420chan now has a web-based IRC client available, right here
Leave these fields empty (spam trap):
Name
You can leave this blank to post anonymously, or you can create a Tripcode by using the float Name#Password
Comment
[*]Italic Text[/*]
[**]Bold Text[/**]
[~]Taimapedia Article[/~]
[%]Spoiler Text[/%]
>Highlight/Quote Text
[pre]Preformatted & Monospace text[/pre]
1. Numbered lists become ordered lists
* Bulleted lists become unordered lists
File

Sandwich

penis pump

Community Updates

420chan now supports HTTPS! If you find any issues, you may report them in this thread
AMD Ryzen by FuckGankinway.scr - Wed, 01 Mar 2017 20:19:29 EST ID:Mw0E3xZ0 No.120684 Ignore Report Quick Reply
File: 1488417569947.jpg -(19638B / 19.18KB, 640x353) Thumbnail displayed, click image for full size. 19638
So what do you guys make of the hype? There's only a few more hours before the review embargos lift and we'll finally get to see independent benchmarks. I'm really hoping AMD can pull of what they're promising because the Dekstop CPU market has been pathetic these last few years and this will really shake things up if the gains AMD is claiming prove true.
>>
AugustusHollerham.php - Thu, 02 Mar 2017 03:23:56 EST ID:Pz4wNt2s No.120689 Ignore Report Quick Reply
>>120684
I'm in the market for a new desktop but they're all already sold out so I won't be getting one for a while at least. Hopefully they fuck with Intel enough to cause the price of chips to come down in general.
>>
ShittingHugglefuck.pcl - Thu, 02 Mar 2017 03:32:56 EST ID:FtUa1Cfw No.120690 Ignore Report Quick Reply
>>120684
Well Intel dropped all of their prices, so they seem to be taking Ryzen pretty seriously. They cut the 6950x by 300 fucking dollars.
>>
DavidPockfield.jar - Thu, 02 Mar 2017 13:50:47 EST ID:bFee7tEo No.120691 Ignore Report Quick Reply
1488480647962.gif -(3095636B / 2.95MB, 500x247) Thumbnail displayed, click image for full size.
Aaaaaaand it's looking like AMD is up to their old tricks again

>Queen of synthetic benchmarks but real-world tests show cheaper Intel CPUs outperforming it.
>Gaming performance is abysmal and a 3 year old i5 beats the 1800x in most situations
>Overclocks for shit with next to no headroom above the ~4.0ghz they're retailing at.
>Using way, way more power than their rated TDP. a 95w AMD 1800x draws more power than a 145w Intel 6900X
>Memory support is fucked with most reviewers incapable of running at higher clocks than 2666mhz. It remains to be said if this is an issue of MoBo firmware or the CPUs ability to handle the bandwidth. Because some cherry-picked engineering samples are running at up to 3600mhz on the same hardware with the same firmware, all early indications point to a CPU issue.
>AMD is requesting reviewers only benchmark it with 4K gaming so the GPU is the bottleneck and the lack of AMD CPU performance seems negligible
>It's been deduced that AMD used trickery for their official benchmarks like running the x99 Intel CPUs with dual-channel memory instead of the quad channel memory that platform supports and 99% of users of said platform would be utilizing
>During the official gaming benchmarks, AMD would regularly zoom in on skyboxes to saturate draw-calls on the Intel test system to artificially hamstring performance
>Between 5% lower IPC than intel, the afformentioned memory issues, and the fact that Intel can reliably clock almost a whole Ghz higher than AMD, the $500 Ryzen is getting its ass kicked by the $330 i7 7700k in Single-Threaded, general Multi-Tasking, and Gaming Workloads.
>Ryzen only excels in niche, Multi-threaded workloads
>>
AugustusHollerham.php - Thu, 02 Mar 2017 14:58:50 EST ID:Pz4wNt2s No.120692 Ignore Report Quick Reply
>>120691
Nice source.
The review embargo seemed a bit fishy, but I don't know if that's a normal thing in the hardware field.
>>
FrederickHundergold.jpeg - Thu, 02 Mar 2017 15:22:25 EST ID:Ki3Nprwe No.120693 Ignore Report Quick Reply
>>120691
It's about what I expected and not as bad as you're saying. I'm probably going to buy one in a few months. However you end up feeling, enjoy AMD's relevance while it lasts.

