 10 power option to high performance due to AM4 Windows 10 drivers being MIA at this point. Then you have lots of the top reviewers pulling a head scratcher in regards to the memory, just flat out giving up and letting it run at gimped speeds when higher speeds were a click away (some Motherboards wouldn't post some would, I guess these reviewers forgot how to operate a PC BIOS). Most of this is obviously AMDs fault, some of it is the Reviewers fault, overall this launch has a bit of a black eye but like any black eye it will get better with time.
|
# ? Mar 3, 2017 10:51 |
|
Apparently, in gaming there is a big impact when HT is on. Seems to confirm the problems with scheduling we were talking about. The question is...is this fixable via BIOS injected microcode? Dante80 fucked around with this message at 11:08 on Mar 3, 2017 |
# ? Mar 3, 2017 11:01 |
|
Dante80 posted:Over at Semi Accurate there is some juicy tidbits about what is going on with the benchmarks being all over the place... If it does then the R5 and R3 launch should show the necessary corrections and improve standing. That and if Zen+ is a bigger better die with the promised 15% increase in IPC (not impossible, the did beat their original 40% mark) it'll be irrelevant anyway - Skylake-X/Kabylake-X/Coffeelake/Cannonlake are still going to be in the same ballpark of each other (like +5%) and that's all of 2017 and most of 2018, so 15% increase on Ivy Bridge-E IPC means only 8-11% behind for Zen+ assuming zero clock improvements. That's projecting too far in the future but apparently Zen can hit higher clocks than Broadwell/Haswell-E using LN2 so their is clearly some room to grow if they can get the voltage threshold for those clocks down. AMD's forever mantra "Wait for the next iteration of our idea", but not really in this case with R5 and R3, all sound like good budget options, no gently caress you I am not buying a G4560 TYOOL2017 for anything serious and LOL at the 7350K. Dante80 posted:Apparently, in gaming there is a big impact when HT is on. Seems to confirm the problems with scheduling we were talking about. Why the 100Hz cap?
|
# ? Mar 3, 2017 11:06 |
|
Percentages instead of FPS? I may be reading this wrong (I don't know French) but I think it means % of performance comparison between the same CPU having SMT on and off. The baseline 100% is with SMT off, and then they compare the performance difference when enabling it. That is why there is a 100.9% in there I think. Dante80 fucked around with this message at 11:12 on Mar 3, 2017 |
# ? Mar 3, 2017 11:08 |
|
Dante80 posted:Percentages instead of FPS? I may be reading this wrong (I don't know French) but I think it means % of performance comparison between the same CPU having SMT on and off. The baseline 100% is with SMT off, and then they compare the performance difference when enabling it. Makes sense, and if so than loving lol that SMT bug guts a Broadwell-E class processor to Ivy Bridge-E, holy poo poo. Wasn't the Phenom TLB bug just as bad? Is this a BIOS, driver or Microcode issue?
|
# ? Mar 3, 2017 11:18 |
|
Dante80 posted:Most of this is obviously AMDs fault, some of it is the Reviewers fault, overall this launch has a bit of a black eye but like any black eye it will get better with time. Yeah it's probably best to forget what happened and go back to sleep until reviews for the R5's are out, which are sure to include the R7 aswell. Launching your top of the line 16 thread CPU arch without having the required Windows drivers in place is pure :AMD: tho. FaustianQ posted:Makes sense, and if so than loving lol that SMT bug guts a Broadwell-E class processor to Ivy Bridge-E, holy poo poo. Wasn't the Phenom TLB bug just as bad? Is this a BIOS, driver or Microcode issue? Some guy made an effortpost on reddit about poorly coded games being super erratic with thread management, and when they switch caches between the two quadcore blocks it's real bad for Zen but Intel is much less affected due to their unified cache. Or something like that. https://www.reddit.com/r/Amd/comments/5x7oaq/ryzens_memory_latency_problem_a_discussion_of/ sauer kraut fucked around with this message at 11:48 on Mar 3, 2017 |
# ? Mar 3, 2017 11:36 |
|
sauer kraut posted:
Game developers have basically been able to rely on Intel being the only serious game in town and thus focus on making their games run fast on the market dominant CPU's when they do optimization. Lots of benefits of a monopoly, everything except price and innovation.
|
# ? Mar 3, 2017 11:51 |
|
I definitely se a 1700 in my future. Or maybe 2700, if they have to release a new chip to get all the nastier bugs fixed. Currently on a haswell, and while it's a great chip, streaming 60fps video at any decent quality causes my game performance to go to absolute poo poo because I need to use the full 4 cores to do it. A sub $400 8core haswell? Yes, please, I'll encode on 6 threads, games run fine on 2.
