Amd Ryzen Threadripper & X399 MegaThread! FAQ & Resources

Page 7 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.


Note that I am discussing power, not efficiency.



14LPP is optimized for sub 3GHz clocks. The efficiency is worse near 4GHz. That is why power consumption skyrocket above this point and the system becomes unstable.



My main argument was based in power measured and a public remark on how the same people that accused the i9 of consuming lots of powers remains silent now that TR consumes more power than the i9.

As a bonus I also computed (using the laws of physics) how much power a TR @4.6GHz would consume.
 


SKL-X is the successor to the Broadwell-E family as has been known and expected by many years. You mention der8auer often. This is what he said:

From my point of view Skylake-X is a pretty impressive CPU. We now have have a 10-core that's a lot cheaper than the Broadwell-E before and it clocks a lot higher. Even with the stock Intel paste I was able to reach 4.8 GHz on the CPU (i9-7900X) using Corsair 280 (mm) AIO. So there is still some headroom, I guess if you use a custom new water cooling you might be able to hit 4.9 GHz on a very very good chip without delidding. So after this test I delidded the CPU and replaced the stock TIM with liquid metal and this helped me push the CPU with an AIO to 5 GHz.

So we had Broadwell-E before who could run like 4.3-4.4 GHz, it cost 1700€. Now we have Skylake-X which is a lot cheaper and we CAN push it to 5 GHz, so what's all this negative press about? I don't really understand it. So from my point of view this is a very impressive CPU, so we have very high single-thread performance on the 10-core and also high MT performance, which we didn't have before in a Broadwell-E (probably comparing to mainstream).

Despite all the bad press, the people posting FUD in forums and rants on pages of Intel products [1], and the biased reviews from the tech media, the indisputable fact is that SKL-X is selling very well. Amazon ranking for new releases chips

Our best-selling new and future releases. Updated hourly.

#3 TR 1950X
#4 i9 7900X
#5 i7 7800X
#6 i7 7820X
#7 TR 1900X
#10 TR 1900X
#11 i9 7920X


[1] https://www.amazon.com/gp/product/B0745486S1/ref=s9_acsd_simh_hd_bw_bxcb_c_x_2_w?pf_rd_m=ATVPDKIKX0DER&pf_rd_s=merchandised-search-11&pf_rd_r=52F601VWB4WCG4A9G553&pf_rd_t=101&pf_rd_p=1381e980-855e-537b-94df-b4f4d642d410&pf_rd_i=229189
 


I see. I think there is something more than just a typo in their database. It seems that the rendering code used in that benchmark list is only designed for CPUs. However, note that they are improving on adding GPU-support to the code. They have also a hybrid version of the renderer and luckily for us they compared CPU times and GPU times

https://labs.chaosgroup.com/index.php/rendering-rd/understanding-v-ray-hybrid-rendering/
 




The link below is based on SALES!

Amazon Best Sellers
Our most popular products based on sales. Updated hourly.



That link you are using to make people into thinking that it's proof they are best seller based on sales! That link states Amazon Hot New Releases
Our best-selling new and future releases. Updated hourly.

Again you offer the same link to a 7900X retailing for $1,499.99 on sale for $999.99. But on closer inspection you can see everyone is buying the 7900X that is being discounted and selling for $962.99! What you are doing is the definition of bias! Click here for the previous post unraveling this misinformation.

You right I miss quoted you! You said:
I am surprised than the SKL-X chips are selling so well despite so many biased reviews and people paid for posting negative comments on forums and stores. Check the nest link. This guy didn't write a review of the Intel chip, he wrote a rant to try to increase sales of AMD chips
You believe there is a conspiracy happening! I agree with Yuka, I'm not going to continue feeding your need to validate your false belief system!

Definition:
Confirmation bias is the tendency people have to favor facts or arguments that confirm the beliefs and positions they already hold. The extreme form of this bias is referred to as “belief perseverance” when people hold onto their beliefs even after they've been proven false.
 


The link that I have given is ordering the chips for best sales: "Our best-selling new and future releases. Updated hourly".

The only difference between your link and mine is that my link only counts sales of the newest chips in the store. That is the reason why my link gives the first and second position in sales to the R3 1200 and R3 1300X, whereas your link gives the first and second position for older Kabylake chips i5-7600k and i7-7700k.

