I guess AMD will have a 32C Rome SKU that destroys this at a significantly lower price and TDP in the near future. It's only hope is clock speed, but TSMC's 7nm seems optimal to around 3.3GHz before power goes up more than linearly with speed, so if AMD were to make a clock-speed targetting 32C SKU they should be able to.
I think they could go past 4.0ghz at 32C and still match Intel’s power budget. The 3950x is only 105W, so if AMD wants to play with 215W, they could aim pretty high. I bet most of the best chiplets are already heading to this market to begin with due to better margins.
I wouldn't bet on it. While there are power saving son the 7nm 'core' dies, there is also power loss on the IF interconnects (all 8 of them) plus the large not-a-Northbridge die that remains on an older process. We saw with Zen that for high core count Epyc more power was consumed by the uncore rather than the cores themselves (https://www.anandtech.com/show/13124/the-amd-threa...
EPYC Rome is not slated to reach 4.0 GHz if leaks are to be believed. However it's important to note that Intel's boost clocks are likely single core boost, with all core boost being a bit lower. Zen 2 has already been shown to be quite the formidable CPU even at lower clocks. Even with a rumored boost of 3.4 GHz, chances are AMD will have a clear win in the data center. After all, when Intel charges 3X more for 10% more performance...well, as the joke goes: "Nobody ever got fired for buying Xeon...until NOW!"
Intel's TDP listings are not the actual TDP, nor are they power consumed. All IO is handled by the IO Die itself, so AMD can have as few or as many chiplets as it wants. Thus, an 8 core, well binned, EPYC ROME CPU could indeed hit a 4.7 GHz Boost if it only has a single chiplet.
Still they can not, the low turbo of Epyc 2 line say something about TSMC low power process huge issues. On Ryzen 3000 line (performance process) the turbo frequence is obtained with a too high voltage applied to the core, moreover on all cores up the cpu overclock is pretty bad and spread an insane power. It is a fact that at least for now the single thread performance of Epyc 2 line is suboptimal, and Intel beat badly the brand new AMD server line in responsiveness. A good reason to say that it is not all rosy at Sunnyvale and many critical workloads are right now forbidden to Epyc.
You seem to have a lot of great information that is nowhere to be found on the sites I frequent. Where can I see the detailed information on Epyc 2 single thread performance? I have not seen anything about clockspeeds so far. What I have seen is a lot of speculation that AMD is saving the best chiplets for TR and Epyc. Considering that TR2 hit the same peak clocks as Ryzen 2000, the conclusion might be to say that Epyc will hit the same peak clocks as Ryzen 3000. But I can only speculate, you seem to have all the answers.
Epyc "ROME" or EPYC 2 as you are calling it is based on the Zen 2 core. Quite frankly, based on what we have seen thus far, it doesn't matter where clock speeds fall (3.4 GHz is the rumored boost clock, and a bit under 10 grand is the rumored top of the line price point for a 64 core, 128 thread "ROME" CPU). Intel is going to have a hard time selling these things when "ROME" launches regardless. After all, would you spend $18,000 on a 28 core CPU just because single core workloads can boost to 4.0 GHz? Or would you instead spend $6000 on a 64 core/128 thread CPU that only boosts to 3.35 GHz? That's 1/3rd the price for more than 2X more cores and a slight clock speed reduction that matters very little for most workloads that these CPUs are going to be handling. Also, as others have mentioned, AMD could easily hit higher clocks than rumored.
> would you spend $18,000 on a 28 core CPU just because single core workloads can boost to 4.0 GHz? Or would you instead spend $6000 on a 64 core/128 thread CPU that only boosts to 3.35 GHz?
For our database server, I'd pick neither. I'd get dual 8-12 core highest boost cpus I can get. They are usually (last few gens) in the 3000-4000$ range.
"For our database server, I'd pick neither. I'd get dual 8-12 core highest boost cpus I can get. They are usually (last few gens) in the 3000-4000$ range."
Wouldn't you do better spending the money on maxing out the RAM and getting as much on fast SSD as you could. Obviously it depends on workload but wouldn't the new 3900X, 64GB RAM and the fastest PCIe4 M.2 SSD's offer better performance for less? I guess it depends where the bottleneck is in modern database servers but my "guess" would be on accessing the data rather than processing it. This isn't advice as I've never spec'ed database servers but I would genuinely be interested in any real world tests out there.
AMD will really have a hard time in the server market where single thread performance matters a lot more than density, power consumption, IO, total cost (including cooling)..... [/s]
I wish for once some one here, would keep there comments to topic, I personally refused to comment on AMD unless someone brings up something against Intel or incase of XPS 15 2in1, my gpu like drivers from AMD and such.
"...I personally refused to comment on AMD unless someone brings up something against Intel..."
That's because you have an irrational personal bias and are incapable of widening your thought horizon to the point of being able to overcome that limit.
"That's because you have an irrational personal bias and are incapable of widening your thought horizon to the point of being able to overcome that limit."
There is no irrational bias - in fact it very rational - I prefer Intel because they are the one that actually create the x86 based CPU and was not cloned from it.
But you notice I mention "Personally", that was one purpose for two reasons 1. I knew it was going to cause a response from AMD bias fan person and it did 2. It is true that others have there personal opinions about products and I just wish they keep it to themselves. Just better to agree to disagree
As for my thought process, you have no idea what experience I have, just to let to know I once work for an operating system and found a bug in IBM 486SLC cpu where chip add it address lines inverted when switching between 286 and 386 protected modes. This was at my first job for almost 7 years and there are few developers on this planet that can even comprehend what I am discussing.
Let's be honest, do the average customer of computer actually read these comments. In fact that average customer care about it and especially desktop chips - server chips have there own specially. But most customers use it for word processing, spreadsheets, internet and emails - but I have 30 years of technical knowledge in these area include low level CPU design and even though I do application program - the low level parts still interest me
Your reply almost sounds like a copypasta where you're just bragging about past experience. Look, it's fine to have preferences and even biases if you can acknowledge them. And it seems like you do acknowledge your own, which is great! But people on the internet also have their personal preferences and it shouldn't blow your mind that people are bringing up AMD on an article like this.
I think regardless of any bias HStewart is right in his criticism that every almost every story is inundated with comments praising AMD and bashing Intel. Regrless of if there comments are right most of the time the form it's done in is not a place for it. And is just off topic, even by internet standards.
Also while I can certainty understand see what HStewart as "bragging" how else would you suggest he states his justification. For example if he said x is try but you disagree and he replies no I'm a doctor x is true. You can't say no your just bragging. Even if that is the case it is not what matters and is just deflecting from the point. Also I don't like to make thing personal but PeachNCream while I empathize with your passion from your past comments I can't take what you say to seriously.
