[EDIT]hmmm.....whats going on at AMD? (and history of 3Dfx Rampage)

[ocre]

Swoosh...
A sigh of relief. Not only was I wrong to be looking at amd situation as negative, it seems I really underestimated the dire and hopeless position nvidia is walking into.

Turns out AMD has a new powerful weapon called RTG And they nvidia is in some serious trouble.

Even worse for nvidia, I think there is some indication, perhaps even verification, that AMD has somehow harnessed the power of the jedi and are putting in inside their soon to be released GPUs. Honestly, its amassing. I didnt even know that jedi power really exist.

How can nvidia compete now. Its over for them. SAD to say, they are toast. Just have to wait till 2017 and then we all can watch nvidia crumble to the mighty RTG and their soon to be released jedi powered unstoppable GPUs.

Amd was never late. This was their plan all along. Just setting the stage for the ultimate smack down, so the world can all bare witness to the ultimate power of the jedi and so we all shall know that it is AMD that possess it.

This is one of the most greatest achievements of all of mankind.

[SteelCrysis]

And while all of you are predicting gloom and doom, AMD scores another juicy contract: https://www.extremetech.com/computing/23...d-services

[RolloTheGreat]

And while all of you are predicting gloom and doom, AMD scores another juicy contract: https://www.extremetech.com/computing/23...d-services

It’s a dual GPU card with 2x AMD Fiji GPUs, 8GB of HBM (4GB per GPU), 1TB/s of total memory bandwidth, and a 300W power draw.

Could Crossfire be coming to the Google cloud?!

"Oooooooooo...K....GGG...ooooooGLE...w..HY...ARE....yooouuu...s...oo....un...EVEN.....t....oday"

Results:
Google cloud powered by Crossfire

[ocre]

And while all of you are predicting gloom and doom, AMD scores another juicy contract: https://www.extremetech.com/computing/23...d-services

Predicting doom and gloom?
Have you had your fingers in your ears? Have you decided to ignore reality?

For a long long time now, AMD has been reporting loss after loss. Its not just some wild predictions to bash on AMD. The reality, doom and gloom sounds rosy if you go back and really look at their performance...take away the name and just look at the figures, its not doom and gloom. The reality is that its just been horrible. Terrible, awful, really really bad, as in not good.

You know, its so bad that people would be jumping from the rooftops if amd nist managed to make a tiny profit...if amd just made one dollar profit this year, now that would have high fives and parties thrown across the nation.

Call me pessimistic but honestly at this point I would worry about anyone who isnt when it comes to AMD at this point. If they dont have a vested interest, then we are looking at a real example of the most perfect definition of the word insane.

You surely must consider that amd would grasp at anything they possibly can to try to get some news and conversation. The company has been struggling badly and only a fool would be denying this.
I am very used to AMD pumping up every tiny little thing that I honestly don't even expect this to ve a story of great riches to them. I could be wrong but I have no faith that this is some huge significant source of cash for them and I completely expect them to end the year with another quarterly loss to pile on to the many many many they have piled up over the last several years.

If I am wrong, than I will smile when you come back gloating in the wake of amd actually turning meaningful profits from this amazing deal. But, without any numbers or figures at all in the article, its probably a much safer bet for me to consider this just another fluff piece spread by amd themselves trying to publicize any and every little thing in an effort to convince people they are still relevant. I mean, the forced emphasis of their "open" software and then the plug about cuda running on amd cards makes it hard to believe this is anything but a fluff piece from AMD.

[RolloTheGreat]

I wasn't planning to do this, because I got this from an old friend who is deep in the industry. It turns out AMD has had a plan all along:

[BenSkywalker]

Extremetech should be ashamed of themselves, holy shit is that article misleading and fully dishonest in its intent-

https://cloudplatform.googleblog.com/201...tform.html

Google Cloud will offer AMD FirePro S9300 x2 that supports powerful, GPU-based remote workstations.

Well, it is AMD parts being used in the cloud, however ET mentions things like HPC, deep learning and AI and, well-

We'll also offer NVIDIA Tesla P100 and K80 GPUs for deep learning, AI and HPC applications that require powerful computation and analysis. GPUs are offered in passthrough mode to provide bare metal performance. Up to 8 GPU dies can be attached per VM instance including custom machine types.

That's from Google, not the apparent AMD shills over at Extremetech. Given the movement of stock prices based on the way this shit was covered, they may want to cover their asses before this goes beyond the FTC and hits the SEC's desk.

