Raevenlord
News Editor
- Joined
- Aug 12, 2016
- Messages
- 3,755 (1.34/day)
- Location
- Portugal
| System Name | The Ryzening |
|---|---|
| Processor | AMD Ryzen 9 5900X |
| Motherboard | MSI X570 MAG TOMAHAWK |
| Cooling | Lian Li Galahad 360mm AIO |
| Memory | 32 GB G.Skill Trident Z F4-3733 (4x 8 GB) |
| Video Card(s) | Gigabyte RTX 3070 Ti |
| Storage | Boot: Transcend MTE220S 2TB, Kintson A2000 1TB, Seagate Firewolf Pro 14 TB |
| Display(s) | Acer Nitro VG270UP (1440p 144 Hz IPS) |
| Case | Lian Li O11DX Dynamic White |
| Audio Device(s) | iFi Audio Zen DAC |
| Power Supply | Seasonic Focus+ 750 W |
| Mouse | Cooler Master Masterkeys Lite L |
| Keyboard | Cooler Master Masterkeys Lite L |
| Software | Windows 10 x64 |
In a Forbes interview conducted by Anthony Leather, AMD officials Senior Vice President and General Manager Jim Anderson, Corporate Vice President of Worldwide Marketing John Taylor, Sarah Youngbauer of AMD's communications team, and James Prior, AMD's Senior Product Manager, have shed some light on the development process for AMD's equivalent of a flash hit - their HEDT, HCC Threadripper chips. Threadripper, which leverages AMD's Zen architecture used in their Ryzen and EPYC processors, makes use of an MCM design with up to 16 cores and 32 threads, with AMD's svelte Zen, 8-core base units linked through the company's high speed interconnect Infinity Fabric.
This has allowed the company to scale designs from four core processors with Ryzen 3, all the way towards the current cream of the crop Threadripper 1950X. It's an extremely scalable design, which brings with it improved yields and some pretty significant cost savings for AMD due to smaller dies. This, in turn, means the company is able to more agressively price their Ryzen and Threadripper processors compared to the competition, at least when it comes to high core and thread counts for the same price bracket - and the success of that business decision is showing.
For our forum lurkers, this article is marked as an editorial.
This modular approach to the base Zen core design, which was meant to be scaled up in dies of 8 cores, has been at the heart of AMD's ability to bring to market a HEDT, HCC platform in what can be only be considered record time. According to Sarah Youngbauer, AMD's Threadripper chips weren't part of the company's original Zen roadmap (which included only the designs that came to be Ryzen and EPYC). Youngbauer said that Threadripper was a labor of love, originating 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." Apparently, it would be a full year of spare-time work being put on Threadripper before the human power behind it ever sought the hierarchy's green light for officially including it in their Zen-based lineup.
James Prior, who was involved with the original team that spearheaded Threadripper's development as a passion - and then a product - says that work began on what they saw as a product gap; something that "(...) stood above Ryzen with more memory bandwidth, cores, PCI-E lanes." A team that numbered between 20 and 30 people that included platform architects, core design engineers, a business unit and marketing team gathered looking for what that product might be. This was almost as if an entire new department had risen amongst AMD, a think tank of sorts that leveraged the company's portfolio, IP, and Zen core design. All of this beginning in 2015 - three years after AMD started working on the Zen design with Jim Keller, and around the same time Zen would be formerly revealed by the company.
But Threadripper was much more than just a pet project, and has turned out to be just that: feasibility studies for Threadripper's development and manufacturing cost against expected demand and prospective price points showed that "it was one of the best-planned products (...) in a long time." This shows in how fast the company delivered on the Threadripper concept, since the actual green light for product development was obtained from Jim Anderson on June 2016. This is little more than a year from green light to product launch, which should have - and likely has - our collective heads shaking in awed disbelief. Jim Anderson himself says that the unusually fast launch was achieved even though initial estimates placed Threadripper's launch sometime in 2018. He says Threadripper "(...) never had a business plan (...) 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."
