Monday, June 11th 2018
Intel's 28-core HEDT Processor a Panic Reaction to 32-core Threadripper
At Computex 2018, we witnessed two major HEDT (high-end desktop) processor announcements. Intel unveiled a client-segment implementation of its "Skylake XCC" (extreme core count) silicon, which requires a new motherboard, while AMD announced a doubling in core-counts of its Ryzen Threadripper family, with the introduction of new 24-core and 32-core models, which are multi-chip modules of its new 12 nm "Zen+" die, and compatible with existing X399 chipset motherboards. With frantic increases in core counts, the practicality of these chips to even the most hardcore enthusiast or productivity professional diminishes. The Computex 2018 demos reek of a pissing-contest between the x86 processor giants, with AMD having an upper hand.
The HEDT segment is intended to occupy the space between client desktops and serious scalar workstations. Intel is frantically putting together a new HEDT platform positioned above its current LGA2066 (X299) platform, built around its Purley enterprise platform, and a variant of the LGA3647 socket (this chip + your X299 motherboard is no bueno). This socket is needed to wire out the 28-core Skylake XCC (extreme core count) silicon, which has a six-channel DDR4 memory interface. The company put up a live demo at the teaser of this unnamed processor, where it was running at 5.00 GHz, which led many to believe that the processor runs at that speed out of the box, at least at its maximum Turbo Boost state, if not nominal clock. Intel admitted to "Tom's Hardware," that it "forgot" to mention to the crowds that the chip was overclocked.
Overclocking the 28-core chip was no small effort. It took an extreme cooling method, specifically a refrigerated heat-exchanger, coupled with a custom motherboard (we suspect GIGABYTE-sourced), to keep the processor bench-stable at 5.00 GHz. Intel's defense to Tom's Hardware was that "in the excitement of the moment," its on-stage presenter "forgot" to use the word "overclocked." Gregory Bryant, SVP client-computing at Intel not only omitted "overclocked" from his presentation, but made sure to stress on "5 GHz," as if it were part of the chip's specifications.
"What's amazing is that trade-off, this actually being a 5 GHz in single-threaded performance frequency and not...having to sacrifice that for this kind of multi-threaded performance, so you've got kind of the best of both worlds. So, you guys want to see us productize that thing? Tell you what, we'll bring that product to market in Q4 this year, and you'll be able to get it," he said.
Rival AMD, meanwhile, showed off its 24-core and 32-core Ryzen Threadripper II processors, with its 24-core part beating Intel's i9-7980XE 18-core chip under ordinary air cooling.
Intel used a multiplier-unlocked derivative of the Xeon Platinum 8180 "Skylake-SP" processor in this demo. The Xeon Platinum 8180 "Skylake-SP" is a $10,000 processor with a 205W rated TDP at its nominal clock speed of 2.50 GHz, with a Turbo Boost frequency of 3.80 GHz. The company achieved a 100% overclock to 5.00 GHz, using extreme cooling, and considering that TDP is calculated taking into account a processor's nominal clock (a clock speed that all cores are guaranteed to run at simultaneously), the company could have easily crossed 350W to 400W TDP stabilizing the 5.00 GHz overclock. If a 205W TDP figures in the same sentence as 2.50 GHz nominal clocks, it doesn't bode well for the final product. It will either have a very high TDP (higher still taking into account its unlocked multiplier), or clock speeds that aren't much higher than the Xeon Platinum 8180.
Consider the AMD EPYC 7601 for a moment, which is the fastest 32-core 1P EPYC SKU. It ticks at 2.20 GHz, with a boost frequency of 3.20 GHz, but has its TDP rated lower, at 180W. Now consider the fact that AMD is building the 32-core Threadripper II with more advanced 12 nm "Zen+" dies, and it becomes clear that the 24-core and 32-core Threadrippers are the stuff of nightmares for Gregory Bryant, not because AMD will make more money out of them than Intel makes out of its 28-core G-man in a football jersey, but because AMD's offering could be cheaper and more efficient, besides being fast. An overall superior halo product almost always has a spillover PR to cheaper client-segment products across platforms; and the client GPU industry has demonstrated that for the past two decades.