This new releases chip best-selling list shows that the entire SKL-X line is selling rather well worldwide despite your repetitive pretension that SKL-X chips are not selling. I read in another forum that "At Mindfactory 175 TR chips have been sold compared to 1040 SKL-X. And that's an AMD retail partner."



Stop saying that the launch price was or $1,499.99 because it is not true. The official launch price was up to $999. Even AMD gives the $999 launch price in slides.

amd-ryzen-threadripper-in-techday-11-638.jpg

amd-ryzen-threadripper-in-techday-12-638.jpg


And all reviews of the i9 mention the $999 price launch: "For the time being power users will have to make do with the 10-core, 20-thread Core i9-7900X, priced at $999 and introduced as the first ever Intel Core i9 processor."

Your claim that the i9 is having massive discounts after TR launch is not true. Amazon is only offering today a 4% discount over the launch price of $999

(962.99 - 999) / 999 = 3.61
 



I concur. Wise decision!
 


AMD is making some really smart calculated moves here on their spare time with this pet project! hahaha Destroying Intel's product pricing schemes!
 
AMD Ryzen Threadripper: The Fascinating Story Behind The Processor That Beat Intel
Antony Leather , CONTRIBUTOR
SEP 5, 2017 @ 07:33 AM


Although computer enthusiasts waited a decade for serious competition to Intel in the desktop processor market, 2017 saw not only AMD's successful mainstream Ryzen processors arrive in March, but also the company's first high-end desktop processor in the form of Threadripper too, which launched in August (see my review of them here).

In many ways, it's even more successful than Ryzen mainstream processors. For $1000, you get more cores and better performance in many tests with Ryzen Threadripper 1950X than Intel's equivalent and highest performing desktop CPU - Core i9-7900X. Where AMD is still a little slower, for example in some older or lightly threaded games, is less of an issue as many people will use Ryzen Threadripper purely for its multithreaded performance. It has to be said that there aren't many reasons to choose Intel in the high end desktop market at this price point.

The battle goes on, of course, with Intel due to release processors with even more cores over the coming weeks, but just where did this somewhat unexpected processor come from? It might surprise you to learn that Ryzen Threadripper wasn’t originally part of AMD’s plans, which back in 2015 only included Ryzen and the EPYC server CPU.

To delve deeper into Threadripper's origins, I spoke to some of AMD’s big guns as far as Threadripper was concerned; AMD Senior Vice President and General Manager Jim Anderson along with Corporate Vice President of Worldwide Marketing John Taylor ,who have both been heavily involved in the Threadripper project; also Sarah Youngbauer of AMD’s communications team, plus James Prior, who’s AMD Senior Product Manager and closely involved in the Threadripper inception.

Forbes: What can you tell us about Threadripper origins and how it came to be?

Sarah Youngbauer:

There’s a unique story surrounding Threadripper, but one that gives testament to the spirit that’s emerging in the company as we go through this transition since 2014 when we announced our Zen architecture. It’s not really a story of roadmaps and long-term planning or huge R&D budgets - it’s a lot more personal than that and stemmed from a skunkworks project and a small group of AMD employees who had a vision of a processor they’d really want in terms of a high-performance PC.

They worked on it in their spare time and it was really a passion project for about a year before they sought the green light from management, which is quite unusual – it was something they really cared about. The result, several years later was Ryzen Threadripper, which is the world’s most powerful desktop processor. Without this group of people Threadripper may not have happened.

Continued from page 1

Forbes: James, you were part of the original team that came up with the idea. What’s your take on it?

James Prior:

Myself and a few others were in a very cross functional team that get together for various different projects and as we got the first hints of what the Zen core performance and efficiency were like and started looking at the internal roadmap, which is a constantly changing thing and noticed a gap between Ryzen and EPYC. Certainly, something that stood above Ryzen with more memory bandwidth, cores, PCI-E lanes. To get to this product, which sounded great to us as enthusiasts, we found we’d only have to change a few details. So we put together this skunkworks team where we had platform architects, people that deal with core design, business unit, marketing team, to work out how to use what’s already here and to go to the boss – Jim Anderson and say we’d like to do this. This was all happening in 2015.

There were lots of late nights, calls, Skype messages and between-meeting conversations where we’d see what progress was being made with Ryzen and EPYC that might be useful to Threadripper and quickly discuss ideas. The cardinal rule here was not to go against the grain. So the 20 or 30 of us spent more and more time working on it until eventually when Jim came on board from Intel and we found out he was an actual CPU enthusiast, this was perfect for taking the project forward.

