AMD AMA - Starts at 12PM EST on Wednesday May 9

[caveman-jim]

Good day James,

I'd like to thank AMD for producing Threadripper and for enabling ECC on it. My Threadripper 1950X is an awesome CPU indeed. I'd like to take this opportunity to suggest some features that I would love to see in the future.

1) RDIMM/LRDIMM support that would make it possible for users to reach (or at least get much closer to) Threadripper's advertised maximum of 1TB, which as of now is woefully inaccessible with support being limited to UDIMMs.

2) A dual configuration of Threadripper would be very welcome.

3) A fully enabled, 32-core Threadripper with all dies populated (instead of the current 2 dies) that is made with the absolute top binned dies available would be an amazing chip and many people (myself included) would gladly pay a substantial premium for such a chip.

Thanks again for producing Threadripper...I hope to see it made even better in the future!

Thanks for your feedback and I'm very glad you like Threadripper. More to come on 2nd Gen Threadripper, later this year!

 

[A-KEY]

Dear James san

Thank you for your kindness.
Will AMD make any Ryzen H (=High Performance Mobile) series against Intel core i HQ series?
We want High Performance Gaming laptops as Alienware, Omen etc. and High Quality laptop as Dell XPS, HP Spectre etc. by Ryzen Mobile.
Please let us know.

Best regards,

 

[caveman-jim]

Thanks for doing this, James!

I have a few of questions:

1. There are rumors that Zen has an amazing wafer success rate. Any comment on Zen's prowess in the fab?

2. Any lessons learned from the issue of Raven Ridge compatibility on older AM4 motherboards? Does AMD plan on changing their chipset release strategy to allow for a better out-of-the-box experience for such CPUs? For example, Raven Ridge compatibility probably wouldn't have been as messy had first-time Ryzen buyers been able to buy a new B450 motherboard rather than a B350 with an older BIOS.

3. How much does the fabrication technology influence the architecture of your processors. For example, is Zen 2 designed specifically with Global Foundries' 7nm process in mind? Since you use TSMC for the the custom SOCs used inside Playstation and Xbox, are architecture tweaks a necessary component of working with their fab?

4. This one is probably more Radeon-related, but what are your thoughts on providing a dedicated graphics cache for future Zen APUs? Adding something like the Xbox One's 32MB of ESRAM might dramatically improve memory bandwidth for future Vega/Navi GPUs that come in the APU lineup.

1 - No, we don't comment on those details. Sorry.

2 - I think the key lesson learned was to begin a public awareness campaign, sooner. I think that the BIOS enablement strategy is better than new motherboard requirement. If you just look at the cost difference, $60-$160 new mobo or $10 for BIOS update service from your retailer (if they offer it), or waiting 1 week to get a free boot kit from AMD... I really don't think many people would prefer to pay the extra cash. It's the unpleasant surprise people were unhappy about, and a larger reaching consumer education campaign would have been a good investment. It's a lesson learned.

3 - The architectures are tuned to the foundry process as part of the productization of the design - the part of the process where the concepts and ideas are modelled into something real. And there's always a boundary of "whats possible" even across multiple foundries, but there's a lot of freedom for the starting point.

4 - Its an interesting idea, currently the focus of designs like "Raven Ridge" is enable thin and light notebooks, and scale down to the mainstream/entry space. ESRAM or other caches add complexity, cost, and ultimately reduce the total market applicability of a product. So we have to balance business needs with designs that can be enabled by adding a small dGPU to a design.

 

[R0H1T]

James, seeing as AMD plans to update Ryzen each year, are the APU's only going to be updated bi annually or in line with major uarch changes i.e. Zen2, Zen3 et al?

Secondly, any particular reason why Ryzen APU's won't do dual graphics, like in the days of Trinity or Richland?

 

[caveman-jim]

If double dipping is okay, I have another question.

1. This is about your potentially future value ULV or low power, sub-15W options, akin to Intel’s Atom and the non-Core-based Pentium product lines. A couple years ago, rumor and leaks had it that a low power line of processors, codenamed Banded Kestrel, was in the pipeline. Now, in spite of Intel Atom’s well-known failure in mobile, fragments and derivatives of it still live on. The resulting market thrives in the form of low cost laptops, tablets, and 2-in-1’s from the likes of Chuwi, Teclast and other far eastern outfits as well as more specialized designs such as the GPD Win and GPD Pocket. If you cannot say anything else, does AMD still plan on addressing this market segment? Beyond that, if so, will it be with Banded Kestrel? Or perhaps does it instead plan on doing so with products using cut down Raven Ridge dies to serve the same purpose?