While Threadripper is a high-end desktop platform, Ryzen 7 was also designed to disrupt the HEDT market - current market ocnjecture and positioning had left rival Intel with a free pass on the high performance CPU market, which led to an (expected, if disappointing) stagnation in both core count and core design. Intel really didn't have anyone urging them to offer better value, to push the design envelope with higher R&D costs and developmental effort, which led to multiple generations of a slightly refined, slightly re-spinned architecture which led to user boredom and stagnation of not only the x86 market, but also software development. AMD's John Taylor says that Ryzen was developed as a somewhat of a HEDT chip in and of itself, since Intel's desktop platforms topped out at a (now measly) 4-core, 8-thread design. high core counts in Intel products were left to the devices of their uber-expensive HEDT lineups, which cost consumers an exponential arm and a leg more than the core count and performance increases justified. So while mainstream desktops would be led by AMD's 8-core, 16-thread Ryzen offerings for almost all workloads, the leadership crown for non-server environments was still left with Intel and its 10-core i7-6950X. John Taylor, James Prior and Jim Anderson wouldn't just let that settle like that; they 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."
Threadripper as a design comes as a scale-down from EPYC, but standing above their Ryzen designs. Thoughts of larger dies than the original MCM design of EPYC and Ryzen for achieving higher core counts were quickly discarded - that would be throwing away what is arguably Zen's greatest asset. AMD's James Prior touches upon and underlies that: "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."
Naturally, when you basically re-use a socket and have only slight discrepancies between two distinct product lines, it becomes easier for platform-enablers to redesign and re-purpose their products for the new market segment. Launching a CPU isn't a solo play, as we've seen with Ryzen's somewhat rushed launch, which included (now increasingly rare) platform issues. At best, it's a perfect marriage of compromises, impossible deadlines and headbutts between CPU makers, platform-enabling companies and motherboard manufacturers. James Prior handily explains the advantages of having two similar designs coming to market, and Zen's scalability: "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."
I'll quote AMD's Jim Anderson on Threadripper's bold choice of name: "Threadripper was just the projects code name, which we were using before we decided what the real name and branding would be. However, we started to like the name and eventually we all started to like it. In the end, we got the necessary team to trademark it and that was that, but it's unusual for a code name to be used as a product name like this."
I think "Threadripper" is an apt name for a product that delivered almost double the amount of cores as the competition's highest core-count CPUs for the HEDT market. Don't expect AMD's Threadripper to sit and wait for Intel's answers, however. While AMD is enjoying the performance crown with its 16-core, 32-thread behemoths as of now, Intel's 18-core response will certainly snatch the HEDT performance crown back from AMD. At least, it will, if AMD doesn't expand on Threadripper's core-counts. AMD's John Taylor himself says he doesn't expect Threadripper's lineup to be finished with the current 8, 12 and 16-core CPUs it currently features; and looking at Threadripper's development history and EPYC roots, it's easy to see it wouldn't take AMD that long - or that much - to increase core counts above what they currently stand at. MCM designs may have their shortcomings, but AMD has shown - and will continue to show - how these are, more likely than not, the industry's future.
View at TechPowerUp Main Site
This has allowed the company to scale designs from four core processors with Ryzen 3, all the way towards the current cream of the crop Threadripper 1950X. It's an extremely scalable design, which brings with it improved yields and some pretty significant cost savings for AMD due to smaller dies. This, in turn, means the company is able to more agressively price their Ryzen and Threadripper processors compared to the competition, at least when it comes to high core and thread counts for the same price bracket - and the success of that business decision is showing.
For our forum lurkers, this article is marked as an editorial.
This modular approach to the base Zen core design, which was meant to be scaled up in dies of 8 cores, has been at the heart of AMD's ability to bring to market a HEDT, HCC platform in what can be only be considered record time. According to Sarah Youngbauer, AMD's Threadripper chips weren't part of the company's original Zen roadmap (which included only the designs that came to be Ryzen and EPYC). Youngbauer said that Threadripper was a labor of love, originating 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." Apparently, it would be a full year of spare-time work being put on Threadripper before the human power behind it ever sought the hierarchy's green light for officially including it in their Zen-based lineup.