>>120692
Engineering samples are always under NDA and review samples are always under embargo, at least for the big three.
>>
FuckGankinway.scr - Thu, 02 Mar 2017 15:36:36 EST ID:Mw0E3xZ0 No.120695 Ignore Report Quick Reply
>>120692
It's all information from
-PCPerspective
-Techspot
-OC3D
-LinusTechTips
-GamersNexus (arguably the most negative review and includes the information regarding AMD's pre-release shenanigans)
-Tech City

Pretty much everyone's showing the same results across the board and the only difference seems to be in the interpretation. Most are content with putting a Cinebench score up on a pedestal and calling it a day while others are contextualizing the target workloads and examing how Ryzen performs poorly in mainstream, single-threaded applications and how the gimpy dual-channel memory is holding it back in workstation applications and CUDA acceleration will completely destroy it in Content Creation applications.
>>
AugustusHeckleham.ace - Fri, 03 Mar 2017 15:02:42 EST ID:7F/QA+Tm No.120702 Ignore Report Quick Reply
If you get one you can finally run gentoo, lol.
>>
RebeccaDuckworth.hlp - Fri, 03 Mar 2017 17:36:49 EST ID:Mw0E3xZ0 No.120703 Ignore Report Quick Reply
So digging deeper, it's pretty evident the launch of Ryzen was rushed and handled poorly. Apparently, board vendors only had 3 weeks to build their motherboards and write the firmware and it's confirmed by AMD that the EFI microcode on MSI and ASUS boards provided to reviewers was extremely buggy, which might account for the very poor memory performance on just about all Motherboards with the exception of those made by Gigabyte. The revised firmware for the public launch on MSI boards supposedly fixes these issues and some reviewers were provided this firmware on request but not all were in contact with MSI before the launch to receive the updated firmware. Most reviews based on benchmarks performed on MSI boards are consistent with the bad firmware on ASUS boards. There's still not a clear picture what's going on with ASUS and whether or not they've resolved the issues yet but the vast majority of reviews were performed on the Asus Crosshair Hero mobo and indicate consistent issues. It does seem that Gigabyte boards are the closest that are working as intended although this is anecdotal but the few reviews based on Gigabyte boards are indicating improved memory performance.

There are also more reports from Reviewers that AMD requested they enable certain OS and UEFI features that favor the AMD platform while requesting that they DISABLE features such as Multi-Core Enhancement (or similiar implementations dependent on the motherboard vendor) and other such High-Precision Event Timers that benefit the Intel platform, essentially stacking the deck in favor of AMD. Altogether, AMD is appearing to be extremely underhanded in the way they're trying to make their product live up to the hype.

I don't know guys. I've always used AMD all the way back in the Win95 days up until Phenom II and this just doesn't live up to the hype for me. The 7% lower IPC than Intel, combined with the 12%+ frequency deficit means that Ryzen just can't compete with Intel in mainstream workloads and it's becoming more and more common for GPU acceleration to handle heavily multi-threaded tasks and something like a GTX 1070 will absolutely destroy both the Ryzen 1800X and the i7 6900x in applications like Adobe Premier and Blender. Ryzen just kinda feels like it's a day late and a dollar short.

However, I do think the Ryzen 7 1700 does have potential to be a decent entry-level content-creation chip if all the aforementioned problems with microcode and application-level optimization get worked out. Combine that with its lower TDP (and the fact that it's not running 1.4v+ through the vcore) which could give it better overclocking headroom than the XFR enabled 1700X and 1800X and If it proves to reliably hit 3.8ghz-3.9ghz, it would also be within the ballpark of Broadwell-E gaming performance. All that and at only $330 is a value you can't argue against.