AMD is already selling 16 cores at $999, and beating Intel's $999 10-core i9-7900X in a variety of HEDT-relevant tasks. The company has already demonstrated that its 24-core Threadripper II is faster than Intel's $1,999 18-core i9-7980XE. It would surprise us if AMD prices this 24-core part double that of its 16-core part, and so it's more likely to end up cheaper than the i9-7980XE.
Intel cannot beat the 32-core Threadripper II on the X299/LGA2066 platform, because it has maxed out the number of cores the platform can pull. The Skylake HCC (high core count) silicon, deployed on 12-core, 14-core, 16-core, and 18-core LGA2066 processors, is already motherboard designers' nightmare, many of whom have launched special "XE" variants of their top motherboard models that offer acceptable overclocking headroom on these chips, thanks to beefed up VRM.
Coming up with a newer platform, namely revising the Purley 1P enterprise platform for the client-segment, with its large LGA3647 socket and 6-channel memory interface, is the only direction in which Intel could have gone to take on the new wave of Threadrippers. AMD, on the other hand, has confirmed that its 24-core and 32-core Threadripper II chips are compatible with current socket TR4 motherboards based on the AMD X399 chipset. It's possible that the next wave of TR4 motherboards could have 8-channel memory interface, wider than that of Intel's Skylake XCC silicon, and both forwards and backwards compatibility with current-generation Threadripper SKUs (at half the memory bus width) and future Threadripper chips.
PC enthusiasts nurse an expensive hobby, but the commercial success of NVIDIA TITAN V graphics card (or lack thereof) shows that there are limits to how many enthusiasts have $3,000 to spend on a single component.
The HEDT segment is intended to occupy the space between client desktops and serious scalar workstations. Intel is frantically putting together a new HEDT platform positioned above its current LGA2066 (X299) platform, built around its Purley enterprise platform, and a variant of the LGA3647 socket (this chip + your X299 motherboard is no bueno). This socket is needed to wire out the 28-core Skylake XCC (extreme core count) silicon, which has a six-channel DDR4 memory interface. The company put up a live demo at the teaser of this unnamed processor, where it was running at 5.00 GHz, which led many to believe that the processor runs at that speed out of the box, at least at its maximum Turbo Boost state, if not nominal clock. Intel admitted to "Tom's Hardware," that it "forgot" to mention to the crowds that the chip was overclocked.
"What's amazing is that trade-off, this actually being a 5 GHz in single-threaded performance frequency and not...having to sacrifice that for this kind of multi-threaded performance, so you've got kind of the best of both worlds. So, you guys want to see us productize that thing? Tell you what, we'll bring that product to market in Q4 this year, and you'll be able to get it," he said.
Rival AMD, meanwhile, showed off its 24-core and 32-core Ryzen Threadripper II processors, with its 24-core part beating Intel's i9-7980XE 18-core chip under ordinary air cooling.
Intel used a multiplier-unlocked derivative of the Xeon Platinum 8180 "Skylake-SP" processor in this demo. The Xeon Platinum 8180 "Skylake-SP" is a $10,000 processor with a 205W rated TDP at its nominal clock speed of 2.50 GHz, with a Turbo Boost frequency of 3.80 GHz. The company achieved a 100% overclock to 5.00 GHz, using extreme cooling, and considering that TDP is calculated taking into account a processor's nominal clock (a clock speed that all cores are guaranteed to run at simultaneously), the company could have easily crossed 350W to 400W TDP stabilizing the 5.00 GHz overclock. If a 205W TDP figures in the same sentence as 2.50 GHz nominal clocks, it doesn't bode well for the final product. It will either have a very high TDP (higher still taking into account its unlocked multiplier), or clock speeds that aren't much higher than the Xeon Platinum 8180.