[SteelCrysis]

https://www.techpowerup.com/227905/amd-z...7-branding

Recent reports peg AMD's upcoming line of microprocessors based on Zen micro-architecture as being labelled SR3, SR5 and SR7 for different hardware tiers (with the SR3 being the lowest-performing, and SR7 being, naturally, the highest-performing). A recent post on Chip hell claims that a leaked slide from an AMD presentation give us these insights, with further information on pricing: it's shown in the roadmap that all Zen SR (Summit Ridge) processors will sell for higher than RMB 1500 ($220).
...
That said, I can't help but feel like this will not be the branding on Zen chips at all, and that the "SRx" moniker is nothing more than a visualization for AMD's roadmap. From a brand perspective, it doesn't make much sense to name AMD's future chips with the platform code name. Platforms change, which would lead to necessary changes in the naming scheme for their Zen processors - and if AMD wants to follow in Intel's footsteps of building strong brand recognition, an annual or biennial change in nomenclature (Summit Ridge = SRx; and an eventual Summit Ridge hypothetical successor being called "Raven's Ridge" = RRx) wouldn't make much sense, breaking continuity in brand awareness and eventually giving rise to confusion among customers.

AMD might instead opt for using ZEN's well-known code-name, imbibing it on their products: "z3", "z5" or "z7" would, in my opinion, make much more sense. AMD is expected to support and iterate on Zen for years to come (much like Intel does under their iX line of processors), and the "zX" moniker would be a symbol of AMD's confidence in ZEN being the horse to carry them to more successful waters. I think AMD would to well to copy Intel's naming scheme, for a variety of reasons: the "zX" capitalizes on Intel's well-known and well-understood branding scheme, and is so instantly recognizable for its proximity in nomenclature that it would increase consumer awareness. Also, I believe it would increase perception of performance being on-par with Intel (AMD's z7 being comparable to Intel's i7) but Z is (along with X) usually (and arguably) associated with higher performance than other nomenclatures (just look at all those "gaming products" with "X"s and "Z"s plastered all over their names).

[RolloTheGreat]

They can call it "Bulldozer II" as long as price/performance is there.

I'd just like to see them make something decent.

[ocre]

Extremetech should be ashamed of themselves, holy shit is that article misleading and fully dishonest in its intent-

https://cloudplatform.googleblog.com/201...tform.html

Google Cloud will offer AMD FirePro S9300 x2 that supports powerful, GPU-based remote workstations.

Well, it is AMD parts being used in the cloud, however ET mentions things like HPC, deep learning and AI and, well-

We'll also offer NVIDIA Tesla P100 and K80 GPUs for deep learning, AI and HPC applications that require powerful computation and analysis. GPUs are offered in passthrough mode to provide bare metal performance. Up to 8 GPU dies can be attached per VM instance including custom machine types.

That's from Google, not the apparent AMD shills over at Extremetech. Given the movement of stock prices based on the way this shit was covered, they may want to cover their asses before this goes beyond the FTC and hits the SEC's desk.

Apparent shills at extremetech?  Its been obvious to me for sometime..
These sites commonly put out fluff pieces which are more like full page ads than anything else.

I would suggest that amd plants many articles and part of their marketing strategy is to get articles like these up on sites all over the world.  The shilling you are accusing extremetech of here is business as usual and its not right to be upset with extremetech for this one case considering that all notable tech sites put out fluff pieces.  Its without question. Now, the individual methods that a company might use to get these sites to put up such articles probably differ.  Considering the reason that these sites exist is to make money, it must be understood that money would be one of those methods.

Controversy usually increases traffic which also means more money.  So there would be some individual "investigations" to write about if they can have enough clicks to make it worth while.  This is important because I dont want to make the impression that every fluff or negative article is directly paid for by a party with interest. But many articles are given to these tech sites on a platter, and its purpose is to serve in a particular companies interest.  Which company can be determined by looking at whatever the substance is.

So while some juicy articles and supposed investigations might be easily persuaded with very little effort, others -I am sure- are pretty much full fledged bought and paid for ads..planted for marketing purposes to try to sway or create a desired image.

[RolloTheGreat]

Yep, tech sites are essentially the marketing arm of tech companies.

However, that is pretty much how it has people with the actual knowledge to review parts in any meaningful way would never waste their time running benchmarks.

So you get a bunch of guys like Apoppin that know how to turn on a computer spoon fed tech info in a way they (and the rest of us) can understand. And they run the benchmarks, paraphrase thee info.

We should probably be glad they do, like I said, no one with the skillset to actually review hardware would do it.

[SteelCrysis]

Rumor says Zen will be released on January 17: https://www.techpowerup.com/227944/amds-...or-usd-300

[SteelCrysis]

https://www.techpowerup.com/227977/amds-...e-a12-9800

AMD's first desktop processors based on the "ZEN" micro-architecture will be priced above the fastest socket AM4 part currently in production, the A12-9800 APU. While AMD hasn't started selling the A-series socket AM4 APUs in the retail channel, the top-end A12-9800, which is faster than the A10-7870K from its previous-generation, is expected to be priced at least the same ($160) or higher. This would mean that even the cheapest ZEN part, under the ZEN SR3 series, likely quad-core; could be priced above $160.