Can't speculate on future products, sorry.

 

[caveman-jim]

James, seeing as AMD plans to update Ryzen each year, are the APU's only going to be updated bi annually or in line with major uarch changes i.e. Zen2 or Zen3 et al?

Secondly, any particular reason why Ryzen APU's won't do dual graphics, like in the days of Trinity or Richland?

Ryzen products will typically follow an innovation - optimzation cycle. So you've seen 1st Gen Ryzen CPUs, that featured the new Zen core architecture using 14nm, which was the innovation cycle; and now 2nd Gen Ryzen CPUs, with Zen+ and 12nm - that's the optimization phase. Similarly, we've launched the Ryzen with Radeon Vega Graphics (also known as Ryzen Mobile) which are an innovation cycle for the processors with graphics style products.

For dual graphics on Ryzen, and in fact on A-series in AM4, we decided not to implement dual graphics. As we move to modern APIs and technologies like explicit multi adapter, better expriences can be had by developers specifically leveraging the power of Radeon graphics in the socket if so desired.

 

[caveman-jim]

Dear James san

Thank you for your kindness.
Will AMD make any Ryzen H (=High Performance Mobile) series against Intel core i HQ series?
We want High Performance Gaming laptops as Alienware, Omen etc. and High Quality laptop as Dell XPS, HP Spectre etc. by Ryzen Mobile.
Please let us know.

Best regards,

Hi A-KEY san

The most recently announced model we have is the Ryzen 7 2700U, which is a very powerful processor for ultra-thin and light notebooks. We'll see more of these processors in notebooks later this year. More info to come!

 

[dcominottim]

Amazing AMA overall, James. One final question from me, please!

I have always craved to learn a basically overview of all the steps involved in getting a CPU released to market. For instance, around 2 years of chip design, around 1 year of validation, and so on. I would love to have the name, basic description, and average duration of all the major steps. It’s truly a pity I’ve never seen a full article about this — just vague comments about “tape in”, “tape out”, etc.

 

[zebrax2]

Thanks James for the AMA

Is it possible to get a short timeline of the Zen from inception to release just so we can get a general idea on CPU development works (at least from AMD side).

 

[A-KEY]

Hi A-KEY san

The most recently announced model we have is the Ryzen 7 2700U, which is a very powerful processor for ultra-thin and light notebooks. We'll see more of these processors in notebooks later this year. More info to come!

Dear James san

Thank you for your prompt reply.
We are looking forward to seeing your new informations.

I'm sorry for that I would like to add a question.
I have a Photon 2 now, so I want a next tablet powered by Ryzen Mobile.
Do you know any successor of something?

Best Regards,

 

[BeepBeep2]

Thank you James for answering our questions.

Is Raven Ridge still on GloFo's original 14nm process that Summit Ridge was produced with, or run with some / all of the "12nm" optimizations, or totally different optimizations than the "12nm" process has? I have seen Raven Ridge being referred to as "14nm+", and Raven Ridge A0 (CPUID 00810F00h) ES parts seemed to be in testing phase around the same time as Summit Ridge A0 (00800F00h) ES parts (late 2016), based on CPU support information contained in AMD platform BIOSes, though the final stepping and retail product for Raven Ridge (B0, 00810F10h) were released much later.

It seems that Raven Ridge, even including the desktop lineup, consists of very low leakage parts, and hit a voltage-frequency wall earlier, whereas Pinnacle Ridge, at least in the case of the 2700X, have higher leakage current than Summit Ridge.

It also seems that while Pinnacle Ridge CPUs exhibit higher temperature / power consumption at a given voltage, it is completely offset (and then some) by a shift in voltage-frequency curve with lower voltage required for given frequency, and higher frequencies achievable.

(1)... depending on what the case really is, could you clarify the differences between this "14nm+" and "12nm"? Are the observations above accurate abstractions?