My boss was sharing a taxi with Jim on the way to Computex 2016 and mentioned Threadripper to him then. He loved it, gave it the green light and found a way to get it on the roadmap. He believed in the idea, especially the way it leveraged existing technology we were using for Ryzen and EPYC, plus the fact as we were so excited about it that we’d already done a lot of the groundwork that would have been involved in the approval process anyway.
In fact, even though it was akin to a pet project up to this point and had no official business plan, the actual feasibility study that took place later concludes it was one of the best-planned products they’d seen in a long time. We’d been that passionate about it during those first few months so even though we were asking for an un-budgeted, un-scoped product to be added to the roadmap, he was willing to say yes to it. The feeling of validation from reading reviews in August was immense – so satisfying and we felt we really knew what the market needed. It was worth all the late nights and additional time we gave up during the initial stages.

Forbes: Jim, what was your reaction when you heard about their proposals?

I started at AMD in June 2015 and we were trying to get the graphics business accelerated, and in June 2016 James and his team brought this concept to me and said they’d been working on it in the background and said I might be interested. I thought it was fantastic! I was really blown away when we worked out we could actually build it – I came into the industry as a CPU architect and it was the kind of product I’d love to build.

I asked for two things – the specifications and also how quickly we could build it. I fell in love with it seeing the specifications, but the timeline was disappointing as it originally pointed at 2018 for launch. I immediately asked them to pull it in by about a year – so summer 2017.

They thought I was crazy, but what was really great is that they were passionate about the project. I’ve never actually told anyone this, but Threadripper never had a business plan – that might raise some eyebrows, but we were building it because we knew it was awesome, because we could and to make it best product we could, even the name had to be big. It’s certainly been one of the most fun products I’ve ever brought to market and I had a great time building a Threadripper PC with my son recently too.

Continued from page 2

Forbes: So, were Ryzen and EPYC the only two processors planned?

John Taylor:

That’s correct. The existence of the EPYC server CPU enabled us to create Ryzen Threadripper, but Threadripper was not part of the original 2017 AMD product roadmap. That only included the Ryzen 3, 5 and 7 ranges of 4, 6 and 8-core CPUs, plus versions of EPYC ranging from eight to 32 cores. Interestingly, some of the original team behind Threadripper joined AMD from the press, with a critical and appreciative eye for PC hardware from years of reviewing a wide range of products. This perspective was core to creating the Threadripper concept and product definition. It is very much a CPU made by enthusiasts, for enthusiasts. The story behind it hasn’t really been told before as we announced and launched Threadripper this summer in a very short period of time. Before May 2017, no one knew it existed, and by August it was in people’s hands, but it’s a very good example of the company AMD is today.
Forbes: There were obviously some key decisions made surrounding Threadripper and its launch, such as using the EPYC-size CPU socket and heatspreader. What were these, how do you think your technology allowed it to come together and what are AMD’s plans for its future?

Jim Anderson:

There were a lot of big decisions as I’m sure James and John will talk about in more detail, but one of the most stand-out for me, which is quite funny, is that I really wanted the retail packaging to be illuminated. Sadly, John and others told me the retail kits would miss the summer target I’d set if they included lighting, which was a shame so I had to give up on that, but as a compromise, the press kits were illuminated.

John Taylor:

Very true and a funny story. We really wanted to make the packaging better and different, though, then any prior CPU and do so for everyone who bought a Ryzen Threadripper, not just a special edition. We built a package for every Ryzen Threadripper CPU that befits the notion of unlocking overwhelming power. Then, with the press kits for reviewers, we went to town and delivered the illumination in a special case that lit up when reviewers opened the special case – all done on a very short timetable to make the August launch.
We will continue to build products for the Ryzen Threadripper Socket TR4 platform, including the just-released Threadripper 1900X 8-core CPU. All of this is possible only through the adage “necessity is the mother of invention.” We needed a method to efficiently bring together AMD and industry standard blocks of processing IP, to enable us to do more with our R&D budget and focus it on our differentiating high-performance cores. The Infinity Fabric our engineering teams invented to accomplish that also lets us create products like Ryzen Threadripper and EPYC.