Two key slides from a presentation AMD intended for Chinese distribution channels, was leaked to the web by ChipHell tech forum members, revealing that ZEN parts will be priced in a segment "above" RMB ¥1,500 (USD $218), assuming that that's a price inclusive of all taxes. The company also put out some performance numbers for the A8-9600 "Bristol Ridge" socket AM4 APU, keeping the "Godavari" A8-7650K as baseline, compared to the Intel Core i3-6100 "Skylake" processor. The A8-9600 was shown to be faster in most tests.

[ocre]

Uhm...

So its gonna be much slower than an i3 6100 accept in graphics apps. Looking across the chart, its faster in the same cases apps that the 7650 is faster in. The 3d mark seems to show a jump but the others are really such a small improvement over their current A8.

Its even slower in the last case.

I dont want to be terribly negative, the power improvement compared to their own A8 should be noted. But, its intel chips they must compete with. And, this is amd....so I can only expect these to be the best of the best cases. Thats really what I am battling with here. And their tdp figuring, its not something I trust. Not this day and age. They have been exaggerating and misrepresenting and its been increasingly worse. Unrealistically so.

Honestly, the situation doesn't seem all that great. Amd has been loud when they dont have all that much to be singing over. They typically flaunt and brag like there is no tomorrow. They were making all kinds of claims and press about Zen the further you look back and then the chip is pushed out and out..things got much quieter.

I mean, obviously they have improved the ipc if we look at these graphs. I just wonder though, isnt their current A8 a 32nm chip? Would you expect a dramatic improvement for a chip that is 14nm? Especially when these are moderate chips.. Intel gained and gained in the moderate and low tdp areas.

I hate to be a drag..but....

[SteelCrysis]

I don't mind you being a drag.

[ocre]

[SteelCrysis]

https://www.techpowerup.com/228272/amd-n...d-december

AMD's next-generation ZEN processor family is bound for a grand entrance this December, with the company planning a special media event that will be live-streamed to the public. Called "New Horizon," AMD is designing the event to be an "exclusive advanced preview" of the processor ahead of its Q1-2017 scheduled launch. The preview will see eSports and Evil Geniuses legend PPD put "Zen" through its paces, according to its announcement. With this AMD appears to be showing off the chip's advanced gaming capabilities. " If you're serious about gaming, this is an event you do not want to miss," AMD's announcement signs off. The event stream goes live on the 13th of December, at 15:00 CST (21:00 hrs UTC).

[ocre]

Can't wait

[SteelCrysis]

https://www.techpowerup.com/228422/myste...phics-card

Benchmarks of an as-of-yet-unknown GPU have surfaced in the Ashes of the Singularity benchmark database - one that could point towards performance levels on an upcoming AMD GPU (the RX 490?). With new cards based on Vega expected to arrive in early 2017 sporting a brand new architecture, hope is they will fill in AMD's utterly lacking (as in, equal to zero) enthusiast-class performance graphics solutions. The new GPU appears to be within spitting distance of the already established GTX 1080, with only a few percentage points separating the two (with the unknown graphics card having an 85.1 average frame-rate in all batches at 1080p standard settings, against 93.4 on a GTX 1080 - with both values potentially varying with particular overclocking characteristics and so on, so please don't take them at face value.

[ocre]

Wonder who could have leaked this? I mean, ashes of singularity!!!!

Its such a massive hit. This game is so popular, so demanding and awesome, so amazing, so phenomenal that it has easily become the standard to which all graphics cards shall be judged. Once AoS alpha was released, it was clear that it was so much more than a modern day Crysis. AoS was the game crytek could have only dreamed of. The second AoS was dreamed up (by an independent company with no ties at all to any gpu maker), it was the game of all games, the most perfect software destined to be the greatest gpu benchmark of all time...
Even before it was in its Alpha form, this was the most significant and most important benchmark that completely invalidated all that came before and all that shall till the end of time.

So this leak, its of course, AoS.
And, you know...Amd loves to brag and show off how awesome their gpus are when playing AoS. Remember the Polaris reveal, amd showed off how much faster and more efficient polaris was than the gm204 pascal.
I mean, wholly smokes!!! Have you guys seen AoS first hand? Its jaw dropping, unlike anything we have ever seen. The most beautiful graphics, technically unmatched on every level. Masterpiece would be an understatement, this game is so far advanced that humans will be extinct before another company can top it. This is the game that shows how amazing PC graphics has become, and everyone who has seen it is still trying to recover. Minds blown completely out of their skulls, pieces of their brains falling from the stratosphere hundreds of miles away.

So, amd is showing us what next level graphics can truly be like on their next level graphics cards.

Here is the most interesting knacks:

1) "sporting a brand new architecture"

-----yeah. I will eat my hat if vega is not a GCN derivative. At best, it will be a GCN evolution. I truly would love to see a brand new design.. But, yeah..

2) its almost as fast as the 1080!!!!

---my vega estimates based on the info we heard a while back: vega would struggle to beat the 1080 much less the pascal titan or the 1080ti.
Its 1070 - 1080.. Its gonna struggle to beat the gp 104.