In case you're interested, some of the reviewers going in depth into the issues they were having with this launch
https://youtu.be/P1dhYDm7SLw?t=38m44s
https://www.youtube.com/watch?v=TBf0lwikXyU
>>
FannyPellystock.uue - Fri, 03 Mar 2017 22:21:23 EST ID:TA5YDbO+ No.120706 Ignore Report Quick Reply
>>120703
3 weeks ? sound insane
>>
AngusNocklefare.sgml - Sat, 04 Mar 2017 19:41:35 EST ID:Mw0E3xZ0 No.120707 Ignore Report Quick Reply
1488674495865.jpg -(35230B / 34.40KB, 800x261) Thumbnail displayed, click image for full size.
So there's more and more evidence that the memory issues are really holding back Ryzen. High-Clocked kits are boosting performance by as much as 10% and altogether, it seems that the platform scales a lot better with RAM speeds due to its architecture compared with Intel.

https://forums.anandtech.com/threads/ryzen-strictly-technical.2500572/
>The northbridge of Zeppelin is officially called as the data fabric (DF). The DF frequency is always linked to the operating frequency of the memory controller with a ratio of 1:2 (e.g. DDR4-2667 MEMCLK = 1333MHz DFICLK). This means that the memory speed will directly affect the data fabric performance as well. In some cases, it may appear that the performance of Zeppelin scales extremely well with the increased memory speed, however that is necessarily not the case.
>In many of these cases the abnormally good scaling is caused by the higher data fabric clock (DFICLK) resulting from the higher memory speed, rather than the increased performance of the memory itself.

Benchmarks are showing this to be consistent and beneficial to some gaming benchmarks.
http://www.eteknix.com/memory-speed-large-impact-ryzen-performance/
https://www.purepc.pl/procesory/premiera_i_test_procesora_amd_ryzen_r7_1800x_dobra_zmiana?page=0,43
>>
GrahamFemmlehood.qif - Sat, 04 Mar 2017 22:13:11 EST ID:Y7Hm+wSB No.120708 Ignore Report Quick Reply
As I recall, the buzz around the 3-6 month delay for Ryzen was chipset related so motherboards at launch not being rock solid maybe shouldn't be surprising.
>>
what do you mean - Sun, 05 Mar 2017 17:14:44 EST ID:xR29hi1x No.120710 Ignore Report Quick Reply
>>120702
>>
ClaraGubblebire.gz - Sun, 05 Mar 2017 22:37:43 EST ID:PUbnWT1r No.120711 Ignore Report Quick Reply
>AMD
a joke. intel for poorfags. if you're building a gaming pc why be a poorfag about it.
>>
PhineasNuvingfoot.msg - Sun, 05 Mar 2017 22:42:07 EST ID:p6545IKk No.120712 Ignore Report Quick Reply
>>120711
If you fix your capitalization and grammar, I will give you a banana sticker.
>>
ClaraGubblebire.gz - Mon, 06 Mar 2017 08:03:46 EST ID:PUbnWT1r No.120713 Ignore Report Quick Reply
>>120712
text me when they're not lagging 2 years behind
>>
AlbertSandlenere.bin - Mon, 06 Mar 2017 08:48:51 EST ID:Pz4wNt2s No.120714 Ignore Report Quick Reply
>>120713
Generally, when you're righting a number less than 10, you should use the word equivalent, i.e. one, two, three.
>>
CarolineHammerhog.smd - Mon, 06 Mar 2017 09:28:19 EST ID:38amYmTX No.120715 Ignore Report Quick Reply
1488810499103.jpg -(9198B / 8.98KB, 220x200) Thumbnail displayed, click image for full size.
>>120714
Next time, try 'e.g.'
>>
AlbertSandlenere.bin - Mon, 06 Mar 2017 09:51:45 EST ID:Pz4wNt2s No.120716 Ignore Report Quick Reply
>>120715
You missed "righting".
>>
ClaraGubblebire.gz - Mon, 06 Mar 2017 10:08:24 EST ID:PUbnWT1r No.120717 Ignore Report Quick Reply
>>120715
or just say "for example"
>>
ElizaBrudgestot.rm - Mon, 06 Mar 2017 19:54:42 EST ID:K+eQksZx No.120722 Ignore Report Quick Reply
Bested by the i5.
Absolutely Pathetic
>>
HenryGinkinshit.rsc - Tue, 07 Mar 2017 14:06:21 EST ID:7F/QA+Tm No.120723 Ignore Report Quick Reply
>>120722
*for gaming.
It smokes intel when it comes to business and multimedia.