Continued from page 3

James Prior:

We started off working out how we could leverage the existing product definition for EPYC and turn it into something else. We also considered working on making a larger die, but that would have put the launch time frame back two years. There’s not really a better example of utilizing the advantage of Zen core and Infinity Fabric to connect them than Threadripper, especially when you consider the time to market advantage and cost advantage. It makes it so much easier to manufacture and to define as well as test.

The efficiency in manufacturing came from the fact that we didn’t need entire wafer runs just to produce Threadrippers, we could use Ryzens and that was the big key in Threadripper’s success. It meant we didn’t need millions of dollars for this one design; we found a way to use our existing great product and make it even more powerful. You can consider the resulting size of the CPU to be a negative aspect – the size of the socket, the heatspreader and number of pins, but it turns out it was a blessing in disguise. We could use the same socket as EPYC and just re-wire things. We’d already defined all the supply and ironed out the issues for EPYC, and this made it really easy to persuade our motherboard partners for example.

Forbes: How did you decide on the features and specifications such as frequencies and memory access modes that can be configured in the Ryzen Master software?

James Prior:

Jim also wanted to make Threadripper the best of the best, which is why the CPUs have such high XFR speeds at 4.2GHz and the same all-core boost as the Ryzen 7 1800X, but on twice the number of cores. We’d also learned a lot about the impact of memory and latency on gaming, which is where the Ryzen Master software’s memory access modes come in. In testing different games, we found that leaving Windows to manage things would see tasks switch between each of the two dies in Threadripper CPUs. Some games used lots of threads, but even with games that didn’t this would still happen, and this introduced a degree of latency.

We looked at ways to constrain these tasks to an individual die and its two memory channels so you get the fastest possible memory access. In addition, some older games aren’t stable in any system with more than 10 cores or 20 threads so we wanted a way for those gamers to be able to run their games too. Similarly, some applications don’t always make use of more cores and threads so that’s why we introduced Legacy Mode, which actually disables one die to focus Windows and applications onto that single die for the lowest latency.

We also have the Local access mode, which keeps the second die active but fills up the local memory first, again lowering latency. So, this is where these modes came from. You don’t have to use these modes, of course, but we wanted to give people the option, especially those that want maximum gaming performance. You also have to reboot when you switch between them, but that’s because Windows really doesn’t like it when you start changing the core and thread counts on the fly – it’s not a limitation of Ryzen Master, but of Windows.
Forbes: The original roadmap obviously had quite a big gap between Ryzen and EPYC. Were there any other plans to launch a high-end desktop (HEDT) processor to compete with Intel here?

John Taylor:

If you go back a few years, our roadmap was focussed on building products like Ryzen 7 from our new Zen core, and we prioritized the desktop market for the first the simple reason that we felt enthusiasts had passionately waited for a new AMD high-performance desktop product for a long time. While Threadripper is a high-end desktop platform, Ryzen 7 was also designed to disrupt the HEDT market as its eight cores were only matched by Intel’s Core i7-6900K – an HEDT product.

However, we still left the leadership crown with Intel as it obviously had the Core i7-6950X, which is a 10-core product. AMDers like James and Jim realized that Infinity Fabric and the EPYC package allowed us to define a version of Ryzen that was even more powerful than Ryzen itself, and used the basics of the EPYC platform to take that ultimate performance crown.

As a result, I think we really surprised the industry, and while Intel has more CPUs coming in September and October, as we sit here today the Ryzen Threadripper 1950X is the most powerful CPU made for the consumer market.

Continued from page 4

Forbes: I’ve been impressed not just with Threadripper’s performance, but also its memory compatibility and stability out of the box, which was much better than Ryzen’s launch. Some teething issues were to be expected of course, but you seem to have learned lessons here.

John Taylor:

There was a teeth-cutting that we were going to have to work through with Ryzen’s launch as the first all new CPU core and platform in some time – a whole ecosystem had to be built around the platform. However, we worked hard to solve the problems, blogged and posted consistently about our progress, and provided regular BIOS updates for motherboards. We built this work in Ryzen Threadripper from the start, so the product launched with a strong stable ecosystem and high-speed memory support.
 


Marketing purposes article that even tries to rewrite the history

It might surprise you to learn that Ryzen Threadripper wasn’t originally part of AMD’s plans, which back in 2015 only included Ryzen and the EPYC server CPU.

