I really appreciate the article + video release. I've never been able to enjoy podcasts due to constant context switching in every part of my day.
@JonMasters2 жыл бұрын
Yes!
@raylopez992 жыл бұрын
I like how the author dumbs down things into a mnemonic, perfect for me, as I'm no longer in this field and skimming and hand waving is just fine.
@FireFalcon2 жыл бұрын
Here Here
@Atiyoga Жыл бұрын
I have been watching STH KZfaq channel a lot and for a long time. Patrick and STH have been doing a fantastic job covering Enterprise Technologies in depth and in width. Thank you and keep up the great work. Your work is very meaningful and helpful!!! Appreciate it.
@ServeTheHomeVideo Жыл бұрын
Wow! Thank you so much!
@jolness12 жыл бұрын
THIS is Patrick from STH! ^ Always makes me smile. And when he says to have an awesome day. Plus the information is great always!
@thejo63312 жыл бұрын
Thanks for including Ampere in this roundup! That stuff is really heating up.
@__--JY-Moe--__2 жыл бұрын
Interested to hear your thoughts on the proposed OCP standard MB form factors!
@jedcheng55512 жыл бұрын
There is a narrative which general purposed and domain specific accelerator swap their importance every 10 years The intel and Arm comparison is definitely an example
@lucasrem2 жыл бұрын
intel was the future that never came, they are stuck in 1970 ARM was always the better design!
@roysigurdkarlsbakk38422 жыл бұрын
…and there's the underdog RISC V - I really look forward to seeing what's going to happen there :)
@ServeTheHomeVideo2 жыл бұрын
Still a long way away from this in the server space.
@roysigurdkarlsbakk38422 жыл бұрын
@@ServeTheHomeVideo Well, it's the new kid on the block, but it seems to be running quickly. Give it a few years and maybe…
@PhilfreezeCH2 жыл бұрын
@@roysigurdkarlsbakk3842 As far as I know there is current,y no real push into the server space (and considering how server companies make buying decisions based on maturity and support, this makes sense). RISCV is currently trying to conquer the low power sector, especially ML on edge and there are some efforts to bring RISCV to low end consumer desktops and similar sized computers (think Raspberry Pi or Intel Atom).
@roysigurdkarlsbakk38422 жыл бұрын
@@PhilfreezeCH I'm aware of this, but given the lack of licensing cost to whomever wants to make a RISC-V chip, I would be surprised if noone catches the chance and tries to make some for the server market.
@AndersHass2 жыл бұрын
@@roysigurdkarlsbakk3842 if you need custom solution for very specific use case then going risc v could be worth it (then you still need to have that custom solution added on).
@johndoh51822 жыл бұрын
More cores = better only if you have a workload that pushes those cores. Most home users can't saturate a current Zen 3 16 core part unless they do video rendering or run a lot of concurrent VMs, which isn't most people for those who feel like dropping a comment about what THEY do. Today's CPUs are SO fast that probably 95% of home users are already served, except in gaming. NEXT generation CPUs from AMD in particular are going to once again push the balance of GPU-CPU back at the GPU, to where only the fastest GPUs will start to put more of the load on the CPU when gaming, and I mean the fastest of the NEXT gen GPUs. Zen 4 CPUs for gaming is going to be insane. Intel will be similar but are going to use more power, probably about 2X what AMD CPUs use to do the same thing. And then to top that off AMD will release Ryzen Vcache parts that will make various games run even faster. Bottom line is unless you're a power user that needs lots of cores and you know who you are, if you want a top end gaming system, Zen 4 is going to be what you're looking for. If you need lots of cores, AMD won't have that until Zen 5 unless you step up to a server or WS CPU in which case they're going up to 96 or 128 cores depending on the variant of EPYC. And I think they're going to smoke pretty much anything on the market, but Zen 5 is going to be more insane. Intel CPUs for server will have to rely on e cores to get high core counts and that's fine for loads that are met by RISC, since that's basically what an e core is. It's a scaled back X86-64 CPU along with being single thread. I'm personally waiting for Zen 5. AMD's improvements with accelerators, most likely a form of big-little along with being on an advanced node means their CPUs are going to be mind blowing IMO. Zen 4 Genoa is 92c/184t. I don't know who competes with that. Is there currently an ARM processor with 184 cores?? Zen 4 Bergamo is 128c/256t. No one touches that. So I don't understand the whole AMD is in the middle thing. Also, regardless of what Intel or ARM offer, there's a LOT of general compute HPC servers that have switched to AMD. Their growth in the world of server has been very impressive and with Zen 4 offering up to 256 threads I don't think Intel's accelerators can make up the difference.