I know, I know it's kind of pointless cuz businesses will use a high end intel cpu for workstations anyway.
AMD's best bet are data-centers at this point where price/performance and multi-threading matters.
>>
AlbertFamblebury.sys - Tue, 07 Mar 2017 14:28:06 EST ID:3XALAJyp No.120724 Ignore Report Quick Reply
1488914886641.png -(21659B / 21.15KB, 582x768) Thumbnail displayed, click image for full size.
>>120723
>It smokes intel when it comes to business and multimedia.
Elaborate on that.
>>
NellSemmerson.rar - Tue, 07 Mar 2017 14:55:39 EST ID:hK4cRdde No.120725 Ignore Report Quick Reply
>>120724
You ever have a hard time telling the difference between willful stupidity and the effects of congenital syphilis?
>>
AlbertFamblebury.sys - Tue, 07 Mar 2017 15:03:59 EST ID:3XALAJyp No.120726 Ignore Report Quick Reply
>>120725
What? I'm asking him to elaborate on the circumstances where Ryzen will perform better than the various Intel Chips. From what I understand Gaming is probably the most demanding process a non-enterprise,non-government, or non-scientific user will undertake.
>>
HenryGinkinshit.rsc - Tue, 07 Mar 2017 16:38:29 EST ID:7F/QA+Tm No.120727 Ignore Report Quick Reply
>>120724
You post a gaming benchmark in response to a statement about workstation performance, you have issues.
http://www.techspot.com/review/1345-amd-ryzen-7-1800x-1700x/page3.html
>>
BetsyShittingdock.cpl - Tue, 07 Mar 2017 17:19:26 EST ID:0AUNs60E No.120728 Ignore Report Quick Reply
>>120727
The image wasn't a response.
I asked you to elaborate.
The question was a response.
>>
CarolineMabbleford.ini - Tue, 07 Mar 2017 20:04:03 EST ID:Ki3Nprwe No.120729 Ignore Report Quick Reply
>>120728
So how did your mom's syphilis work out?
>>
CarolineGacklefield.dao - Tue, 07 Mar 2017 21:17:48 EST ID:K+eQksZx No.120730 Ignore Report Quick Reply
>>120729
You're an obvious troll.
>>
CarolineCimmledock.pdf - Wed, 08 Mar 2017 11:27:59 EST ID:3KZXtrJX No.120732 Ignore Report Quick Reply
>>120727

The link you posted isn't really relevant to business or workstation either. Most people looking for workstation and multitasking aren't going to buy gaming cpu's like i7's they're going to buy xeons and those benchmarks are far a few, I'm having trouble finding someone who has actually sat down and compared. Most I can find is sites like cpu.userbenchmark.com which mostly compare stats and not real world benchmarks. Even so the 1800x only comes out ontop in cinebench against the i7's. At most these cpu's might be useful for YouTube content creators who don't want to spend $1000 on a 6900k just to be a few seconds faster on a two tasks and i'm not so sure Youtubers use Excel that often so that brings it back down to cinebench. As for these cpu's being used in businesses and data centers not so sure that's gonna be very common either. Looking at the techspot benchmarks it looks like someone can buy a cheaper 7700k and overclock it to get similar if not better performance than the 1800x in the few spots it out performs.
>>
CorneliusNeshlot.bfc - Sat, 11 Mar 2017 03:41:59 EST ID:7F/QA+Tm No.120751 Ignore Report Quick Reply
>>120732
Yeah that's true and I already mentioned that businesses in need of a high performance workstation probably will still use a high end Intel CPU, which includes Xeons. 6 and 8 core versions of them aren't available right now with a recent architecture though. And on it's own a Xeon won't significantly beat an i7 if the number of cores are the same.
It still would be interesting to see how a Broadwell 8 core Xeon would do in comparison. Still it's not really a fair comparison since those cost 3-4 times as much as the high end Rizon.
>>
SimonPeckleham.vxd - Sun, 12 Mar 2017 10:06:49 EST ID:Mw0E3xZ0 No.120757 Ignore Report Quick Reply
1489327609387.jpg -(354847B / 346.53KB, 1212x1920) Thumbnail displayed, click image for full size.
There's been a lot of talk about the Windows Scheduler not assigning tasks to threads properly and holding back Ryzen's performance but that alternative fact has been completely debunked.
https://www.pcper.com/reviews/Processors/AMD-Ryzen-and-Windows-10-Scheduler-No-Silver-Bullet