It only mentions EPYC (SP3) and RyZen (AM4) and omits to mention Snowy Ownl (SP4).

As I have been saying for years, TR didn't appear in the roadmaps by 2015/2016. The only planned sockets then were AM4, SP4, and SP3:

AM4 --> single-die; dual-channel; 4/6/8 core (desktop)
SP4 --> MCM2; quad-channel; 8/12/16 core (microserver and workstation)
SP3 --> MCM4; octo-channel; 16/24/32 core (server and HPC)

Then the SP4 socket failed to get traction from the industry, which raised doubts about what do with all those 8/12/16 core CPUs and, in the other hand, Skylake-X was much more powerful than what AMD expected, and in a last minute change somewhat around 2016 AMD started to merge the SP4 and SP3 platforms in the new SP3r2 socket, releasing the 8, 12 and 16 core ThreadRipper. It is also about 2016 when AMD started to remain silent about the SP4 socket.
 


You should read the Forbes interview about how TR came about. It was not on roadmaps because it was only approved 12 months before launch, and it was kept secret.
 
HW News: X399 Outstripping X299, PCIe 5.0, 1900X CCXs
Gamers Nexus
Published on Sep 7, 2017
Talking about this last week's hardware news, including X399 vs. X299 sales, PCIe 4.0 & 5.0, and new peripherals.

[video="https://www.youtube.com/watch?v=3616WecsCBM&ab_channel=GamersNexus"][/video]
Speaking with vendors the word on the street is that looking at X299 vs. X399 the venders with whom we have spoken said that the first few days to the first week, depending on which vendor, of sales of the X399 products, related products surpassed the first month of Intel X299 products sales from those same vendors, not retailers, but actual manufacturers.
 


I did read it and that is why I know the article tries to rewrite the history and omits to mention the SP4 platform and the corersponding 8/12/16 core CPUs.
 


It omits to mention any platforms or codenames. Zen is not a platform - at least not in the way it's said in the article.

This is Forbes, not Hexus, Anand or SemiAccurate. idk why Forbes would deviate from a business/human centered story on TR to talk about an irrelevant scrapped product from an irrelevant scrapped product line. You expect them to talk about Zeppelin too?
 


The Forbes article claims (bold emphasis mine)

It might surprise you to learn that Ryzen Threadripper wasn’t originally part of AMD’s plans, which back in 2015 only included Ryzen and the EPYC server CPU.

The bolded part is not true, because RyZen is AM4 platform, and EPYC is SP3 platform and AMD also planned Snowy Owl which used the SP4 platform. The problem is not Forbes but the AMD people giving the interview and trying to rewrite the history to hide the SP4 platform, which appeared in roadmaps one or two years ago but now is silently gone and AMD no longer talks about it.
 


http://wccftech.com/amd-cpu-roadmap-leak-7-nm-starship-14nm-naples-snowy-owl-zen-core/ ?

I can't find any slides from before that.

Cheers!
 


Just because it was leaked in 2017 doesn't mean anything. Pay attention to the slides.
 


Instead of being a jerk, you could provide a link at least? And I did look at the slides. They mention the sockets and that is a product paired with Naples at the top. Are you absolutely sure you're not just confusing the codenames with the actual product names that they were released under?

Cheers!
 


A link to what?

What confusion? It is all well-established in AMD internal docs

d0a61489904667.png



 


And how do you correlate the silicon piece to the socket name in that table you gave?

I can see two whole families there going to different sockets, but I have no idea what they are. For all I can see there, Ryzen can be put pretty much anywhere; just like TR and just like EPYC.

Your thesis is that TR = Snowy Owl, but I just don't see how you can come to that conclusion from any slide nor document I've seen so far.

Cheers!
 


I can correlate because I know the information. I have known it for while. Even before leaks, as the one that you linked to.

No, my thesis is not "TR = Snowy Owl". The TR4 socket used in ThreadRipper is a fusion of the old SP4 socket used in Snowy Owl and the SP3 socket used in EPYC.
 


I don't think anyone understands you Juan, but it's fine.

And doesn't that statement validate AMD's story?

Cheers!
 


My friends and colleagues say otherwise



So, I devote some time to explain that the Forbes article is a piece of marketing that tries to rewrite the history and you interpret my words as I am validating AMD history. ROFL!
 
Status
Not open for further replies.