Lets hope I am wrong...Lets hope this leak is bogus, or its not the long awaited vega

[SteelCrysis]

AMD's MI25 machine-learning acceleration GPU is based on Vega: https://www.techpowerup.com/228617/amds-...ccelerator
Unknown Vega prototype with 8 GB of HBM2 VRAM is able to play Doom at 4K 60 FPS using Vulkan: https://www.techpowerup.com/228635/amd-v...ver-60-fps

[SteelCrysis]

https://www.techpowerup.com/228601/amds-...processors

The fact that this is even possible also begs the question as to why is it so - are the yields not good enough on the RX 460 chips that AMD is being forced to "lock" portions of a higher-end chip? I would wager this isn't likely, since Polaris 11 "Baffin" is such a small chip already that yield issues shouldn't have such effects. Most likely, we'd guess that AMD is re-purposing the mobile versions of the RX 480 (RX 480M, with some now suspicious-looking 1024 stream processors) that haven't passed power requirements for the 35 W those are rated at - but which surely correctly function at the RX 460's "up to 75 W" TDP. That might mean that AMD designed the original "Baffin" with 1024 stream processors for mobile devices, and carved the RX 460 as a way to recycle eventual defective dies that would surface for that particular SKU, not passing certification. But yields being better than expected, and demand for RX 460 cards outweighing the defective chips' output, AMD was forced to soft lock the extra stream processors on the "Baffin" SKU to satisfy demands.

[SteelCrysis]

https://www.techpowerup.com/228648/amd-r...t-zen-cpus
Here comes Zen:

AMD has apparently chosen RYZEN (pronounced "risen") as the marketing name for the first consumer enthusiast processors based on the "ZEN" micro-architecture. Slated for market availability in Q1-2017, the first RYZEN part will feature 8 cores, 16 threads, 20 MB of total cache (8x 512 KB L2 + 2x 8 MB L3), and a clock speed of over 3.40 GHz. The chips will be built in the socket AM4 package, and will be launched alongside a new wave of motherboards with AMD X370 chipset.

Given the relatively modest clock speeds by AMD's standards, it's safe to say that AMD has made big IPC gains, and is now tapping into those gains. Intel, on the other hand, is clocking its upcoming Core i7-7700K at 4.00-4.40 GHz. AMD is also introducing enthusiast-segment features with RYZEN, including XFR (eXtended frequency range), Smart Prefetch, a tightly tuned power-management system built from the ground up, called AMD Pure Power, and its related feature, AMD Precision Boost.

[ocre]

AMD's MI25 machine-learning acceleration GPU is based on Vega: https://www.techpowerup.com/228617/amds-...ccelerator
Unknown Vega prototype with 8 GB of HBM2 VRAM is able to play Doom at 4K 60 FPS using Vulkan: https://www.techpowerup.com/228635/amd-v...ver-60-fps

Oh god!!!

deep learning, AI? Where have I heard of that before?

Oh, yeah...
Nvidia.

Does AMD have to follow every foot step? Can they not find ways and forge their own paths to success?

Nvidias deep learning efforts are a little more than the same gpus with a new sticker. They have been actively building business partners and networks...ecosystems, tons of software and support.

Its kind of interesting to see amd now all of a sudden sporting deep learning solutions. Oh, and announcing them before the official vega graphics card revealing.

What? Its like, huh? So..surely people are dancing in the streets cause AND just announced their much anticipated, latest, brand new of deep learning solutions. What all their fans have been waiting so long for.

Yeah...truly odd

[ocre]

https://www.techpowerup.com/228601/amds-...processors

The fact that this is even possible also begs the question as to why is it so - are the yields not good enough on the RX 460 chips that AMD is being forced to "lock" portions of a higher-end chip? I would wager this isn't likely, since Polaris 11 "Baffin" is such a small chip already that yield issues shouldn't have such effects. Most likely, we'd guess that AMD is re-purposing the mobile versions of the RX 480 (RX 480M, with some now suspicious-looking 1024 stream processors) that haven't passed power requirements for the 35 W those are rated at - but which surely correctly function at the RX 460's "up to 75 W" TDP. That might mean that AMD designed the original "Baffin" with 1024 stream processors for mobile devices, and carved the RX 460 as a way to recycle eventual defective dies that would surface for that particular SKU, not passing certification. But yields being better than expected, and demand for RX 460 cards outweighing the defective chips' output, AMD was forced to soft lock the extra stream processors on the "Baffin" SKU to satisfy demands.

The idea that amd had the bright idea to fill such a high volume slot like that?

"hey, I got a great idea, we can take salvaged 480m chips to create the 460 line."

Its truly bizarre logic. See, the article just got thru making the case that yields of polaris 11 could not possibly be bad enough that amd would have to use cut down chips. And then poses the idea that these were rejected 480m chips that wouldnt pass in the 35watt power envelop. Okay...
But how does that explain then being cut down? Even if these chips couldn't make it as a 35watt chip, the 460 has over double the tdp spec. Why are these hypothetically rejected mobile chips cut? He just isnt making any sense at all. There is a massively generous tdp threshold for the 460 vs 480m. He goes a long way to try to convince people (and possibly himself) that this is not a result of Polaris 11 yields but chips have to be cut sometimes in order for them to fit into a desired power envelop. That is still an issue of silicon missing its target, not yielding to expectations.