|
# ? Mar 3, 2017 11:55 |
|
I'm still balls in, getting a 1700 though instead of a 1800X (since apparently there is no performance binning, only factory overclocks). Planet Coaster will use all 16 threads (Hail Cobra!), so I am perfectly fine with it...C:
|
# ? Mar 3, 2017 12:10 |
|
Btw...the most hilarious part about the launch is that AMD supplied the reviewers mainly with the top ASUS board (like in 7-8 out of 10), and this proved to be the worst performer out of the bunch in OCing, memory, stability etc. ASROCK Taichi and the Gigabyte K7/5 were much better. https://www.youtube.com/watch?v=mW1pzcdZxKc Also, there seems to be a pattern forming (have read at least 3 reviewers mention it) in that Ryzen gets worse av/max FPS but better min and "smoother gameplay". At this point in time I think I can attribute that to biased reviewers (smoothness, really?), but it would be fun to see how this unfolds in later reviews. Dante80 fucked around with this message at 12:37 on Mar 3, 2017 |
# ? Mar 3, 2017 12:33 |
|
White Rock posted:Game developers have basically been able to rely on Intel being the only serious game in town and thus focus on making their games run fast on the market dominant CPU's when they do optimization. Even if there are more things in software that can be done between AMD and x86 game devs, this is a fundamental problem with the architecture. It's not just something that can be fully fixed with optimization. The Windows Scheduler most likely won't be able to take care of everything perfectly when all is said and done.
|
# ? Mar 3, 2017 12:44 |
|
Dante80 posted:Also, there seems to be a pattern forming (have read at least 3 reviewers mention it) in that Ryzen gets worse av/max FPS but better min and "smoother gameplay". At this point in time I think I can attribute that to biased reviewers (smoothness, really?), but it would be fun to see how this unfolds in later reviews. You do realize you can quantify "smoothness" by graphing frame times, right? The better reviewers out there will back up their impressions with that info.
|
# ? Mar 3, 2017 12:46 |
|
TheJeffers posted:You do realize you can quantify "smoothness" by graphing frame times, right? The better reviewers out there will back up their impressions with that info. All the reviewers that talk about better smoothness also show lower minimum framerates in benchmarks against the intel products. Or don't show min framerates at all. Granted, they are talking about playing the game instead of running the benchmarks, but the disparity is tough to swallow. Right? I gave an example from the video above. See the benches, and then watch this part. https://www.youtube.com/watch?v=mW1pzcdZxKc&t=924s Dante80 fucked around with this message at 13:26 on Mar 3, 2017 |
# ? Mar 3, 2017 13:23 |
|
Pawn 17 posted:At least read up on the basics of microeconomics before you postulate something like that. Competition does affect luxury good pricing (competition shifts the supply curve). the "basics of microeconomics" are simplified nonsense for children who can't understand calculus and are very general ideas put in algebraic graph form if you think they are how anything actually works you didn't understand what you were being taught just like if you thought molecules are actually wooden blocks with rods between them
|
# ? Mar 3, 2017 14:15 |
|
People seem to confuse frame time with minimum fps.
|
# ? Mar 3, 2017 14:27 |
|
Minimum FPS in benchmarks usually happens on map loads/streaming in new content anyway, and is a useless statistic. 99th percentile is a bit better, but really, you want frametime graphs to see what the gently caress.
|
# ? Mar 3, 2017 14:31 |
|
Truga posted:Minimum FPS in benchmarks usually happens on map loads/streaming in new content anyway, and is a useless statistic. 99th percentile is a bit better, but really, you want frametime graphs to see what the gently caress. I want to see DigitalFoundries frametime video but I guess they're busy benchmarking all the Switch games
|
# ? Mar 3, 2017 14:33 |
|
Rastor posted:We apparently have different definitions of "mainstream" desktop computer use. twitch is streaming 2 million unique people per month now. god knows how many people are uploading gaming vids to youtube every week. god knows why, I don't get it, but lots of people do regardless of whether they have an audience or not. recording video of games is totally mainstream. not that broadcasting video of games to zero people is a really great reason to buy a $500 CPU, intel or AMD.
|
# ? Mar 3, 2017 14:46 |
|
Klyith posted:twitch is streaming 2 million unique people per month now. god knows how many people are uploading gaming vids to youtube every week. god knows why, I don't get it, but lots of people do regardless of whether they have an audience or not. recording video of games is totally mainstream. Well, even AMD themselves are making their own SKUs above $320 look plenty bad already. Pay more for factory OC and XFR = lol
|
# ? Mar 3, 2017 15:04 |
|
sauer kraut posted:Some guy made an effortpost on reddit about poorly coded games being super erratic with thread management, and when they switch caches between the two quadcore blocks it's real bad for Zen but Intel is much less affected due to their unified cache. Or something like that. I wonder if you could fix that by setting core affinity. Maybe not. I wonder if you could break that by setting core affinity. Might be worth a test. Force a benchmark to run on a couple cores here, a couple cores there.
|
# ? Mar 3, 2017 15:05 |
|
Maxwell Adams posted:I wonder if you could fix that by setting core affinity. I dunno man, but I never heard any of this core switching thing negatively affecting C2Qs with the same split L2 cache except they are connected across stone age FSB. Looks like it's something inherently flawed about the Ryzen design.
|
# ? Mar 3, 2017 15:15 |
|
I've watched a bunch of comparison videos where they do the left side Intel and right side AMD and I am nearly sure the AMD stay smoother under the heavier scenes. Its very hard to tell but its there.