AMD articles were historically "inundated" with comments praising Intel and bashing AMD before Zen came out. This is what tech discussion pages look like - especially when the balance between the two remaining x86 CPU competitors begins to shift. Given that the article is about how much Intel charge for an extra 300Mhz on your 28 core CPU, the imminent appearance of a competitor who tends to provide better value is most certainly an on-topic discussion.
HStewart is not a logical individual. His claim to authority here makes as much sense as claiming expertise on dental health based on having once been a proctologist. If you find yourself in agreement with him, I'd encourage taking a short break from the internet.
You're mistakenly projecting your Intel bias outward. Like most people, I don't care what branding is on a CPU. It's amusing that your defense is to attempt to lawyer over the word "personally" and then make claims about your personal experience that do not justify your blind loyalty.
I actually don't care either, what I don't like it people bashing Intel - with AMD statements - I never bash AMD products unless they bash Intel first. Just sounds like a bunch of uneducated discussing bashing another product because one thinks one product.
I not saying Intel needs does not need to make improvements, In fact AMD helps Intel moving toward the future - I just think most people really don't care - about high core counts and high frequencies.
It not blind loyalty, I mention before it because Intel created the original product - x86 processors and because of stupid IBM back in the 1980's, AMD came in to picture because of 2nd source. My first computer that I purchase had an AMD processor on it was 386 clone - not sure the place advertise that fact.
Times have change now for most people and organizations, desktop are part of past - work have notebook connected to monitors. so you can bring the computer to board room. Days of desktop computer in work force is going way.. even now we have workstation based notebook with Xeon processors
But if there is a threat to Intel, it not the desktop AMD chips but instead iPad and Android tablet, which at my last job, I would have died if I supported Android on phone or tablet. So for windows based computing I prefer Intel because of long history of personal good results with them.
Probably if you want to say what has hurt me with AMD is there video, back in older once AMD purchase ATI - I did try AMD card that supposedly had good TV recording, I had bad luck with and I never switch from NVidia. But laptops became mainstream over the years and video card became less important. I still have nVidia on my Lenovo Y50 and I like the layout of Dell XPS 15 2in1 and even though I have concerns with it video, I was willing to give it try. It has some incompatibilities with two things 1,. Older games and 2. some professional 3d graphic software like Vue 2016. And also Steam VR.
But my life style is changes and graphics is becoming less important, but I still have older tech interested of younger days.
So main concern here not against AMD but blasting of AMD on topics discussing Intel product. So what does anybody really care. This kind of attitude surely will not make want to purchase an AMD product. I believe this attitude of blasting pro-AMD into Intel product would hurt AMD. I would say the same thing on any product
> Yes but x64 is just an extension of x86 and it would have came in time.
Maybe - but not from Intel, who didn't have a 64-bit x86 roadmap as their Itanium vision was rejected by the marketplace so it came down to AMD to invent x86_64 and give customers what they wanted rather than being force fed something expensive, complicated, incompatible and badly planned. As usual it is AMD that is innovating in the x86 space, while Intel does as little as possible and sweats it's assets/monopoly.
i dont think intel had ANY plans to extend x86 to 64 bit, like CityBlue said, intel tried to transition to 64 bit via itanium, but because it had nothing to do with x86, EVERYTHING software wise would of needed to be recompiled to Itanium 64 bit, and the industry didnt want to do that. to be honest, some the cpu innovation has not come from intel, but from amd.
I dare you, even double dare you, to use X86 without the 64 bit instruction these days. Intel or AMD.
even if X86-64 needs the base X86 instruction set, Intel would be useless today without the 64bit extension. There's one huge factor you're forgetting about X86, it's limited to roughly 4GB of ram. AMD beat Intel to the punch, it's a simple fact of life.
> As for my thought process, you have no idea what experience I have, just to let to know I once work for an operating system and found a bug in IBM 486SLC cpu where chip add it address lines inverted when switching between 286 and 386 protected modes. This was at my first job for almost 7 years and there are few developers on this planet that can even comprehend what I am discussing.
Your 20-30 year old experience is irrelevant to the current predicament that Intel is in. AMD is winning this generation. Accept it. Intel will come back eventually; there's no doubt about that. But they are not in the winning position RIGHT NOW. And nothing you can say will change that.
And you're not the only one on these boards with 20-30 years experience in these issues. Most of us just don't feel the need to brag about it, nor use it as a blunt hammer to try to force people to think that you know what you're talking about.
"There is no irrational bias - in fact it very rational - I prefer Intel because they are the one that actually create the x86 based CPU and was not cloned from it." Bias is bias, whether you consider it rational or irrational. Actually by definition every bias is irrational, so "rational bias" is functionally equivalent to "logical prejudice". They are both contradictions.
The reason you "prefer Intel" is irrational (though most of our choices are irrational). The creator of an ISA, whether it's x86 or ARM or anything, *can* fall behind some other company which licensed that ISA. Intel are much bigger than AMD so they did not fall behind AMD (yes, they did - not in market share, of course, I mean in desktop CPU performance. AMD has now surpassed Intel in IPC and power efficiency, *almost* matched them in single core performance and continues to surpass them in multicore performance) due to lower funding.
In fact their much bigger size and much more money available for R&D are precisely what made them fall behind : it made them overconfident enough to try multiple new technologies at the same time with their 10nm node, making it so immensely complex that they still have not managed acceptable yields with it, not even with its supposedly "fixed" 2nd gen 10nm+ iteration that Ice Lake is to be fabbed with.
Intel fell behind due to node issues, not μarch issues. If they had faced no issues with their 10nm node -for instance if they had made it quite simpler and less dense- they would have already released Ice Lake (Sunny Cove + Gen11 iGPU), Tiger Lake (Willow Cove + Gen12 iGPU and even Alder Lake (Golden Cove + Gen13? iGPU).
No comment about your computing experience, which I fail to see its relevance here. In this context it smells a bit like an argument from authority fallacy to me.
What is truly remarkable about this guy is that every single desktop chip we run nowadays is based on an AMD design. If Intel had its druthers, we'd all be running IA-64 Itaniums right now.
So, I guess in this case, he should be shunning EMT64 as a 'clone', right?
HStewart " I prefer Intel because they are the one that actually create the x86 based CPU and was not cloned from it. " so does that mean if AMD released a cpu that was in EVERY way better then what intel has, in performance, power usage, cores, etc, while costing much less, you would still buy the intel chip, just be cause of the above quote ?? if your answer is yes, then im sorry, but that is the dumbest thing i have ever seen, and shows you either really are the biggest intel fan on here, or, you have more money then you know what to do with it.
Skeptical123. " in his criticism that every almost every story is inundated with comments praising AMD and bashing Intel. " have you NOT seen his previous posts ?? in practically EVERY article about AMD, that is good news for the company, HStewart will post some crap about how its not a big deal, and then go on to try to turn the good amd news, into bad, then go and praise intel in some way, what you are saying in your 1st paragraph, is the EXACT same thing HE DOES.