Many chips have come back too power hungry and have come to the market as a cut down version of its original planned spec.

There is the another possibility. What about apple? Did apple ask for exclusivity? Was tonga released as a cut down chip because of an apple contract which gave them a unique and exclusive chip just for them? Wasnt there some exclusive nvidia chip that nvidia provided for apple? I remember something like that a while back.

Anyhow, does apple have anything to do with this? Perhaps exclusive rights of a desktop model?
Kind of strange that amd would cut a deal like that, for a tiny segment, while making their pc customers have to settle for less.
Its just that AMD has so little to offer already and then cutting that down....Yeah

Also, its sad to see the 480m is absolutely nothing like the desktop chip. I mean, its way smaller. We were all told that the 480 was a game changer and amd was gonna disrupt mobile, that nvidia was gonna be in trouble. But, here we are. The 480m is not even half of a desktop 480. Its nowhere close. Thats kind of pathetic

[SteelCrysis]

https://www.techpowerup.com/228981/amd-r...-promising

French tech print magazine "Canard PC" is ready with early benchmarks of an AMD Ryzen 8-core processor. The scan of a page from its Ryzen performance review article got leaked to the web, revealing three key performance takeaways. In the first selection of tests, Canard PC put Ryzen through synthetic CPU-intensive tests that take advantage of as many CPU cores/threads as you can throw at them. These include the likes of H.264 and H.265 video encoding, WPrime, Blender, 3DSMax 2015, and Corona. Ryzen was found to be faster than the quad-core Core i7-6700K, and the six-core i7-6800K, but somewhere between the i7-6800K and the eight-core i7-6900K.

The next selection of tests focused on PC gaming, with a list of contemporary AAA titles, including "Far Cry 4," "Battlefield 4," "The Witcher 3: Wild Hunt," "Anno 2070," "GRID: Autosport," and "ARMA III." Here, the Ryzen sample was found to be underwhelming - it was slower than the Core i5-6600 quad-core chip clocked at 3.30-3.90 GHz; but faster than the i5-6500, clocked at 3.20-3.60 GHz. The fastest chip in the table is the i7-6700K (4.00-4.20 GHz). The reviewer still notes that Ryzen has a decent IPC gain unseen from the AMD stable in a while.



In the final segment, the reviewers tested the power-consumption of the processor. AMD rates the TDP of the Ryzen 8-core chip at 95W, which was desperately needed from a chip built on the 14 nm node. Here it was noted that Ryzen made a tremendous performance/Watt leap over the 32 nm FX-8370 "Vishera." It consumes 93W, just under the 96W consumed by the Core i7-6900K eight-core chip, and slightly more than the 85W consumed by the 22 nm Core i7-4790K "Devil's Canyon" quad-core chip. The 14 nm i7-6800K six-core chip draws 83W, and the quad-core 14 nm i7-6700K draws 62W.

In all, the reviewer concludes that Ryzen could give the DIY performance CPU market the stir it badly needed, and could give Intel a shake-down, but it boils down the pricing.

[SickBeast]

It looks great, actually.

[SteelCrysis]

http://www.digitimes.com/news/a20161226PD208.html
Things are looking better for AMD.

AMD's next-generation 14nm Ryzen series processors will be unveiled in the first quarter of 2017 and the new platform will be officially released at the end of February and enter global mass shipments in March, according to sources from motherboard players. Optimistic about its high price/performance ratio, motherboard players have been rather aggressive about placing orders for related motherboards recently and some even believe demand for the new platform will increase AMD's share in the worldwide desktop processor market in the second quarter of 2017.
...
The sources pointed out that they have started receiving high-end X370, mainstream B350 and entry-level A320 chipsets from AMD and are now in the final stage of finishing related testing and motherboard designing.

[ocre]

Thanks for all the links SteelCrysis.

I am really am surprised by the info. Zen seems better than I would have guessed. I am though, kind of afraid to let this info excite me. Its supposed to be an engineering chip that is clocked at 3.1 ghz with very little boost. This would mean very substantial ipc gains. There is a lot left to ponder. Like, what is the clock speed characteristics and limitations of their 14nm node. As this is a very huge element in how Zen may compete.

The data presented is very positive. And the gap in IPC appears to be closing. What a great thing that would be. I must still admit that i am not without reservations. I am fearful of hype being built up. These are limited test and its all unofficial to boot. I would hate to see hype once again come back to bite another one of AMDs big and long awaited releases. So i choose to try to contain all of that until there are some real and official data.