|
# ? Mar 3, 2017 16:32 |
|
Truga posted:I definitely se a 1700 in my future. Or maybe 2700, if they have to release a new chip to get all the nastier bugs fixed. A 7700K wouldn't be enough? That was going to go into my new desktop this year but I'm at least considering AMD now because it'd be nice to have a bit of a guarantee that I'll be able to swap out the CPU when they get better in a couple years. Palladium posted:I dunno man, but I never heard any of this core switching thing negatively affecting C2Qs with the same split L2 cache except they are connected across stone age FSB. Looks like it's something inherently flawed about the Ryzen design. If the scheduler is moving a high-demand workload (like a game!) to a different core, that sounds like a problem with the scheduler. Err, unless it found one clocked way higher, but that shouldn't happen often enough to make a noticeable impact unless the scheduler is written with invalid assumptions about how the hardware acts. Which, oh hey look the drivers aren't out yet! Welp
|
# ? Mar 3, 2017 16:49 |
|
amd gif right here
|
# ? Mar 3, 2017 16:57 |
|
Paul MaudDib posted:This release is a shitshow, I am pissed that AMD rushed this so bad that there is only one (lovely, half-baked) chipset and no mITX boards. GRINDCORE MEGGIDO posted:Ryzen's going to be on my radar when they respin it, there are mature bioses and board variety (itx). Biostar is starting to show this off: Racing X370-GTN, because you use your Mini ITX system for racing I guess?
|
# ? Mar 3, 2017 17:01 |
|
You mean finally somebody that knows how to make motherboards
|
# ? Mar 3, 2017 17:10 |
|
The socket layout is pretty . The wasted space is the width of an m.2 slot.
|
# ? Mar 3, 2017 17:16 |
|
There's nothing Racing about these boards e: vvvvvv haha jesus buglord fucked around with this message at 17:21 on Mar 3, 2017 |
# ? Mar 3, 2017 17:18 |
|
Is that two sata ports up at the edge, then a couple more against the pcie slot?
|
# ? Mar 3, 2017 17:19 |
|
Interestingly enough, Ryzen suffers no problems due to SMT in most other applications. Far from that (it is essentially the new perf/$ king for many tasks) Something is loving up the games, since the mean performance is less than the straight IPC and clock disparities would lead us to believe before launch (we always knew that it would not top a 500mhz higher clocked i7, we did not expect them though to do significantly worse than the HEDT chips).
|
# ? Mar 3, 2017 17:51 |
|
Boiled Water posted:The socket layout is pretty . The wasted space is the width of an m.2 slot. Where are you seeing wasted space on that thing? The label?
|
# ? Mar 3, 2017 17:53 |
|
Munkeymon posted:Where are you seeing wasted space on that thing? The label? The two black plastic bits, if they were removed and the holes put closer to the socket you should be able to just squeeze something in there.
|
# ? Mar 3, 2017 17:57 |
|
Where is AVX2 expected to matter and do any of these reviews look at AVX2 256-bit wide commands?
|
# ? Mar 3, 2017 17:57 |
|
Dante80 posted:Interestingly enough, Ryzen suffers no problems due to SMT in most other applications. Far from that (it is essentially the new perf/$ king for many tasks) Ryzen really does seem to kick all asses when it comes to compiling.
|
# ? Mar 3, 2017 18:06 |
|
Twerk from Home posted:Where is AVX2 expected to matter and do any of these reviews look at AVX2 256-bit wide commands? AVX and AVX2 are primarily intended to improve handling of certain video encoding, database processing, and general server workloads. The intended areas for it to help are similar to the areas that the various SSE revisions were intended to help.
|
# ? Mar 3, 2017 18:11 |
|
So, the impression I'm getting is that Ryzen 7 effectively renders Intel's X99 platform moot, but lags noticeably behind Intel's mainstream chips when it comes to gaming, and that the Core i5 7600k remains the default CPU choice for gaming builds (for now).
|
# ? Mar 3, 2017 18:16 |
|
|
# ? May 21, 2024 14:03 |
|
Delusibeta posted:So, the impression I'm getting is that Ryzen 7 effectively renders Intel's X99 platform moot, but lags noticeably behind Intel's mainstream chips when it comes to gaming, and that the Core i5 7600k remains the default CPU choice for gaming builds (for now). More or less. Remember, that it is also Intel that renders the X99 moot (for new builds of course), since they are changing sockets anyway soon. With regards to gaming, I think that the AMD 6c/12t and 4c/8t offerings might push i5s relevance, especially if the AM4 platform surpasses some of the day 1 problems it has in the following months. At least in the perf/$ context. Again though, it seems that Intel is also going to do that anyway in 2018, with 6c/12t descending on the mainstream platform. Which btw will also kill 1151 I think. At least AM4 will stay in place for the next 3-4 AMD revisions. Dante80 fucked around with this message at 18:24 on Mar 3, 2017 |
# ? Mar 3, 2017 18:20 |