The Windows Scheduler handles Ryzen's cores and SMT just fine and assignment is identical to Broadwell-E. The issue seems to be when processes change to a different Core Complex(CCX). The architecture is more or less behaving similarly to a dual-proc setup with two sets of 4 cores, rather than a true 8 core processor like Broadwell-E. Anytime a process bounces from the first set of 4 to the second, there's an inordinately high amount of latency as it passes through the Core Complex's "Infinity Fabric" that links the two together, whereas on Intel, the latency is identical across all cores. However, it does appear that when processes bounce between cores on the same CCX, Latency is much, much lower than Intel's architecture. Ryzen's Multithreaded workload performance is outstanding and single-threaded performance is 5%-10% lower than Intels as expected but Multi-Tasking and Gaming Workloads are proving to be far worse than they should be given Ryzen's specs on paper. The high latency between CCXs very well could be a piece to this puzzle.
>>
breakabond - Sun, 12 Mar 2017 19:24:03 EST ID:fVWGPSRj No.120758 Ignore Report Quick Reply
just a quick glance in this thread, but the pattern seems to be
>Person A: Ryzen is bad because [proof]
>Person B: No [scapegoat] is just holding Ryzen back
>Person A: Ad hominem
>Repeat ad nauseum...
>>
LillianClapperfield.pm3 - Sun, 12 Mar 2017 20:28:46 EST ID:kFrnwkeE No.120759 Ignore Report Quick Reply
Here is what I wanna fucking know -

Is it better than the creaky ass AMD Phenom 2?
Can I afford it?
What is the performance between this and the comparably priced Intel product?

Bang for my far too few bucks is what I want to see.
>>
PollySibblemin.wsz - Mon, 13 Mar 2017 09:24:50 EST ID:Mw0E3xZ0 No.120763 Ignore Report Quick Reply
>>120759
>Is it better than the creaky ass AMD Phenom 2?
Yes
>Can I afford it?
Can you afford $330+ the cost of buying into a new platform (so around $800 if you need to include RAM and an OS)?
>What is the performance between this and the comparably priced Intel product?
Compared to the i7 7700k, Multithreaded performance is significantly faster, singlethreaded performance is about 10%-15% worse when running both at stock (this difference increases when overclocking the 7700k), and Gaming performance is erratic and nonsensical on the Ryzen chips and everyone is scrambling to understand what's going on with that.

It'll be worth it to wait for Ryzen 5 if you really care about bang for the buck.
>>
CharlesHuckleway.lit - Wed, 15 Mar 2017 11:44:29 EST ID:Pz4wNt2s No.120769 Ignore Report Quick Reply
>>120759
>Can I afford it?
Post your online banking details and I'll check for you.
>>
LydiaHadgebury.pm6 - Thu, 16 Mar 2017 12:43:26 EST ID:Mw0E3xZ0 No.120771 Ignore Report Quick Reply
http://www.anandtech.com/show/11202/amd-announces-ryzen-5-april-11th
Unfortunately, its confirmed that the R5 lineup is going to be a shaved down 8 core with lower frequencies. The 6-core CPUs are no surprise there but what's surprising is the 4 core models are retaining both CCXs and shaving two cores off each, rather than utilizing only one 4-core CCX. There is no way this won't put undue burden on the data fabric and given the performance issues when threads crosstalk between CCXs on the 8 core models, The 4 core models are looking to be dreadful in my opinion.
>>
EmmaGugglestock.icl - Thu, 16 Mar 2017 18:05:40 EST ID:qaJ3iinO No.120772 Ignore Report Quick Reply
>>120771
Well that's weird. I wonder what the yields are like and what they plan to put out for laptops.
>>
LydiaHadgebury.pm6 - Thu, 16 Mar 2017 19:04:29 EST ID:Mw0E3xZ0 No.120773 Ignore Report Quick Reply
>>120772
The APU rumors aren't completely concrete at the moment but it's speculated that they'll use a single 4 core CCX joined with some sort of GCN architecture for the iGPU. I'm interested how those will turn out because I've seen anecdotes of people disabling entire CCXs on their 1800Xs and seeing improved performance in some games (and worse performance in others).
>>
BeatriceClayshit.dev - Thu, 23 Mar 2017 10:19:53 EST ID:Mw0E3xZ0 No.120789 Ignore Report Quick Reply
So Rumors for the high-end X399 platform are starting to crop up.
https://www.overclock3d.net/news/cpu_mainboard/amd_s_rumoured_180w_16-core_ryzen_part_will_reportedly_run_at_3_1_3_6ghz/1