@badrnaji84852 жыл бұрын
thanks for your reply very helpful to understand the market
@gg-gn3re2 жыл бұрын
AMD's purchase of xilinx definitely makes it look like they want huge addition of accelerators and maybe FPGAs on their chips as well. It will be interesting if we see general purpose CPUs with FPGAs and programs taking advantage of it.. like during the startup of a software, lock your FPGAs into the type of data that program processes and it runs 40x faster or whatever
@hikaru-live2 жыл бұрын
Intel went there then backed out of it, putting an Stratix into a Xeon package.
@poofygoof2 жыл бұрын
Lessons were definitely learned on the 6138P. It will be interesting to see what happens with FPGAs and CXL.
@poofygoof2 жыл бұрын
@@hikaru-live It was Arria, not Stratix, in the 6138P.
@coa81092 жыл бұрын
FPGAs don’t necessarily speed up the processing but they are definitely the most efficient
@samlebon98842 жыл бұрын
They would plug the FPGAs to Infinity Fabric; that way they would play nice the CPU/GPU ecosystem. IF is AMD"s secret sauce
@meco2 жыл бұрын
What about a Tiny Mini Micro addition with ARM SBC’s? The Rockchip RK3588 boards seem to be a great homeserver candidate. (around 150$ish)
@ServeTheHomeVideo2 жыл бұрын
That would probably be its own series if we do it.
@HiltonT692 жыл бұрын
T-Rex the series - Tiny ARMs. :)
@everyhandletaken2 жыл бұрын
@@HiltonT69 lol
@arm-power2 жыл бұрын
RK3588 has Cortex A76 cores from 2018. - IPC around AMD Zen 1 or early Skylake - this core is used in Ampere Altra and Graviton 2 - much smaller than x86 core, at 7nm each A76 core has about 1.2 mm^2 incl 512 kB L2 cache . today's new Cortex X3 has 76% higher IPC than A76.... - it's shame RK3588 in 2022 has 4 year old A76 when 5 gen newer X3 is available - BTW the Cortex X3 has 30% higher IPC than AMD Zen 3 in Milan EPYCs.... which means 18% higher IPC than new upcoming Zen 4 (1.3 / 1.1 = 1.18) Graviton 3 - uses 5nm TSMC and Neoverse V1 core (derived from Cortex X1 with added 256-bit SVE SIMD extension) - Neoverse V1 core has about 5% higher IPC than Zen 3 (in general code) - SVE is for HPC (competitor to AVX512) and supports vector and MATH math (AVX512 lacks matrix math) - SVE supports FPU width scalable from 128-bit up to 2048-bit (AVX512 has 512-bit fixed) - SVE was developed for and 1st used in Japanese super computer Fugaku (Fujitsu ARM A64 FX cpu) - Fugaku was the fastest supercomputer on the world (only CPU, there are no GPUs used at all) now in 2022 dethroned by GPU SCs - still today the fastest CPU-only SC Interesting bits: - since 64-bit ARMv8 ISA is not compatible with 64-bit ARMv8/9 ISA (different encoding) there is needed two decoders for backward compatibility - since Cortex X2 designers in ARM Holding removed 32-bit compatibility and saved transistors (x86 cannot do that due to using extensions in variable length CISC coding, sure this has pros and cons) - pure 64-bit CPUs for max performance and IPC
@meco2 жыл бұрын
@@arm-power While everything you said is true I don’t think the newer cores could compete from a price perspective (or even be available) for consumers. I just replaced my x86 home “storage” server with an RK3568 Board and am still able to fully use the 2.5GbE-Nic while cutting power draw by at least 3x on full and up to 10x on idle.