"they are the one that actually create the x86 based CPU and was not cloned from it" isn't a rational basis for a bias. The property of 'who did the thing first' has no bearing whatsoever on how good these companies are currently.
Given that AMD created the x64 superset of x86 that Intel subsequently borrowed back, even according to your way of seeing things Intel and AMD are now arguably even.
I'm forced to conclude that you really don't have a rational basis for your preference.
Except AMD is expected to do very well against Intel in the high core count chips. I think it's perfectly reasonable to mention AMD here. Seeing the 3700x/3900x reviews in multithreading I think Intel is about to get stomped in the server space. AMD's SMT looks very nice this generation. With that said I still am reserved and will wait to see reviews.
We'll keep our comments on topic when you do. Guess when that will be; somewhere north of 10 years after hell freezes over.
And you admit that you only respond when someone speaks against Intel (and don't even care whether they deserve it). And saying "I knew it was going to cause a response from AMD bias fan person" proves that you're just trolling, and that's all you ever do. Don't be butthurt that you're being called out on it.
Look, we all get that Intel is the Lord and Saviour to you. But this isn't about Intel. It's about your blind zealotry of a company that doesn't care about you, and you EXTREMELY ANNOYING and TROLLING comments. You are looking to get a rise out of people, and we will keep calling you out for it.
sa666666666 you are also forgetting that he will also speak against AMD if they announce some good news, and say some BS about it to make it look bad, then right after, praise intel in some form, the announcement a bit ago about the Cray supercomputer is and example of that.
only if some one bash intel ?? are you serious ??? you gladly bash amd ANY chance you get, even if no one said anything negative about intel 1st... you will find some way to bash them, then turn around and praise intel.. come on HStewart, enough with the bs. your long winded post above.. is also complete bs. you even go as far as making things up, to make amd look bad, and intel to look good.
HStewart, even if some one didnt bash intel in their comment, but instead suggested AMD as a better alternative, cause of price, performace, power usage, or all 3, you take it as an intel bash, and reply with an amd bash. so you are the one that bashes amd 1st most of the time, not the other way around.
What surprises me is that there's like a dozen people saying, "Hey, you're being cluelessly brand loyal to a fault," and his mind can still find an escape route which doesn't endanger his poorly grounded thinking. A non-obsessed, rational person would have gone through some sort of internal self-assessment a long time ago and concluded that dialing it back a bit at the very least, would be a good idea.
i think he kind of admitted he is bias with this line " I prefer Intel because they are the one that actually create the x86 based CPU and was not cloned from it " and he never answered my direct question to him about if AMD released a cpu that was in EVERY way better then what intel has, in performance, power usage, cores, etc, while costing much less, if he would still buy the intel chip.. my guess, is he would still by the intel chip because is blind loyalty to intel wouldnt allow him to buy a non intel chip, even if the intel chip cost 2 or 3x more. " In fact AMD helps Intel moving toward the future " so intel cant move forward WITH OUT AMD's HELP ?? thats ludicrous. intel is QUITE capable of moving forward on its own, but it CHOOSES not to, instead, intel CHOOSES to stagnate the cpu market like it has over the last few years, and milk its customers for as much as they can.
probalby. in server environments the amd cpu will be much better. but in latency sensitive tasks maybe not. a lot of people running high end autodesk software like maya or 3dsmax would probalby run better on intel cpus. besides the cpu is such a small cost of a server that the price isnt that significant.
and support both hardware and software support is an intel selling feature. if a chip goes down intel will send out a courier and replace it. even for your desktop chip. amd are missing some key software tools that intel have. mostly people just use the intel tools though so thats not a huge difference.
the final most important piece of the puzzle is supply. even if everybody in the world wants to go amd they cant. amd just cant make enough chips to replace even half of intels sales. people have to buy intel. they could buy amd out and then move some stuff over to arm and they would still have to buy intel. at least somebody would
Curious about the 24C range, since 6C=>8C chiplets is like a >50% price jump (b/c yields I guess). Also, of course, about AMD's pricing and Intel's (pricing/product) response.
I always wonder why Intel has so many SKUs. I don't think anyone wants this many choices since they are hardly different. I like to see choices but in such a small increments, nah? Is this marketing? Production forces them to it e.g. silicon variation?
These are Intel's equivalent to AMD's 'semi custom' service, were Intel will produce an SKU to the request of a specific vendor for a specific product. It's why the 'list price' is a bit of a misnomer, as they're not listed and that price doesn't really reflect what the companies buying these variants are paying.
Yes, except that at least two of those are likely to be artificial limitations for product segmentation (read: extracting the most profit from those who will pay for it).
Is this a bad thing? Not necessarily. But it's definitely a thing, and annoying if you want those features.
No segmentation, die harvesting. When you have a very large die, you have lots of individual component parts you can fuse off if they have a defect. Cores, PCIe interfaces, memory interfaces, etc. If you did not harvest these dies and only sold 'perfect' dies as a single SKU line, then you would have a very small volume of parts and high prices for those parts. By finely binning dies into a large number of SKUs based on yield, then you have many more sellable dies and customers can buy ones that lack features they do not use to reduce outlay.
Or some of both. Unlike Skylake, Intel seems to be including 26 cores worth of L3 on some 24-core Cascade Lake chips, which implies that yields are up (and that Intel would rather have a bigger performance gap between their 24- and 28-core parts).
all those skus arent for us. they are for oem's. intel sells so many more chips to people like dell and hp than to us that we are basically insignifcant. thats why all the sales numbers with amd beating intel are not nearly as damning for intel as people think...yet.
but anyway if oems can save 5 dollars per chip that can add up to millions. so more skus allows oems to tune their costs to their liking. you can pretty much overlook them. they arent meant for us
without rome even available the recent intel refresh / intel already dropped it cpu pricing with 30%.(for the 1000s unit price)
but still horrible what you have to pay for those high core count sku. knowing spectre, melt and zombie i wonder if those high end buyers ever think twice and revise there portfolio/buying behavior
I think a big difference with these CPU's is that they are not just Dual CPU setups but 4 or even 8 CPU motherboard options. I believe this is why they are call Xeon Scalable.
Plus Xeons have advance instruction sets like AVX 512 that not found in other systems and also chipsets are have better IO.
Yes it old but my Dual Xeon 5160 was faster than most machines until the i7's came out.
Keep in mind 4 16core cpu';s is same amount of cores as 2 32core cpus. There is also likely IO difference with multiple cpus and likely it because of less core per cpu, can be run at faster speed.
But the speed difference when using AVX512 vs normal AVX2 is significant and more than clock speed could ever make the difference.
And if you don't have AVX512 based application, should be able to disable it and run at full speed.