(2) A lot of enthusiasts are overclocking their Ryzen CPUs, with many users hitting 4 GHz and above for all-core overclocks. For Summit Ridge, Robert Hallock mentioned that for long-term use, it would be advisable to stay under a maximum of 1.425v, and mentioned that 1.1v SoC / uncore voltage would be helpful for memory overclocks. Are you able to give guidance on the voltage tolerance of Pinnacle Ridge, considering its different leakage characteristics?

 

[Ryzenenthusiast]

Hey James,

I would like to ask couple of questions about the design of the Zeppelin die.

First of all architecturally is it possible to add 1 more CCX to the CPU die or increase the CCX Core count to lets say 6 without breaking something on the architecture?

Second Question: Are you considering to move into a single CCX design for Desktop products on the future iterations of Zen architecture?

Third: Global Foundries has 2 nodes for 7nm. One is ultra high performance with more power consumption(Namely IBM node). It can work on 5+ Ghz easily, but it consumes a lot of power. The other is extremely power efficient which lets a processor like 2700U to run around 5W TDP. Are you considering to use different nodes for different product groups in the future? Or will the "one size fits all" strategy continue?

Fourth and last: Zen core does not have an AVX512 unit, for that reason, your IPC is vastly inferior to the Intel's architecture on this area. I know it uses a lot of power. Was lowering the power consumption of the cores your main goal for excluding AVX512 unit from the core itself(since GLOFO's 14nm is not able to match Intel's 14nm in terms of perf/w) and is there a possibility for you to include an AVX512 unit on the next iterations of Zen core?

 

[cbn]

In the past AMD offered both 15W and 35W APUs for mobile.

However, as shown by the specs above there doesn't appear to be much difference in performance comparing the 15W and 35W on paper because only the CPU clock (not under load by the iGPU) is shown.

In contrast, Independent testing found here for Carrizo @ 15W/25W vs. 35W/42W did show a very significant difference in clockspeed when both the CPU and iGPU were both active.

15-25W:

35-42W:

So to help promote better understanding of products does AMD have plans for a more comprehensive labeling on future higher and lower watt APUs? CPU and iGPU clocks when both are simultaneously loaded?

 

[ET]

Thanks for all the answers, James.

It's great that AMD promised AM4 socket compatibility until at least 2020, but the BIOS situation is problematic, and even though the boot kit solution did work for me, it's still somewhat of a hassle on both the user and AMD's side.

Some motherboards offer the option to flash BIOS without a CPU, but that's a rare feature available only at the high end.

Has AMD given any thought to having some base functionality which will enable future processors to at least be able to flash a BIOS via USB, even if the rest of the chip and interfaces require an update to work? Using the ARM core, having a standard USB controller, ... I don't know, I'm not familiar with what's needed to achieve this, or what prevents new chips from working with an older BIOS, but surely it's not impossible.

 

[Rancor]

Hi James, does AMD devote many resources to Linux development for Ryzen processors? A couple of marginality problems have held me back from purchasing in the past. Others like myself are waiting to hear if Zen+ is affected by the Linux soft lockup bug. It would be great if AMD could be more forthcoming about what's being worked on. Hopefully generating even more trust and enthusiasm for the platform.

 

[Minsky]

Hi James

Heterogeneous computing was considered an important element in the days of Bulldozer especially around the KaverI launch. With the advent of Zen is heterogeneous computing still considered important and if so then what kind of future are we looking at?

 

[CatMerc]

Is having to support the same platform year after year holding you back in some ways? Would it be easier to change motherboard specifications with every new generation you release?

 

[NTMBK]

Hi James,

Is there any news about AMD's HIP platform coming to Windows? Speaking as a CUDA developer, I'd love to be able to try porting our code to run on AMD cards, but we haven't got resources or desire to rewrite everything for OpenCL- especially not the limited subset of OpenCL that is supported by both AMD and NVidia. And supporting two different codebases is out of the question. HIP would be great, but sadly our software stack is tied to Windows.

Thanks!

 

[wahdangun]

Hi James,

Since amd gpu/cpu was good at mining, will amd create more specific instructions set for crypto ?

And will amd embrace mining community ?