@ander14822 жыл бұрын
Thank you so much Patrick, very useful to get a quick idea of the situation.
@OkkelsSkov2 жыл бұрын
This video leaves me with an interesting question: what does STH think will drive the direction of server market technologies over the next say 5 years, is the solution demands from ie. software and cloud providers or is it the technological progress and iteration from the chip makers/solution providers?
@ServeTheHomeVideo2 жыл бұрын
Cloud providers. The Enterprise server vendors (Dell, HPE, and Lenovo) are no longer setting direction in the market.
@MrHav1k Жыл бұрын
This was very well done and really articulates the current landscape in an easy to digest manner. Thank you!
@johndoh51822 жыл бұрын
AMD is about to equal "Our cores are faster than yours" with Zen 4. Depending what you want to do with a system, like if you want to game, AMD is THE solution for Zen 4. Intel CPUs will be good, but they're going to use more power in trying to do what AMD does.
@technicallyme2 жыл бұрын
I'm not this will be true when raptor lake comes out. Seems like they are leaning into cache at least from the rumors
@johndoh51822 жыл бұрын
@@technicallyme Actually it will be MORE true with Raptor lake. I was talking about Zen 4 vs. Raptor Lake. Intel is adding more e cores if you're talking about desktop parts, AND they are clocking the CPus even higher. I just saw a leak which says the 13900K will use up to 350W as opposed to the 240W for the 12900K. And that makes sense and is what I expected. If you add 8 more e cores, that's about another 50W right there when running all-core loads (really that's the only reason why you buy CPUs like this). And NOW you clock 8 p cores and 16 e cores even faster, but Intel was already at the border of efficiency with the 12900K, in fact most reviewers felt Intel had already gone too far with it. So that puts you in that 340 - 350W territory. I don't see anything good about that. When METEOR LAKE comes out, we'll see, but Raptor Lake is just an improved Alder Lake. Meteor Lake is supposed to be on "Intel 4", and if it is and Intel adds more performance cores then they have a much better part. And I was really referring to all-core clock behavior. AMD is going to have much higher power limits than they did before which is going to allow their all-core clock speeds to be much faster than for Zen 3. This is where AMD is going to shine, because what it will take Intel to do with 350W, AMD is going to do about the same thing with 235W.
@Freshbott22 жыл бұрын
@@johndoh5182 We can't really know till it happens. It's easy to speculate but how long have people been saying this. When I built my PC everything seemed to say AMD was the better choice factoring power etc. In reality Intel offered lower cost and higher min frames and performed better across different memory clocks. I know that's not relevant to what these CPUs are meant for but it does tell you that power level and cores doesn't tell you everything. For constant CPU loads eg. Cycles the CPU I got is identical to the AMD one it was typically compared with. If your load doesn't fully feed all your cores all the time the e cores could be a big deal. You've kind of just said Intel has just consumed more TDP with extra cores and higher clocks to get performance and then that AMD has done exactly the same. The difference will probs come down to cache. AMD having extra room to bump up wattage might just lead them to hit the same wall Intel did where it gains them basically nothing.
@jfkastner2 жыл бұрын
Well done video, thanks. Hopefully Virtualization license costs and overhead will not 'eat up' all the gains from those 'new' architectures
@ServeTheHomeVideo2 жыл бұрын
In theory, that is where Intel's acceleration helps
@jurepecar90922 жыл бұрын
More like this please! :) And looking forward to kungpeng review, someone (we both know) was offering it to me three years ago ;)
@CommonCentsRob2 жыл бұрын
Luv this dude! Always over the top like 80's rock disc jockey. 😂😂😂
@michaelclement13372 жыл бұрын
Great summary. Most people want something compact and portable that plays videos, I'm not sure if CPU's are the issue for them. However there's a significant market for where performance is significant
@DrRussell2 жыл бұрын
For someone who really needs to learn so much from you, it would be really useful to know the optimal use cases for the different chips, from a software availability and hardware deployment perspective. For example, will Freenas be optimally served by an ARM system? Sorry if this is an oversimplification.