AVX 512 is not in mainstream computers until Sunny Cove, but it would be bad thing if AVX 512 not active slows things down. This part could be wise it not main stream yet.
" AVX 512 is not in mainstream computers until Sunny Cove " um... i think you might be wrong on that, even just a little, as skylake X and cannon lake both have avx512
The other thing the scalable has going for it I believe is pci-e lanes. I use scalable to build out full lanes of pci-e based flash storage arrays. No AMD options that fit the bill, even the consumer side they still seem very focused on direct cpu performance.
I don’t see AMD losing to the scalable any time soon.
Intel always release these special order tray only processors. In the last series of xeon there was a tray only I used all the time because lower power and full speed and the cost difference was about 300$. I can’t tell you how many times customer cooking equipment failed and I was happy with the lower heat cpu resulting in no hysteria.
This doesn't make much sense. EPYC has a lot more PCIe lanes available than Xeon Scalable - 128 PCIe 3.0 on the first generation. Surely that would be superior for building PCIe-based flash storage arrays?
EPYC motherboards have 128 PCIe lanes direct from the CPU (whether single socket or dual socket).
Xeon SP has 36 PCIe lanes direct from the CPU socket. In order to match an EPYC system, you NEED a 4 socket motherboard, 4 separate CPUs, and 4x the number of RAM sticks to fill it. How is that "better"? You're paying extra for the multi-socket motherboard, extra for the multi-socket CPUs, extra for the DIMMs, extra, extra, extra.
If you want PCIe lanes, especially for a storage system where CPU speed isn't really relevant to total I/O throughput, you're better off going with an EPYC system.
To me this sounds like a customer willing to throw money at Intel for even higher binned than normal chips. On the one hand it seems a bit silly, on the other - provided Intel can find enough chips to meet the desired performance target - charging that customer 50% more is a big pile of free money.
Yep. Don't know that this is the use case, but it's worth a lot to high-frequency traders to be just a *little* lower-latency than competitors, which could drive stuff like this. Wonder if that price would hold up in a world where trades settle on the hour. :)
I'm curious what other off roadmap chips Intel has.
No Cascade Lake has on package Omnipath though Intel has hinted that they are holding them back until the 200 Gbit controllers are ready. Though at this pace, that'll be alongside Cooper Lake or Ice Lake-SP now.
Intel also had on package FPGA options with only one model publicly announced previously: the Gold 6138P. However there were supposed to be far more options off road map. Intel still is talking about this but with the 10 nm delays and lack of public parts, I would presume that all this is now off roadmap.
This CPU is DOA. Rome 64C will offer almost twice the throughput in MT workloads at the same or lower power consumption, and likely half the price or less.
Intel Xeon basically is only relevant heavy AVX code, up to avx512, that requires lots of branches or dependency, otherwise you should probably be writing it for a GPU or vector processor. (A GTX 1660 Ti has higher FP32 throughput than this Xeon 28C at an order of magnitude less money).
Intel needs its 10nm node right now, to have any chance of even competing against Rome based epyc. Even Intel internal memo acknowledged this.
The title reminds me a little bit of sketch by Daniel Tosh. I'll have to clean it up obviously, but the gist is that he's explaining to his girlfriend that it is not her versus a '6' he meets on the road. It is the girlfriend versus EVERY '6' he meets on the road.
So to say that $5500 for 300 MHz is a bit disingenuous. Instead, it is more like ALL of the 300 MHz. I mean, sheesh 28 cores! Insane! So is $5500, but the comparison holds.
I've seen Intel CPU variants in AWS which are unlisted on ark.intel.com. Some are a bit quicker with a few more cores. I've always assumed that when Intel do the final die test, they fuse out bad cache and cores. Ones which happen to be perfect, not needing the spare silicon, can then be sold at a premium to select customers.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
81 Comments
Back to Article
psychobriggsy - Thursday, July 18, 2019 - link
I guess AMD will have a 32C Rome SKU that destroys this at a significantly lower price and TDP in the near future.It's only hope is clock speed, but TSMC's 7nm seems optimal to around 3.3GHz before power goes up more than linearly with speed, so if AMD were to make a clock-speed targetting 32C SKU they should be able to.
MonkeyPaw - Thursday, July 18, 2019 - link
I think they could go past 4.0ghz at 32C and still match Intel’s power budget. The 3950x is only 105W, so if AMD wants to play with 215W, they could aim pretty high. I bet most of the best chiplets are already heading to this market to begin with due to better margins.edzieba - Thursday, July 18, 2019 - link
I wouldn't bet on it. While there are power saving son the 7nm 'core' dies, there is also power loss on the IF interconnects (all 8 of them) plus the large not-a-Northbridge die that remains on an older process. We saw with Zen that for high core count Epyc more power was consumed by the uncore rather than the cores themselves (https://www.anandtech.com/show/13124/the-amd-threa...eek2121 - Thursday, July 18, 2019 - link
EPYC Rome is not slated to reach 4.0 GHz if leaks are to be believed. However it's important to note that Intel's boost clocks are likely single core boost, with all core boost being a bit lower. Zen 2 has already been shown to be quite the formidable CPU even at lower clocks. Even with a rumored boost of 3.4 GHz, chances are AMD will have a clear win in the data center. After all, when Intel charges 3X more for 10% more performance...well, as the joke goes: "Nobody ever got fired for buying Xeon...until NOW!"eek2121 - Thursday, July 18, 2019 - link
Intel's TDP listings are not the actual TDP, nor are they power consumed. All IO is handled by the IO Die itself, so AMD can have as few or as many chiplets as it wants. Thus, an 8 core, well binned, EPYC ROME CPU could indeed hit a 4.7 GHz Boost if it only has a single chiplet.Gondalf - Thursday, July 18, 2019 - link
Still they can not, the low turbo of Epyc 2 line say something about TSMC low power process huge issues. On Ryzen 3000 line (performance process) the turbo frequence is obtained with a too high voltage applied to the core, moreover on all cores up the cpu overclock is pretty bad and spread an insane power.It is a fact that at least for now the single thread performance of Epyc 2 line is suboptimal, and Intel beat badly the brand new AMD server line in responsiveness.
A good reason to say that it is not all rosy at Sunnyvale and many critical workloads are right now forbidden to Epyc.