 

[ThatBuzzkiller]

Hi James,

Is there any news about AMD's HIP platform coming to Windows? Speaking as a CUDA developer, I'd love to be able to try porting our code to run on AMD cards, but we haven't got resources or desire to rewrite everything for OpenCL- especially not the limited subset of OpenCL that is supported by both AMD and NVidia. And supporting two different codebases is out of the question. HIP would be great, but sadly our software stack is tied to Windows.

Thanks!

I too am very curious about ROCm for Windows ...

I believe ROCm support for Windows was supposed to be delivered during last year at the end of the last quarter according to an AMD developer which leaked that detail in a comment in their developer blog ... (we could see Windows support coincide with the release of ROCm 2.0 either by the end of this quarter or conservatively by the end of the next quarter)

That being said, I think it's hinted that James specializes in technical marketing for their CPU team so I am extremely interested in knowing if their engineers are considering hardware extensions such as AVX-512 and transnational memory support such as Intel TSX or if there are patent roadblocks due to the end of their cross-licensing agreement with Intel that could potentially prevent their implementation ...

 

[Ket_MANIAC]

Hello James,

First of all, thank you for this opportunity to allow us to interact with you and thus with the company. I believe a company that interacts and listens to its customers will always have goodwill in the market. And yes, a big AMD fan. Typing this right now on my 7 year old Phenom II X4 955 OC'd to 3.7 GHz. Still kicks azz.

Anyway, to the questions now. I have been following Ryzen like an eagle since last year and being a system integrator, have recommended them over Intel at any price point and sold quite a few systems since. My questions are:

1) Considering the fact that the Raven Ridge APU's have been out now for almost 3 months, why didn't AMD shift the B450 board launch ahead? This could have in theory mitigated the backlash the APUs had to face because of lack of BIOS ready boards on launch day. I myself make sure to update the BIOS of boards when a customer demands so but many retailers here in India have no clue and still have old boards on stock which are incompatible with the 2200G and 2400G. Fantastic products that they are, the average public has no clue about the things such as BIOS updates and when an uninformed buyer buys so and finds out the CPU and board are incompatible they blame the company which produces negative feedback due to word of mouth. Couldn't this be solved with better marketing and a note that buyers need to buy latest boards or boards with latest BIOSs' in a flyer inside the CPU box or like on the CPU box.

2) I just love the Precision Boost Overdrive function and the way AMD has tweaked the 2000 series to the edge out of the box. It is sad that reviewers and even enthusiasts don't know much about it and are trying to manually OC their CPU while AMD has a fanatastic Auto OC which performs better. I read somewhere that this feature is still in beta phase? Is it true and if yes, when does it become official?

3) Again, why no B450 boards yet? Are more products coming up for the 2000 series lineup?

4) Last and most important question, why is there still a lack of Ryzen powered laptops? The options are few and far and if they are, they are either hobbled by poor OEM decisions like Dell has been doing with its AMD laptop series or lack of driver support which is creating negative views for otherwise good products like the HP Envy 360. Also, why isn't talking to these OEM's about it? Many websirtes have claimed its the OEMs who are deliberately trying to make AMD products look bad. Dell has a complete lineup of utter garbage configurations which are poorly placed against their Intel counterparts. I was hoping to get myself a Ryzen 5 2500U laptop but pickings are slim.

5) Not a question, simply love the work you guys did with the IMC of the 2000 series. Can't wait for ThreadRipper 2 and the 2800X that you guys have been saving up.

Thank you for your patience and giving us this opportunity once again.

 

[ET]

And will amd embrace mining community ?

I hope not. Mining is a stupid idea so hopefully AMD won't be encouraging it.

 

[CatMerc]

AMA wa yesterday, guys. You missed it.

Nope, it's open for 24 hours. Its been 20.

 

[bearmoo]

We spoke with a lot of media outlets, including Anandtech, when we briefed them as part of the launch of 2nd gen Ryzen. We have new designs coming from every global OEM this year, so you'll see a lot more choice in notebooks!

Maybe a quick follow up question: can you divulge what silivon process Vega Mobile is made in? I believe eatlier roadmap presented by Mark Papermaster put Vega in the 2018 12nm refresh category so I am wondering if Vega Mobile is the one?

 

[eva02langley]

Anandtech just did a deliding article on the 2400G. What is the cost of using solder over paste?

It seems to be a negligible price for the benefit.