See, even if these are real benches in a legitimate environment (where nothing is artificially gimped or scales purposefully tipped), there are things that must be considered. Will the new chip be consistently capable of distributing load across cores in all cases? Will there be stalls and wasted cycles? How about prefetching, cache, and branch prediction?
Its one thing to do well in canned multicore test or a very predicable data crunch that deals with blocks of data that break down evenly in a perfectly predetermined manner. But to make use of many cores is much more difficult in random work loads.

There are so many areas where intel has been making improvements that dont shine so much in multicore benchmarks dealing with large chunks of static data that is evenly broken down and totally predicable. With such a small set of benchmarks here and the lack of specific results, its just so limited and impossible to see any detail. And that is one of my larger concerns here. Why is that so? Why do we have these results and why are they so vague? Jumbled together to offer no resolution. Just some averages of averages

i dont know.

So, yeah. I am truly starting to feel excited for zen and Amd. But, i feel have to keep my excitement down. I guess it wont be a bad thing to hold back my expectations. If its great, i will still have to wait. We all have to.
But the seeds are now planted. And its gonna spread all over. RyZen now has bigger shoes to fill. People are now declaring Zen to no be a Sandy or Ivy like IPC. Its now being touted as better than Haswell on many forums. And its gonna have to meet that mark. Which i truly hope its at least close..cause thats more than i ever expected. if its better, bring it

[SteelCrysis]

https://www.techpowerup.com/229103/amd-a...m-c-tested
Just to be clear, this isn't Zen.

German PC enthusiast "Crashtest" clinched a sweet combo of an AMD A12-9800 "Bristol Ridge" socket AM4 APU with an ASUS A320M-C entry-level micro-ATX motherboard, for 200€. Pairing it with 8 GB of dual-channel DDR4-2133 memory, the platform was put through the AIDA64 test-suite. In the memory front, the platform performs on-par with older platforms at comparable DDR3 bandwidth. The K15.6 integrated memory controller isn't producing the kind of memory bandwidth as the Core i7-6700K with dual-channel DDR4-2133 memory from AIDA64's internal reference bench table.

In the CPU-related tests, the APU has about the same performance as its predecessors, such as the A10-7850K. The chip features two "Excavator" x86-64 CPU modules, making up four cores, and is clocked at 4.20 GHz. There are performance upticks seen in tests such as Hash and VP8, where the chip likely benefits from new instruction sets.

[SteelCrysis]

https://www.techpowerup.com/229204/amd-a...pabilities

But no, bar the name, it doesn't have almost anything to do with AMD's renowned Freesync. Instead, according to Videocardz, AMD are apparently expanding the Freesync functionality to more than just an adaptive synchronization technology - it might eventually become a software stack unto itself, embedded within AMD's drivers. This would be a smart move from AMD, since they would be taking advantage of Freesync's brand and name recognition on the market as a way to promote new features.

And what does this "Freesync 2" amount to? Basically, it stands as a way to reduce input lag from HDR processing of an image, by substituting the entirety of the HDR transport and display tone mapping usually needed to render a HDR image by a faster, driver level, Freesync 2 transport (which apparently already includes display tone mapping before the image is sent to the monitor), thus reducing that tone mapping processing time. Another part of the Freesync 2 update is the ability to instantly switch between display modes (SDR and HDR). AMD is also touting Freesync 2 to display "over 2x perceivable brightness and color volume over sRGB", though that honestly looks like more of a byproduct of using an HDR source than a Freesync 2 achievement. We expect to have more information over the coming days.

http://www.tomshardware.com/news/amd-fre...33248.html

AMD expects FreeSync and FreeSync 2 to coexist, so you’ll continue seeing new FreeSync-capable displays even after FreeSync 2-certified models start rolling out. Any AMD graphics card that supports FreeSync will support FreeSync 2, as well.

[SteelCrysis]

https://www.techpowerup.com/reviews/Inte...ce/26.html
A very interesting finding:

We initially planned to get this review done and over with just two GPUs, the high-end GeForce GTX 1080, and the performance-segment Radeon RX 480. When processing our results, we noticed an interesting trend that the RX 480 was posting bigger gains due to increased CPU speed than the GTX 1080. This phenomenon is more visible in the lower 1080p resolution, than in 1440p or 4K (where the RX 480 doesn't provide playable frame-rates in most games anyway). We initially dismissed this as behavior a performance-segment GPU designed for lower resolutions would exhibit. We then pulled out a GeForce GTX 1060 6 GB, a comparable graphics card from NVIDIA to the RX 480, and noticed that such gains were not noticeable.

From this we can conclude that the RX 480 has a slightly higher CPU overhead than the two NVIDIA cards. Graphics drivers need the CPU to crunch some serial processing loads at various stages of the graphics rendering pipeline, before they get fed to the GPU. Previously AMD's driver overhead was a vague assumption backed by very few benchmarks; this article is conclusive evidence, that the AMD graphics ecosystem definitely has more CPU overhead than the one from NVIDIA.