Death666Angel - Thursday, July 18, 2019 - link
You seem to have a lot of great information that is nowhere to be found on the sites I frequent. Where can I see the detailed information on Epyc 2 single thread performance? I have not seen anything about clockspeeds so far. What I have seen is a lot of speculation that AMD is saving the best chiplets for TR and Epyc. Considering that TR2 hit the same peak clocks as Ryzen 2000, the conclusion might be to say that Epyc will hit the same peak clocks as Ryzen 3000. But I can only speculate, you seem to have all the answers.eek2121 - Thursday, July 18, 2019 - link
Epyc "ROME" or EPYC 2 as you are calling it is based on the Zen 2 core. Quite frankly, based on what we have seen thus far, it doesn't matter where clock speeds fall (3.4 GHz is the rumored boost clock, and a bit under 10 grand is the rumored top of the line price point for a 64 core, 128 thread "ROME" CPU). Intel is going to have a hard time selling these things when "ROME" launches regardless. After all, would you spend $18,000 on a 28 core CPU just because single core workloads can boost to 4.0 GHz? Or would you instead spend $6000 on a 64 core/128 thread CPU that only boosts to 3.35 GHz? That's 1/3rd the price for more than 2X more cores and a slight clock speed reduction that matters very little for most workloads that these CPUs are going to be handling. Also, as others have mentioned, AMD could easily hit higher clocks than rumored.remosito - Wednesday, July 24, 2019 - link
> would you spend $18,000 on a 28 core CPU just because single core workloads can boost to 4.0 GHz? Or would you instead spend $6000 on a 64 core/128 thread CPU that only boosts to 3.35 GHz?For our database server, I'd pick neither. I'd get dual 8-12 core highest boost cpus I can get. They are usually (last few gens) in the 3000-4000$ range.
mike_bike_kite - Tuesday, August 6, 2019 - link
"For our database server, I'd pick neither. I'd get dual 8-12 core highest boost cpus I can get. They are usually (last few gens) in the 3000-4000$ range."Wouldn't you do better spending the money on maxing out the RAM and getting as much on fast SSD as you could. Obviously it depends on workload but wouldn't the new 3900X, 64GB RAM and the fastest PCIe4 M.2 SSD's offer better performance for less? I guess it depends where the bottleneck is in modern database servers but my "guess" would be on accessing the data rather than processing it. This isn't advice as I've never spec'ed database servers but I would genuinely be interested in any real world tests out there.
Irata - Friday, July 19, 2019 - link
AMD will really have a hard time in the server market where single thread performance matters a lot more than density, power consumption, IO, total cost (including cooling).....[/s]
web2dot0 - Friday, July 19, 2019 - link
Server market care about single thread?!? You mean most of the time the CPU is idling LOL.Irata - Friday, July 19, 2019 - link
You missed the [/s] tag.HStewart - Thursday, July 18, 2019 - link
I wish for once some one here, would keep there comments to topic, I personally refused to comment on AMD unless someone brings up something against Intel or incase of XPS 15 2in1, my gpu like drivers from AMD and such.PeachNCream - Thursday, July 18, 2019 - link
"...I personally refused to comment on AMD unless someone brings up something against Intel..."That's because you have an irrational personal bias and are incapable of widening your thought horizon to the point of being able to overcome that limit.
HStewart - Thursday, July 18, 2019 - link
"That's because you have an irrational personal bias and are incapable of widening your thought horizon to the point of being able to overcome that limit."
There is no irrational bias - in fact it very rational - I prefer Intel because they are the one that actually create the x86 based CPU and was not cloned from it.
But you notice I mention "Personally", that was one purpose for two reasons
1. I knew it was going to cause a response from AMD bias fan person and it did
2. It is true that others have there personal opinions about products and I just wish they keep it to themselves. Just better to agree to disagree
As for my thought process, you have no idea what experience I have, just to let to know I once work for an operating system and found a bug in IBM 486SLC cpu where chip add it address lines inverted when switching between 286 and 386 protected modes. This was at my first job for almost 7 years and there are few developers on this planet that can even comprehend what I am discussing.
Let's be honest, do the average customer of computer actually read these comments. In fact that average customer care about it and especially desktop chips - server chips have there own specially.
But most customers use it for word processing, spreadsheets, internet and emails - but I have 30 years of technical knowledge in these area include low level CPU design and even though I do application program - the low level parts still interest me
jdw912 - Thursday, July 18, 2019 - link
Your reply almost sounds like a copypasta where you're just bragging about past experience. Look, it's fine to have preferences and even biases if you can acknowledge them. And it seems like you do acknowledge your own, which is great! But people on the internet also have their personal preferences and it shouldn't blow your mind that people are bringing up AMD on an article like this.Skeptical123 - Thursday, July 18, 2019 - link
I think regardless of any bias HStewart is right in his criticism that every almost every story is inundated with comments praising AMD and bashing Intel. Regrless of if there comments are right most of the time the form it's done in is not a place for it. And is just off topic, even by internet standards.Also while I can certainty understand see what HStewart as "bragging" how else would you suggest he states his justification. For example if he said x is try but you disagree and he replies no I'm a doctor x is true. You can't say no your just bragging. Even if that is the case it is not what matters and is just deflecting from the point. Also I don't like to make thing personal but PeachNCream while I empathize with your passion from your past comments I can't take what you say to seriously.
Spunjji - Monday, July 22, 2019 - link
AMD articles were historically "inundated" with comments praising Intel and bashing AMD before Zen came out. This is what tech discussion pages look like - especially when the balance between the two remaining x86 CPU competitors begins to shift. Given that the article is about how much Intel charge for an extra 300Mhz on your 28 core CPU, the imminent appearance of a competitor who tends to provide better value is most certainly an on-topic discussion.HStewart is not a logical individual. His claim to authority here makes as much sense as claiming expertise on dental health based on having once been a proctologist. If you find yourself in agreement with him, I'd encourage taking a short break from the internet.
PeachNCream - Thursday, July 18, 2019 - link
You're mistakenly projecting your Intel bias outward. Like most people, I don't care what branding is on a CPU. It's amusing that your defense is to attempt to lawyer over the word "personally" and then make claims about your personal experience that do not justify your blind loyalty.HStewart - Friday, July 19, 2019 - link
I actually don't care either, what I don't like it people bashing Intel - with AMD statements - I never bash AMD products unless they bash Intel first. Just sounds like a bunch of uneducated discussing bashing another product because one thinks one product.I not saying Intel needs does not need to make improvements, In fact AMD helps Intel moving toward the future - I just think most people really don't care - about high core counts and high frequencies.
It not blind loyalty, I mention before it because Intel created the original product - x86 processors and because of stupid IBM back in the 1980's, AMD came in to picture because of 2nd source. My first computer that I purchase had an AMD processor on it was 386 clone - not sure the place advertise that fact.
Times have change now for most people and organizations, desktop are part of past - work have notebook connected to monitors. so you can bring the computer to board room. Days of desktop computer in work force is going way.. even now we have workstation based notebook with Xeon processors
But if there is a threat to Intel, it not the desktop AMD chips but instead iPad and Android tablet, which at my last job, I would have died if I supported Android on phone or tablet. So for windows based computing I prefer Intel because of long history of personal good results with them.