Consider the AMD EPYC 7601 for a moment, which is the fastest 32-core 1P EPYC SKU. It ticks at 2.20 GHz, with a boost frequency of 3.20 GHz, but has its TDP rated lower, at 180W. Now consider the fact that AMD is building the 32-core Threadripper II with more advanced 12 nm "Zen+" dies, and it becomes clear that the 24-core and 32-core Threadrippers are the stuff of nightmares for Gregory Bryant, not because AMD will make more money out of them than Intel makes out of its 28-core G-man in a football jersey, but because AMD's offering could be cheaper and more efficient, besides being fast. An overall superior halo product almost always has a spillover PR to cheaper client-segment products across platforms; and the client GPU industry has demonstrated that for the past two decades.
AMD is already selling 16 cores at $999, and beating Intel's $999 10-core i9-7900X in a variety of HEDT-relevant tasks. The company has already demonstrated that its 24-core Threadripper II is faster than Intel's $1,999 18-core i9-7980XE. It would surprise us if AMD prices this 24-core part double that of its 16-core part, and so it's more likely to end up cheaper than the i9-7980XE.
Intel cannot beat the 32-core Threadripper II on the X299/LGA2066 platform, because it has maxed out the number of cores the platform can pull. The Skylake HCC (high core count) silicon, deployed on 12-core, 14-core, 16-core, and 18-core LGA2066 processors, is already motherboard designers' nightmare, many of whom have launched special "XE" variants of their top motherboard models that offer acceptable overclocking headroom on these chips, thanks to beefed up VRM.
Coming up with a newer platform, namely revising the Purley 1P enterprise platform for the client-segment, with its large LGA3647 socket and 6-channel memory interface, is the only direction in which Intel could have gone to take on the new wave of Threadrippers. AMD, on the other hand, has confirmed that its 24-core and 32-core Threadripper II chips are compatible with current socket TR4 motherboards based on the AMD X399 chipset. It's possible that the next wave of TR4 motherboards could have 8-channel memory interface, wider than that of Intel's Skylake XCC silicon, and both forwards and backwards compatibility with current-generation Threadripper SKUs (at half the memory bus width) and future Threadripper chips.
PC enthusiasts nurse an expensive hobby, but the commercial success of NVIDIA TITAN V graphics card (or lack thereof) shows that there are limits to how many enthusiasts have $3,000 to spend on a single component.
160 Comments on Intel's 28-core HEDT Processor a Panic Reaction to 32-core Threadripper
I'll be disappointed if the chiller doesn't come in the retail box cause i'm set
They presented a sample running at 5GHz without explicitly saying it was overclocked. Anyone with half a brain should have figured it was overclocked, but this wasn't an actual reveal of a product. So why fret if the end product won't be exactly like what was shown there?
I mean, it could be the end of the world for you, but for me, it's like Bugatti showcasing a 2000BHP prototype only to find out later it can't actually output that much power unless special conditions are met. Guess what, it makes zero difference to me: I don't have the money to buy a Bugatti any more than I have the cash to blow on this upcoming* Intel platform.
*If it ever sees the light of day, in the light of Threadripper2 reveal.
Why choose such a PSU... maybe Intel was improvising too much or maybe the system really needs a huge amount of power... :ohwell:
No i dont feel sorry for intel. Dont get me wrong i am still an intel believer, but the reason i believe for intels missing platform to compete againts amd threadripper 2 chips.
Is intels own dam fault for being lazy the later years, because they had no competition for years and now that they do, they dont have platforms ready to take on amd latest platform. With first gen threadripper intel panic with the 14, 16 and 16 core i9 chips and now they do it again with a server based platform because they dont have any desktop platform ready. Thats what intel gets for being lazy and not considered that amd some day might have something that cut give intel a scare.
I've looked at a couple of retailers here and that brand isn't being sold. I suspect they'd have to release a 3KW version for the UK if they were (or 2.5KW or so to allow for inefficiencies in the PSU).
Finally, did you notice how skinny that mains cable was? I'd bet money it was running hot with all that current flowing through it.
Again, this is the transcript of what they said at the event -That second portion DIRECTLY addresses the potential consumer - they say "You", to the audience. YOU can get the best of both worlds. YOU don't have to sacrifice single thread in order to have 28 cores.
Except, that's clear bullshit, isn't it? The customer can't get the best of both worlds. The customer *does* have to sacrifice single-thread to get 28 cores.