[SteelCrysis]

https://www.techpowerup.com/229339/amd-r...-ghz-boost

The folks at hardwareluxx managed to get some quality alone time with AMD's Ryzen demonstration boot at CES 2017, and it has to be said they used their time well. They managed to bring up Windows' System page, as well as its Device Manager, which seemingly confirmed that the Ryzen sample at use, though an engineering sample it was, was set at 3.6 GHz base clock with the capability to boost up to 3.9 GHz on a whim - up 200 MHz from the base clock speed of the sample used at AMD's New Horizon Event, where even at those speeds, the chip was shown beating an 8 core, 16 thread i7 6900K. You can see those clocks at the below screenshot, where "1D3601A2M88F3_39/36_N" (the code for the engineering sample Ryzen chip) makes it clear this is an F3 stepping processor, with the 39 referring to the boost clock, and the 36 referring to its base clock.

This goes right into AMD's claims of 3.4 GHz being the lowest frequency a Ryzen consumer processor would carry. It seems AMD is quickly galloping towards the finish line here, and as Lisa Su said at the New Horizon presentation, Ryzen chips can only improve until their promised launch, with an already rumored F4 stepping of the processor carrying a rounded-up, 4 GHz boost clock.

[SteelCrysis]

http://www.pcworld.com/article/3155129/c...years.html

AMD spent four years developing its Ryzen microprocessor, and that’s the same length of time the company expects it to live on in its ongoing war with Intel in the PC processor space.

Mark Papermaster, AMD’s chief technology officer, confirmed the four-year lifespan in a conversation here at CES 2017 in Las Vegas, though he declined to discuss specifics.
...
With everything riding on Ryzen’s launch, AMD isn’t taking chances. Jim Anderson, senior vice president and general manager of AMD’s Computing and Graphics business, told PCWorld that Ryzen chips will be available from day one. “We’re not going to do a paper launch,” he said, referring to a “launch” where customers have to wait weeks or months for the products to actually arrive. “We’ve done that before. We’re not going to mess with it.”

While Anderson’s responsible for bringing Ryzen to market—“you don’t have any idea how many hours I and my team have spent on this,” Anderson said—it’s Papermaster who has to think of the future. When asked how long Zen would last, compared to Intel’s two-year tick-tock cadence, Papermaster confirmed the four-year lifespan and tapped the table in front of him: “We’re not going tick-tock,” he said. “Zen is going to be tock, tock, tock.”
...
AMD’s last Bulldozer architecture debuted in 2011 and persisted through the Piledriver, Steamroller, and Excavator updates. If Papermaster’s words are to be taken literally, it seems AMD plans to iteratively improve its Ryzen chips through an additional three generations, about one per year.

What those improvements will be, of course, is anyone’s guess. But Papermaster said he’s a believer in architecture improvements that go beyond simple manufacturing, something he’s previously referred to as “Moore’s Law Plus.” One of the things AMD has tried to do with the Zen architecture is “go wide,” increasing the number of instructions that the chip can process, and Zen uses predictive tables and machine learning to improve its ability to ask and process data before it’s actually needed, known as branch prediction.

Operationally, though, the next step is putting Ryzen in the hands of journalists and customers. AMD’s already set the stage for the latter, revealing over a dozen motherboards and 16 performance PCs already committed to Ryzen. The company has also revealed that every Ryzen chip will be overclockable.

On the horizon, of course, lurks Intel’s 10nm Cannon Lake, most likely due this fall, and the still somewhat mysterious Optane technology. That’s of little concern at AMD, though. For now, the company’s focused on releasing Ryzen, and it’s full speed ahead.

[SteelCrysis]

Here are the systems from AMD's earlier Zen demos:

[SteelCrysis]

Does this Vega card only needs a single PCI-e power connector, or am I just looking at 2 PCI-e connectors spaced far apart on the PSU?

[SteelCrysis]

https://www.techpowerup.com/229540/amd-b...t-disabled

With AMD's Ryzen chips launch being ever closer to us, details about its product line - which still remain mysterious enough - eventually begin to slip. Reportedly, AMD's entry-level Ryzen chips - the SR3 line of processors, if previous leaks ring true, will be made up of 4-core processors with AMD's SMT (Simultaneous Multi-Threading), the equivalent to Intel's HT (Hyper-Threading) disabled. These will be, apparently, true 4-core processors, without any additional logical processors exposed by SMT.

If reports about AMD's line-up being composed of 8-core and 6-core processors, then with this news, we can now theoretically paint the numbers on AMD's Ryzen line-up. As it stands with this new information, it could be composed of entry-level four-core parts (under the SR3 product stack, and a base clock of 3.4 GHz at the minimum for any Ryzen-based part, according to AMD); a midrange six-core, twelve-thread part (under the SR5 moniker; I don't figure AMD would disable their much-lauded SMT on this six-core part); and the top-of-the-line, SR7 8-core, 16-thread chip we've seen in so many benchmarks and leaks.