Probably if you want to say what has hurt me with AMD is there video, back in older once AMD purchase ATI - I did try AMD card that supposedly had good TV recording, I had bad luck with and I never switch from NVidia. But laptops became mainstream over the years and video card became less important. I still have nVidia on my Lenovo Y50 and I like the layout of Dell XPS 15 2in1 and even though I have concerns with it video, I was willing to give it try. It has some incompatibilities with two things 1,. Older games and 2. some professional 3d graphic software like Vue 2016. And also Steam VR.
But my life style is changes and graphics is becoming less important, but I still have older tech interested of younger days.
So main concern here not against AMD but blasting of AMD on topics discussing Intel product. So what does anybody really care. This kind of attitude surely will not make want to purchase an AMD product. I believe this attitude of blasting pro-AMD into Intel product would hurt AMD. I would say the same thing on any product
PeachNCream - Friday, July 19, 2019 - link
TL;DRQasar - Friday, July 19, 2019 - link
you should of read it Peach, at the least, its good for a laughMrPoletski - Thursday, July 18, 2019 - link
You know AMD, not Intel, created x64 right? that's what we're all using now.HStewart - Friday, July 19, 2019 - link
Yes but x64 is just an extension of x86 and it would have came in time.I pretty sure we will not x128 or x256 but you never know in the future.
Basically x64 is nothing with x86 base anyway.
CityBlue - Sunday, July 21, 2019 - link
> Yes but x64 is just an extension of x86 and it would have came in time.Maybe - but not from Intel, who didn't have a 64-bit x86 roadmap as their Itanium vision was rejected by the marketplace so it came down to AMD to invent x86_64 and give customers what they wanted rather than being force fed something expensive, complicated, incompatible and badly planned. As usual it is AMD that is innovating in the x86 space, while Intel does as little as possible and sweats it's assets/monopoly.
Korguz - Sunday, July 21, 2019 - link
i dont think intel had ANY plans to extend x86 to 64 bit, like CityBlue said, intel tried to transition to 64 bit via itanium, but because it had nothing to do with x86, EVERYTHING software wise would of needed to be recompiled to Itanium 64 bit, and the industry didnt want to do that. to be honest, some the cpu innovation has not come from intel, but from amd.steepedrostee - Thursday, July 18, 2019 - link
"...the low level parts still interest me."I bet they do..
Samus - Thursday, July 18, 2019 - link
Intel created x86, but every modern program and operating system are x86-64.Take a guess who created that...
HStewart - Friday, July 19, 2019 - link
Yes but x64 is just an extension of x86 and it would have came in time.I pretty sure we will not x128 or x256 but you never know in the future.
Basically x64 is nothing with x86 base anyway.
Xyler94 - Monday, July 22, 2019 - link
I dare you, even double dare you, to use X86 without the 64 bit instruction these days. Intel or AMD.even if X86-64 needs the base X86 instruction set, Intel would be useless today without the 64bit extension. There's one huge factor you're forgetting about X86, it's limited to roughly 4GB of ram. AMD beat Intel to the punch, it's a simple fact of life.
sa666666 - Thursday, July 18, 2019 - link
> As for my thought process, you have no idea what experience I have, just to let to know I once work for an operating system and found a bug in IBM 486SLC cpu where chip add it address lines inverted when switching between 286 and 386 protected modes. This was at my first job for almost 7 years and there are few developers on this planet that can even comprehend what I am discussing.Your 20-30 year old experience is irrelevant to the current predicament that Intel is in. AMD is winning this generation. Accept it. Intel will come back eventually; there's no doubt about that. But they are not in the winning position RIGHT NOW. And nothing you can say will change that.
And you're not the only one on these boards with 20-30 years experience in these issues. Most of us just don't feel the need to brag about it, nor use it as a blunt hammer to try to force people to think that you know what you're talking about.
Santoval - Thursday, July 18, 2019 - link
"There is no irrational bias - in fact it very rational - I prefer Intel because they are the one that actually create the x86 based CPU and was not cloned from it."Bias is bias, whether you consider it rational or irrational. Actually by definition every bias is irrational, so "rational bias" is functionally equivalent to "logical prejudice". They are both contradictions.
The reason you "prefer Intel" is irrational (though most of our choices are irrational). The creator of an ISA, whether it's x86 or ARM or anything, *can* fall behind some other company which licensed that ISA. Intel are much bigger than AMD so they did not fall behind AMD (yes, they did - not in market share, of course, I mean in desktop CPU performance. AMD has now surpassed Intel in IPC and power efficiency, *almost* matched them in single core performance and continues to surpass them in multicore performance) due to lower funding.
In fact their much bigger size and much more money available for R&D are precisely what made them fall behind : it made them overconfident enough to try multiple new technologies at the same time with their 10nm node, making it so immensely complex that they still have not managed acceptable yields with it, not even with its supposedly "fixed" 2nd gen 10nm+ iteration that Ice Lake is to be fabbed with.
Intel fell behind due to node issues, not μarch issues. If they had faced no issues with their 10nm node -for instance if they had made it quite simpler and less dense- they would have already released Ice Lake (Sunny Cove + Gen11 iGPU), Tiger Lake (Willow Cove + Gen12 iGPU and even Alder Lake (Golden Cove + Gen13? iGPU).
No comment about your computing experience, which I fail to see its relevance here. In this context it smells a bit like an argument from authority fallacy to me.
blppt - Thursday, July 18, 2019 - link
What is truly remarkable about this guy is that every single desktop chip we run nowadays is based on an AMD design. If Intel had its druthers, we'd all be running IA-64 Itaniums right now.So, I guess in this case, he should be shunning EMT64 as a 'clone', right?
LOL
Qasar - Thursday, July 18, 2019 - link
HStewart " I prefer Intel because they are the one that actually create the x86 based CPU and was not cloned from it. " so does that mean if AMD released a cpu that was in EVERY way better then what intel has, in performance, power usage, cores, etc, while costing much less, you would still buy the intel chip, just be cause of the above quote ?? if your answer is yes, then im sorry, but that is the dumbest thing i have ever seen, and shows you either really are the biggest intel fan on here, or, you have more money then you know what to do with it.Skeptical123. " in his criticism that every almost every story is inundated with comments praising AMD and bashing Intel. " have you NOT seen his previous posts ?? in practically EVERY article about AMD, that is good news for the company, HStewart will post some crap about how its not a big deal, and then go on to try to turn the good amd news, into bad, then go and praise intel in some way, what you are saying in your 1st paragraph, is the EXACT same thing HE DOES.
duploxxx - Friday, July 19, 2019 - link
so HStewart is only running 32bit code? since the 64bit code that Intel is using is actually sourced from AMD :DSpunjji - Monday, July 22, 2019 - link
"they are the one that actually create the x86 based CPU and was not cloned from it" isn't a rational basis for a bias. The property of 'who did the thing first' has no bearing whatsoever on how good these companies are currently.Given that AMD created the x64 superset of x86 that Intel subsequently borrowed back, even according to your way of seeing things Intel and AMD are now arguably even.