Without even addressing the point that they didn't mention overclocking here, it is ABUNDANTLY CLEAR that Intel in the space of those two sentences, have promised a level of performance in both single and multi-thread, that this product cannot possibly deliver to consumers. And they *know* they're doing it - Unless you want to make the argument that this entire exchange (Which comprises really the entire worthwhile content of their demo), was also a mistake, and they "forgot" that the consumer has to make a tradeoff?
There was no damage done here. Move along.
2 - No, I don't take things companies say at face value, however there is a matter of degrees. When a bicycle company says their new frame is 17% stiffer than last years, I know that's a cherrypicked measurement of a certain stressor, but that the measurement itself is true in some context and a consumer can expect a product that reflects the claim - even if the claim is hyperspecific and not significant in the grand scheme.
When Intel say "You can get single-threaded performance frequency <...> not having to sacrifice that for this kind of multithreaded performance" and it turns out that the consumer cannot actually get that? That's very different.
3 - There absolutely was damage done here. The tech press reporting this story have done the work of Intel's marketing department. They have successfully allowed Intel's announcement of a chip that doesn't exist and will never exist (A 5GHz, 28 Core part), to overshadow AMD's announcement of a product that does exist and is coming to market in a matter of a few months in a form that the consumer can expect will match the demonstration AMD provided.
That's a HUGE corporate win for Intel, achieved only via deception and misleading the press. It is not "Nothing" - It has real, and significant, financial consequences for the companies involved.
btw, I've got and Intel CPU, so you can't just brush me off as an AMD apologist.
I really don't care. I'm unsubscribing from these threads as I really don't have anything else to add.
No need to rage quit the thread! :)
It's not like Intel will send me a cookie if I point out enough times that no puppies have been harmed during the process ;)
You are completely ignoring this point, because, as qubit said, you are insistent on being an Intel apologist.
It would be a stronger point in favour of your position if you were able to articulate how this confusion has not been of benefit to Intel in the tech press. The fact you refuse to do so is because you are unable to do so - Intel have clearly gained from the confusion.
The argument is whether Intel *intended* to benefit from the confusion. Repeating that "nobody was hurt" is facile - AMD were hurt by it. This is self-evident. If Intel had mentioned the word "overclocked" in their presentation, AMD would not have been hurt by the announcement. That is also self-evident.
@GlacierNine Not sure what marketing effect are you talking about. Did you do market research and found people who still believe Intel will launch a 28 core, 5GHz product?
Over and out.
Now at what point are you going to address the substance of this instead of cherry picking individual sentences to respond to?
I think it's quite clear to anyone reading this debate you're not having with us, that you've been deliberately straying from the topics anyone is trying to get you to address, because they're inconveniently well founded and you can't make a reasonable point against them.
*cough*Polaris will offer 2.5x the perf/W*cough*
Once again, you attack only the part of anyone's argument that you think you can manage to make a point about, while completely ignoring the main substance of any other post in this thread that makes a better point. What AMD has done in the past is not at all relevant here - it is simple whataboutery and does not address the reasons anyone would have an issue with what intel have done here, or what AMD have done in the past.
Let's just ask you a few very pointed questions here:
1 - Intel lied. This is beyond question at this point. The only reason they would lie is to gain a competitive advantage in the marketplace. This is also beyond question. Bearing in mind the two previous points, why would anyone not have an issue with the fact they did so?
2 - Bearing in mind the same two points as before, why would Intel, rather than coming clean and making a proper statement about it, choose to only contact those few media outlets that ran followup stories mentioning the chiller, in order to provide their excuse of having forgotten to mention the overclock, and why would their statement be so short regarding what is a fairly major gaffe, potentially affecting their credibility within the tech community?
3 - Why would intel make the specific claim "You can get single-threaded performance frequency <...> not having to sacrifice that for this kind of multithreaded performance" in their demonstration, while demonstrating a part running at 5GHz, massively overclocked, if their intention was not to mislead the audience to believe that this part was capable of running at this speed in consumer's hands? Bearing in mind that in advance of the demonstration, Intel had already announced and *released* a part that ran at 5GHz stock speeds.