>16 Cores
>32 Threads
>Quad Channel DDR4
>Way more PCIe lanes
>3.1ghz/3.6ghz with 180w TDP (though other sites have stated 2.4ghz/2.8ghz at 150w so take this with grain of salt)

If the specs prove to be true, the estimated price of around $1000 and if the extra cores scale up correctly, This is going to obliterate Intel's high-end desktop CPUs
>>
WalterBrimmercocke.jpg - Fri, 24 Mar 2017 17:17:57 EST ID:7F/QA+Tm No.120794 Ignore Report Quick Reply
>>120789
>This is going to obliterate Intel's high-end desktopworkstation CPUs
FTFY


Current consumer desktop apps and games don't utilize a massively multicore system yet, and as for workstations well. Most of the typical high performance massively parallel apps are using GPGPU today. And for those enterprise/research applications that can't be done using CUDA/OpenCL there is Intel's Xeon Phi.

What AMD is doing is pitching exclusively to the consumerist hardware enthusiast market, which is sadly pathetic.
180W, yeah fuck those guys.
>>
AlbertPickwill.pub - Fri, 24 Mar 2017 20:02:47 EST ID:Mw0E3xZ0 No.120795 Ignore Report Quick Reply
>>120794
The HEDT platform is supposed to compete head-to-head with X99. The monstrous Naples CPUs are the ones that will compete with Xeon.

>What AMD is doing is pitching exclusively to the consumerist hardware enthusiast market
Most casual users are adopting smartphones and tablets for every-day use so the Desktop Market is becoming more and more of a niche for enthusiasts an power users. How exactly is bringing down the barrier to entry on Enthusiast-grade components a bad thing?
>>
CharlesBunham.ico - Sat, 25 Mar 2017 01:34:15 EST ID:YUkDrNLy No.120798 Ignore Report Quick Reply
>>120795
Yeah for real. Barely anyone I know has a laptop anymore. Let alone a desktop.

All the work/school computers are based on zero clients/VM's.

Everybody just uses an ipad or iphone now.
>>
ThomasWabberbanks.swp - Sat, 25 Mar 2017 16:10:28 EST ID:7F/QA+Tm No.120799 Ignore Report Quick Reply
>>120798

Straight consumers never had any point of owning a PC/Laptop, but if you do have any business in creating any type of digital content you need one.

>Everybody just uses an ipad or iphone now.
Yeah Apple has very a very effective marketing tactic about pitching their crap. Go ahead and read the butthurt post of "creatives" who drank the coolaid that they can throw away their PC and use an ipad pro instead.

So I agree the masses don't need a PC and they only did in the past decade because there were no better alternatives tailored for consumption only.


Back to the point on the intended consumer base of AMDs new line of processors. By hardware enthusiast I mean people who see their PC as a hobby so they are willing to spend a significant share of their income on it. And in principle there is nothing wrong with that, Nvidia's Gforce Titan line is that. But they still sell models they can produce in significant numbers.

What AMD is doing is overstretching their semiconductor manufacturing process in order to produce few expensive chips with even fewer functioning ones because of the low yield resulting from a massive silicon area.
It's a resource intensive marketing stunt and I won't support that.
>>
AlbertPickwill.pub - Sat, 25 Mar 2017 17:26:58 EST ID:Mw0E3xZ0 No.120800 Ignore Report Quick Reply
>>120799
What the hell are you on about? First off, AMD isn't manufacturing the silicone, GlobalFoundries is and second, AMD's yields are just fine and they're in the process of rolling out their mainstream products with HEDT and Server parts following suit.
>>
BarnabyPaggleworth.drv - Sun, 26 Mar 2017 07:11:31 EST ID:7F/QA+Tm No.120801 Ignore Report Quick Reply
>>120800
Who manufactures the chips doesn't matter, what does is how they are designed.
For each same architecture die area is directly proportional to power consumption.
And the yield is directly proportional to the reciprocal of the die area.