I'm forced to conclude that you really don't have a rational basis for your preference.
FreckledTrout - Thursday, July 18, 2019 - link
Except AMD is expected to do very well against Intel in the high core count chips. I think it's perfectly reasonable to mention AMD here. Seeing the 3700x/3900x reviews in multithreading I think Intel is about to get stomped in the server space. AMD's SMT looks very nice this generation. With that said I still am reserved and will wait to see reviews.sa666666 - Thursday, July 18, 2019 - link
We'll keep our comments on topic when you do. Guess when that will be; somewhere north of 10 years after hell freezes over.And you admit that you only respond when someone speaks against Intel (and don't even care whether they deserve it). And saying "I knew it was going to cause a response from AMD bias fan person" proves that you're just trolling, and that's all you ever do. Don't be butthurt that you're being called out on it.
Look, we all get that Intel is the Lord and Saviour to you. But this isn't about Intel. It's about your blind zealotry of a company that doesn't care about you, and you EXTREMELY ANNOYING and TROLLING comments. You are looking to get a rise out of people, and we will keep calling you out for it.
Qasar - Thursday, July 18, 2019 - link
sa666666666 you are also forgetting that he will also speak against AMD if they announce some good news, and say some BS about it to make it look bad, then right after, praise intel in some form, the announcement a bit ago about the Cray supercomputer is and example of that.HStewart - Friday, July 19, 2019 - link
Only if some one bash Intel, you miss that part. I am sure some one in the discussion bash it.Or in case of Cray supercomputer bash Aurora.
Korguz - Friday, July 19, 2019 - link
only if some one bash intel ?? are you serious ??? you gladly bash amd ANY chance you get, even if no one said anything negative about intel 1st... you will find some way to bash them, then turn around and praise intel.. come on HStewart, enough with the bs. your long winded post above.. is also complete bs. you even go as far as making things up, to make amd look bad, and intel to look good.Qasar - Friday, July 19, 2019 - link
HStewart, even if some one didnt bash intel in their comment, but instead suggested AMD as a better alternative, cause of price, performace, power usage, or all 3, you take it as an intel bash, and reply with an amd bash. so you are the one that bashes amd 1st most of the time, not the other way around.AshlayW - Saturday, July 20, 2019 - link
He's financially invested. :)azfacea - Thursday, July 18, 2019 - link
HStewart again LUL. He thinks we don't know who he is LULPeachNCream - Thursday, July 18, 2019 - link
What surprises me is that there's like a dozen people saying, "Hey, you're being cluelessly brand loyal to a fault," and his mind can still find an escape route which doesn't endanger his poorly grounded thinking. A non-obsessed, rational person would have gone through some sort of internal self-assessment a long time ago and concluded that dialing it back a bit at the very least, would be a good idea.arashi - Friday, July 19, 2019 - link
You cannot make him admit his bias when his job as a marketing peon with a made up a/iamverysmart backstory depends on that.Qasar - Friday, July 19, 2019 - link
i think he kind of admitted he is bias with this line " I prefer Intel because they are the one that actually create the x86 based CPU and was not cloned from it " and he never answered my direct question to him about if AMD released a cpu that was in EVERY way better then what intel has, in performance, power usage, cores, etc, while costing much less, if he would still buy the intel chip.. my guess, is he would still by the intel chip because is blind loyalty to intel wouldnt allow him to buy a non intel chip, even if the intel chip cost 2 or 3x more. " In fact AMD helps Intel moving toward the future " so intel cant move forward WITH OUT AMD's HELP ?? thats ludicrous. intel is QUITE capable of moving forward on its own, but it CHOOSES not to, instead, intel CHOOSES to stagnate the cpu market like it has over the last few years, and milk its customers for as much as they can.bobhumplick - Thursday, July 18, 2019 - link
probalby. in server environments the amd cpu will be much better. but in latency sensitive tasks maybe not. a lot of people running high end autodesk software like maya or 3dsmax would probalby run better on intel cpus. besides the cpu is such a small cost of a server that the price isnt that significant.and support both hardware and software support is an intel selling feature. if a chip goes down intel will send out a courier and replace it. even for your desktop chip. amd are missing some key software tools that intel have. mostly people just use the intel tools though so thats not a huge difference.
the final most important piece of the puzzle is supply. even if everybody in the world wants to go amd they cant. amd just cant make enough chips to replace even half of intels sales. people have to buy intel. they could buy amd out and then move some stuff over to arm and they would still have to buy intel. at least somebody would
web2dot0 - Friday, July 19, 2019 - link
AMD will have a 64C Rome SKU that will destroy all Intel SKYs. And it won't be $15k. The clock is ticking ... and tocking ... LOL.twotwotwo - Friday, July 19, 2019 - link
Curious about the 24C range, since 6C=>8C chiplets is like a >50% price jump (b/c yields I guess). Also, of course, about AMD's pricing and Intel's (pricing/product) response.chada - Monday, July 22, 2019 - link
cores, cost, threads. Pick 1 or 2 depending on your needs.mmrezaie - Thursday, July 18, 2019 - link
I always wonder why Intel has so many SKUs. I don't think anyone wants this many choices since they are hardly different. I like to see choices but in such a small increments, nah? Is this marketing? Production forces them to it e.g. silicon variation?TheWereCat - Thursday, July 18, 2019 - link
Its just a ton of different bins mostlyedzieba - Thursday, July 18, 2019 - link
These are Intel's equivalent to AMD's 'semi custom' service, were Intel will produce an SKU to the request of a specific vendor for a specific product. It's why the 'list price' is a bit of a misnomer, as they're not listed and that price doesn't really reflect what the companies buying these variants are paying.HStewart - Thursday, July 18, 2019 - link
I am curious what L M and Y mean with same specsSarahKerrigan - Thursday, July 18, 2019 - link
M: Supports 2TB RAML: Supports 4.5TB RAM
S: Speed Select
HStewart - Thursday, July 18, 2019 - link
Thanks so this shows one of reasons why there is so many different products.GreenReaper - Thursday, July 18, 2019 - link
Yes, except that at least two of those are likely to be artificial limitations for product segmentation (read: extracting the most profit from those who will pay for it).Is this a bad thing? Not necessarily. But it's definitely a thing, and annoying if you want those features.