James Prior, who was involved with the original team that spearheaded Threadripper's development as a passion - and then a product - says that work began on what they saw as a product gap; something that "(...) stood above Ryzen with more memory bandwidth, cores, PCI-E lanes." A team that numbered between 20 and 30 people that included platform architects, core design engineers, a business unit and marketing team gathered looking for what that product might be. This was almost as if an entire new department had risen amongst AMD, a think tank of sorts that leveraged the company's portfolio, IP, and Zen core design. All of this beginning in 2015 - three years after AMD started working on the Zen design with Jim Keller, and around the same time Zen would be formerly revealed by the company.
But Threadripper was much more than just a pet project, and has turned out to be just that: feasibility studies for Threadripper's development and manufacturing cost against expected demand and prospective price points showed that "it was one of the best-planned products (...) in a long time." This shows in how fast the company delivered on the Threadripper concept, since the actual green light for product development was obtained from Jim Anderson on June 2016. This is little more than a year from green light to product launch, which should have - and likely has - our collective heads shaking in awed disbelief. Jim Anderson himself says that the unusually fast launch was achieved even though initial estimates placed Threadripper's launch sometime in 2018. He says Threadripper "(...) never had a business plan (...) 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."
While Threadripper is a high-end desktop platform, Ryzen 7 was also designed to disrupt the HEDT market - current market ocnjecture and positioning had left rival Intel with a free pass on the high performance CPU market, which led to an (expected, if disappointing) stagnation in both core count and core design. Intel really didn't have anyone urging them to offer better value, to push the design envelope with higher R&D costs and developmental effort, which led to multiple generations of a slightly refined, slightly re-spinned architecture which led to user boredom and stagnation of not only the x86 market, but also software development. AMD's John Taylor says that Ryzen was developed as a somewhat of a HEDT chip in and of itself, since Intel's desktop platforms topped out at a (now measly) 4-core, 8-thread design. high core counts in Intel products were left to the devices of their uber-expensive HEDT lineups, which cost consumers an exponential arm and a leg more than the core count and performance increases justified. So while mainstream desktops would be led by AMD's 8-core, 16-thread Ryzen offerings for almost all workloads, the leadership crown for non-server environments was still left with Intel and its 10-core i7-6950X. John Taylor, James Prior and Jim Anderson wouldn't just let that settle like that; they 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."
Threadripper as a design comes as a scale-down from EPYC, but standing above their Ryzen designs. Thoughts of larger dies than the original MCM design of EPYC and Ryzen for achieving higher core counts were quickly discarded - that would be throwing away what is arguably Zen's greatest asset. AMD's James Prior touches upon and underlies that: "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."
Naturally, when you basically re-use a socket and have only slight discrepancies between two distinct product lines, it becomes easier for platform-enablers to redesign and re-purpose their products for the new market segment. Launching a CPU isn't a solo play, as we've seen with Ryzen's somewhat rushed launch, which included (now increasingly rare) platform issues. At best, it's a perfect marriage of compromises, impossible deadlines and headbutts between CPU makers, platform-enabling companies and motherboard manufacturers. James Prior handily explains the advantages of having two similar designs coming to market, and Zen's scalability: "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."
I'll quote AMD's Jim Anderson on Threadripper's bold choice of name: "Threadripper was just the projects code name, which we were using before we decided what the real name and branding would be. However, we started to like the name and eventually we all started to like it. In the end, we got the necessary team to trademark it and that was that, but it's unusual for a code name to be used as a product name like this."
I think "Threadripper" is an apt name for a product that delivered almost double the amount of cores as the competition's highest core-count CPUs for the HEDT market. Don't expect AMD's Threadripper to sit and wait for Intel's answers, however. While AMD is enjoying the performance crown with its 16-core, 32-thread behemoths as of now, Intel's 18-core response will certainly snatch the HEDT performance crown back from AMD. At least, it will, if AMD doesn't expand on Threadripper's core-counts. AMD's John Taylor himself says he doesn't expect Threadripper's lineup to be finished with the current 8, 12 and 16-core CPUs it currently features; and looking at Threadripper's development history and EPYC roots, it's easy to see it wouldn't take AMD that long - or that much - to increase core counts above what they currently stand at. MCM designs may have their shortcomings, but AMD has shown - and will continue to show - how these are, more likely than not, the industry's future.
View at TechPowerUp Main Site
Last edited by a moderator:
I'll give this a proper read a bit later, when I can do it justice.
"However, we all started to like the name and eventually we all started to like it."