I find it strange that AMD would cut the SMT out of any of its processor lines, though - at least, from all of its processors. My educated guess would be that AMD is planning to release a special-edition part (or a specific part number) just like Intel does in its i3, i5 and i7 product lines to differentiate between multiplier-locked (non-K processors) and multiplier-unlocked (K processors, such as the i5-7600K). Though, with all AMD Ryzen processors having an unlocked multiplier, like the company has often announced, this differentiation might be between SMT-disabled and SMT-enabled chips - perhaps with AMD bringing back their Black Edition line of processors for this particular use-case. It just seems strange for AMD to shed one of their vaunted technologies (which would allow them to improve their performance at little to no cost added) completely, considering the comeback the company is planning to accomplish.

https://www.techpowerup.com/229537/amds-...-2017s-gdc

It would appear as if AMD has seemingly (and not purposely) given us some lights as to when their hotly anticipated Ryzen chips will hit the wild - and some of our wallets. While they haven't specifically given an exact date, the talk to be given by AMD at the annual Game Developer Conference (GDC) reads: "Join AMD Game Engineering team members for an introduction to the recently-launched AMD Ryzen CPU followed by advanced optimization topics". The "recently-launched tidbit is the most important, since it places Ryzen's launch necessarily before the end of GDC - which will happen between February 27th and March 3rd. AMD has, in the meantime, altered the original headline by cutting the "recently launched" tidbit altogether.

It would seem that GDC wouldn't be the best stage from where to announce AMD's upcoming star - they surely would want it to be a star of its own show, instead of competing for attention with all the news, events and announcements that are bound to come out during both GDC and the Mobile World Congress, which happens at the same week. As such, I believe we should take March 3rd as the end-date at which Ryzen has already been effectively launched, though I would expect AMD to host its own event even before February 27th. And with AMD's own Jim Anderson, senior vice president and general manager of AMD's Computing and Graphics business, saying that AMD is done with paper launches, that would mean retail availability for AMD's Ryzen processors before that February 27th date.

[SteelCrysis]

https://www.techpowerup.com/229550/amds-...-2017-leak

According to WCCFTech, AMD's next-generation Vega architecture of graphics cards will see its launch on consumer graphics solutions by May 2017. The website claims AMD will have Vega GPUs available in several SKUs, based on at least two different chips: Vega 10, the high-end part with apparently stupendous performance, and a lower-performance part, Vega 11, which is expected to succeed Polaris 10 in AMD's product-stack, offering slightly higher performance at vastly better performance/Watt. WCCFTech also point out that AMD may also show a dual-chip Vega 10×2 based card at the event, though they say it may only be available at a later date.
...
AMD plans to roll out the "Navi" architecture some time in 2019, which means the company will use Vega as their graphics architecture for two years.

[SteelCrysis]

https://www.techpowerup.com/229740/amds-...-onslaught

But as the "out with the old, in with the new" adage still stands, retailers are now clearing inventory of their Radeon Pro Duo graphics cards, sometimes offering almost 50% off from the original launch price of $1499. Newegg, for example, has the card for $799 on both their North American and Asia Pacific online stores.

[SteelCrysis]

http://www.dsogaming.com/news/bethesda-m...b-of-vram/

First things first. According to Bethesda, these are the official requirements for Fallout 4’s upcoming High-Resolution Texture Pack.

Recommended PC Specs
-Windows 7/8/10 (64-bit OS required)
-Intel Core i7-5820K or better
-GTX 1080 8GB/AMD Radeon RX 490 8GB
-8GB+ Ram

What’s really interesting here is that NVIDIA’s GTX980Ti is not listed in there. The GTX980Ti is packed with 6GB of VRAM and from the looks of it, it won’t be able to offer an optimal experience with this high-resolution texture pack.

[SteelCrysis]

http://wccftech.com/sk-hynix-hbm2-radeon-vega-2017/

There have been slight modifications to the catalog. Previously, SK Hynix had listed both 1.6 Gbps and 2.0 Gbps HBM2 for availability in Q4 2016 but that doesn’t seem to be the case anymore. Currently, only one product has been mentioned. The specific part number is “H5VR32ESM4H-H1K” and comes in the 5mKGSD package. This is a part holding a capacity of 4GB on a 4-Hi stack that operates at 1.2V. The package operates at 1.6 Gbps with a total bandwidth output of 204.8 GB/s.
...
Some calculations reveal that the Radeon Vega graphics card with HBM2 will feature a lower bandwidth than the competing enthusiast product. The card will be considered flagship by AMD and NVIDIA’s flagship right now is the Titan X (Pascal). This card features a memory bandwidth of 480 GB/s on a GDDR5X interface that carries a 12 GB VRAM buffer.

We know this as fact that Vega GPU comes with two HBM2 stacks that are 4-Hi configured and hold 4GB capacity. This would indicate that the card would feature up to 8 GB VRAM until higher density memory or more number of stacks are available on the HBM2 products. This also gives us a total bandwidth of 409.6 GB/s which is lower than AMD’s first generation HBM product, the Fury X, which operated with a bandwidth of 512 GB/s.