edzieba - Friday, July 19, 2019 - link
No segmentation, die harvesting. When you have a very large die, you have lots of individual component parts you can fuse off if they have a defect. Cores, PCIe interfaces, memory interfaces, etc. If you did not harvest these dies and only sold 'perfect' dies as a single SKU line, then you would have a very small volume of parts and high prices for those parts. By finely binning dies into a large number of SKUs based on yield, then you have many more sellable dies and customers can buy ones that lack features they do not use to reduce outlay.Elstar - Tuesday, July 23, 2019 - link
Or some of both. Unlike Skylake, Intel seems to be including 26 cores worth of L3 on some 24-core Cascade Lake chips, which implies that yields are up (and that Intel would rather have a bigger performance gap between their 24- and 28-core parts).bobhumplick - Thursday, July 18, 2019 - link
all those skus arent for us. they are for oem's. intel sells so many more chips to people like dell and hp than to us that we are basically insignifcant. thats why all the sales numbers with amd beating intel are not nearly as damning for intel as people think...yet.but anyway if oems can save 5 dollars per chip that can add up to millions. so more skus allows oems to tune their costs to their liking. you can pretty much overlook them. they arent meant for us
yeeeeman - Thursday, July 18, 2019 - link
The sooner AMD gets out Rome with 64 cores the better.duploxxx - Friday, July 19, 2019 - link
true for competition and that is already showing.without rome even available the recent intel refresh / intel already dropped it cpu pricing with 30%.(for the 1000s unit price)
but still horrible what you have to pay for those high core count sku. knowing spectre, melt and zombie i wonder if those high end buyers ever think twice and revise there portfolio/buying behavior
HStewart - Thursday, July 18, 2019 - link
I think a big difference with these CPU's is that they are not just Dual CPU setups but 4 or even 8 CPU motherboard options. I believe this is why they are call Xeon Scalable.Plus Xeons have advance instruction sets like AVX 512 that not found in other systems and also chipsets are have better IO.
Yes it old but my Dual Xeon 5160 was faster than most machines until the i7's came out.
SarahKerrigan - Thursday, July 18, 2019 - link
They do indeed support 4s and 8s configurations - but 4s+ x86 is very much a niche market.HStewart - Friday, July 19, 2019 - link
Keep in mind 4 16core cpu';s is same amount of cores as 2 32core cpus.There is also likely IO difference with multiple cpus and likely it because of less core per cpu, can be run at faster speed.
Korguz - Friday, July 19, 2019 - link
yes.. but if some one has a 4s board with 4 16 core cpus, and they upgrade those to 32 core.. they have twice the cores.. in the same space.FreckledTrout - Thursday, July 18, 2019 - link
All of that is true but Intel does downclock for AVX512. It will be interesting to see these chips go against EPYC.HStewart - Friday, July 19, 2019 - link
But the speed difference when using AVX512 vs normal AVX2 is significant and more than clock speed could ever make the difference.And if you don't have AVX512 based application, should be able to disable it and run at full speed.
AVX 512 is not in mainstream computers until Sunny Cove, but it would be bad thing if AVX 512 not active slows things down. This part could be wise it not main stream yet.
Qasar - Friday, July 19, 2019 - link
" AVX 512 is not in mainstream computers until Sunny Cove " um... i think you might be wrong on that, even just a little, as skylake X and cannon lake both have avx512ravib123 - Friday, July 19, 2019 - link
The other thing the scalable has going for it I believe is pci-e lanes. I use scalable to build out full lanes of pci-e based flash storage arrays. No AMD options that fit the bill, even the consumer side they still seem very focused on direct cpu performance.I don’t see AMD losing to the scalable any time soon.
Intel always release these special order tray only processors. In the last series of xeon there was a tray only I used all the time because lower power and full speed and the cost difference was about 300$. I can’t tell you how many times customer cooking equipment failed and I was happy with the lower heat cpu resulting in no hysteria.
Spunjji - Monday, July 22, 2019 - link
This doesn't make much sense. EPYC has a lot more PCIe lanes available than Xeon Scalable - 128 PCIe 3.0 on the first generation. Surely that would be superior for building PCIe-based flash storage arrays?phoenix_rizzen - Wednesday, July 24, 2019 - link
EPYC motherboards have 128 PCIe lanes direct from the CPU (whether single socket or dual socket).Xeon SP has 36 PCIe lanes direct from the CPU socket. In order to match an EPYC system, you NEED a 4 socket motherboard, 4 separate CPUs, and 4x the number of RAM sticks to fill it. How is that "better"? You're paying extra for the multi-socket motherboard, extra for the multi-socket CPUs, extra for the DIMMs, extra, extra, extra.
If you want PCIe lanes, especially for a storage system where CPU speed isn't really relevant to total I/O throughput, you're better off going with an EPYC system.
DanNeely - Thursday, July 18, 2019 - link
To me this sounds like a customer willing to throw money at Intel for even higher binned than normal chips. On the one hand it seems a bit silly, on the other - provided Intel can find enough chips to meet the desired performance target - charging that customer 50% more is a big pile of free money.twotwotwo - Friday, July 19, 2019 - link
Yep. Don't know that this is the use case, but it's worth a lot to high-frequency traders to be just a *little* lower-latency than competitors, which could drive stuff like this. Wonder if that price would hold up in a world where trades settle on the hour. :)Kevin G - Thursday, July 18, 2019 - link
I'm curious what other off roadmap chips Intel has.No Cascade Lake has on package Omnipath though Intel has hinted that they are holding them back until the 200 Gbit controllers are ready. Though at this pace, that'll be alongside Cooper Lake or Ice Lake-SP now.
Intel also had on package FPGA options with only one model publicly announced previously: the Gold 6138P. However there were supposed to be far more options off road map. Intel still is talking about this but with the 10 nm delays and lack of public parts, I would presume that all this is now off roadmap.
AshlayW - Saturday, July 20, 2019 - link
This CPU is DOA. Rome 64C will offer almost twice the throughput in MT workloads at the same or lower power consumption, and likely half the price or less.Intel Xeon basically is only relevant heavy AVX code, up to avx512, that requires lots of branches or dependency, otherwise you should probably be writing it for a GPU or vector processor. (A GTX 1660 Ti has higher FP32 throughput than this Xeon 28C at an order of magnitude less money).
Intel needs its 10nm node right now, to have any chance of even competing against Rome based epyc. Even Intel internal memo acknowledged this.
chada - Monday, July 22, 2019 - link
The title reminds me a little bit of sketch by Daniel Tosh. I'll have to clean it up obviously, but the gist is that he's explaining to his girlfriend that it is not her versus a '6' he meets on the road. It is the girlfriend versus EVERY '6' he meets on the road.So to say that $5500 for 300 MHz is a bit disingenuous. Instead, it is more like ALL of the 300 MHz. I mean, sheesh 28 cores! Insane! So is $5500, but the comparison holds.
speculatrix - Wednesday, July 24, 2019 - link
I've seen Intel CPU variants in AWS which are unlisted on ark.intel.com. Some are a bit quicker with a few more cores.I've always assumed that when Intel do the final die test, they fuse out bad cache and cores. Ones which happen to be perfect, not needing the spare silicon, can then be sold at a premium to select customers.