Again Nvida, Intel and others have/had their own enthusiast CPUs too but they only produce them after surplus. And if you think AMD doesn't have to pay for bad silicon because they use a third party foundry you're delusional.
>>
MarthaHicklestock.udf - Sun, 26 Mar 2017 09:54:53 EST ID:Mw0E3xZ0 No.120802 Ignore Report Quick Reply
>>120801
Nvidia has always launched a new generation with their flagship Titan GPUs and all subsequent cards are built on stripped-down architecture based on defective silicone. Intel does the same by typically launching their stripped-down two core i3 and Pentium parts a few months after they launch their 4 core i5s and i7s. This is done because AMD, Intel, and Nvidia will typically have healthy yields on fully-functional silicone and will need to build up a stock of the stripped-down components to be used as low-end parts. This is again happening with AMD's stripped-down Ryzen 5 and Ryzen 3 that will be rolled out next month.

Where you seem to be getting confused is on Intel's HEDT "E" processors which are subjected to a different binning and validation process than their mainstream counterparts due to having more complex architecture. These are the only "high end" parts that are rolled out last (that I can think of) and it's not even an apples to apples comparison at that point because their based on fundamentally different architecture at that point, just as AMD's HEDT platform will be based on fundamentally different architecture that will involve a different binning and validation procedures than that of the mainstream components.

You keep saying their Yields are bad but there's zero evidence to support that. In fact, supply of Ryzen 7 is starting to pile up due to a motherboard shortage.
>And if you think AMD doesn't have to pay for bad silicon because they use a third party foundry you're delusional.
No, they'll buy the bad silicon and sell it as low-end parts, just like Intel and Nvidia. That's a standard procedure in chip manufacturing.

>For each same architecture die area is directly proportional to power consumption.
And since Ryzen is based on a 14nm architecture means everything is just fine. The fact that an AMD 8core running at 65w can compete with an Intel 8core running at 140w is fucking amazing and the rumors of a 150w-180w 16 core is just as impressive.
>>
BarnabyPaggleworth.drv - Sun, 26 Mar 2017 13:56:39 EST ID:7F/QA+Tm No.120803 Ignore Report Quick Reply
>>120802
I don't have time to go into all of your points, but just let me tell you that a 16 core die for a CPU with a 180W TDP won't be used as a 4 or 8 core one.

Oh and i7s aren't an enthusiast product and that brings me to this:
Stop posting fanboyish claims, no the 8 core ryzen doesn't compete with intels 8 core line since it has even an issue of surpassing intels 4 core CPUs in half of the benchmarks.
>>
MarthaHicklestock.udf - Sun, 26 Mar 2017 16:53:56 EST ID:Mw0E3xZ0 No.120804 Ignore Report Quick Reply
>>120803

>i7s aren't an enthusiast product
Are you seriously trying to say Broadwell-E isn't an enthusiast platform?

>it has even an issue of surpassing intels 4 core CPUs in half of the benchmarks.
In some Gaming workloads and single-threaded work loads, yes, Ryzen can't compete with the sheer speed of Intel CPUs that can push 5ghz. When it comes to multi-threaded workloads, The 8 core Ryzen CPUs demolish the 4 core Intel CPUs, Destroy the 6 Core Intel CPUs, and go toe to toe with the 8 Core Intel CPUs and they compete with 95% of Intel's performance at

>a 16 core die for a CPU with a 180W TDP won't be used as a 4 or 8 core one.
No fucking shit Sherlock! a 16 Core die for a CPU with a 180W TDP will be used as a 16 Core CPU and possibly shaved down to 12 and 14 core parts depending on how they're binned. What a fucking novel concept!


Report Post
Reason
Note
Please be descriptive with report notes,
this helps staff resolve issues quicker.