@ServeTheHomeVideo2 жыл бұрын
TrueNAS might be good one day on Arm, but for now it is x86. TrueNAS Scale may be the one that jumps more easily one day to Arm. Many storage vendors use QuickAssist for compression/ encryption as mentioned.
@Blissy11752 жыл бұрын
Really looking forward to Zen4 this coming gen, hope they keep the pressure on Intel so the two keep one-upping each other in various metrics.
@Hari-ur9ve2 жыл бұрын
ARM-has not been to successful in getting good hold of server market space due to absence of supportive ecosystem, proven software modules used in servers compiled for ARM (time-tested?) and many more.. would love to hear your views on next videos on when ARM will be a able to get a good foot hold on the same !
@ServeTheHomeVideo2 жыл бұрын
It is finally happening. That is something that I discussed when I spoke at the Cavium ThunderX2 launch in San Francisco. It is just taking a LONG time. That said, you are right that there is still a gap. Even yesterday's Solidigm D7-5520 NVMe SSD review we did on the main site yesterday there is a difference between Intel, AMD, Arm (Ampere/ Huawei), and IBM Power9 in terms of NVMe performance.
@arm-power Жыл бұрын
@@ServeTheHomeVideo Graviton 2 itself created 10% server market share in just one year. Cortex A76 core in Graviton 2 was the break point for ARM, because it has IPC similar to Intel Skylake or AMD's Zen 2. That's huge performance jump from Cortex A72 used in first AWS Graviton (A72 has IPC of AMD's Bulldozer or K8 Opterons which is really bad). A72 is 3rd core from A15/A57/A72/A73/A75 1st "high performance OoO" line up where A15 was fairly terrible. I have A72 in my Raspberry Pi 4 and it is very inconsistent, pretty good in scalar integer code (running Gnome linux desktop), while very slow in everything FPU/SIMD intensive (no wonder because A72 needs to break every 128-bit NEON vector into 2x 64-bit because it has 64-bit HW SIMD only (A76 has 4x more performance because it can compute 2x 128-bit FMA per cycle), also crypto SHA instructions are very slow which was probably OK for smartphone but it's pain at server/desktop). Simply put A72 is missing some key instruction features due to mobile focused design. A76 has double IPC than A72 and it is 1st core from completely new "desktop class" core line up (A76 / A77 / X1 / X2). Engineers learned lesson since A15. I have A76 at Linux desktop Debian/Ubuntu and it's HUGE jump from A72. X1 core in Graviton 3 is still A76 line up but 50% higher IPC and 4x 128-bit FMA doubles the A76 in Graviton 2. X1 has IPC at Intel's Raptor Lake level (little better than Zen 3). X3 core is 1st from entire new line-up. At least Graviton 4 and Nvidia Grace server CPUs should use this core in 2023. Cortex X3 has 20% higher IPC than X1 which put X3 as the most performant server CPU core in 2023. I really like how many executing units X3 has: 6x ALU + 2x Branch units = 8-wide integer scalar core. For comparison Zen 4 has only 4x ALU + 1x Branch unit and Intel Raptor lake has 5x ALU. Raw performance is about 50% higher for X3 on paper so those 20% IPC advantage is just beginning. X4 / X5 / X6 has pretty solid base to build on. Another topic would be special instructions for AI/ML like Intel's AMX. ARM has matrix support already in SVE from 2016 but there are SME (Scalable Matrix Extension) and brand new SME2. This is heavy focus to matrix math from ARM. Intel Sapphire Rapids with AMX is 1st gen and it's not even at market yet. ARM has 1st matrix gen out in HW (A64FX and Graviton 3), 2nd matrix gen is about to come out soon (in Neoverse V2 core in 2023?) and 3rd matrix gen has already closed specifications (CPU cores are developed right now, it takes 3 years at least). I do worry about AMD because they didn't mentioned when they are gonna support AMX. If AMX support will take them same long time as AVX512 did (almost 10 years delay after AVX512 announcement), then AMD is in trouble because ARM is already working on 3rd martrix gen HW and AMD has nothing.
@woomidyo2217 Жыл бұрын
As a data engineer, I think more memory and i/o bandwidth could be another key factor for the competition. Data processing implies distributed computing and over tons of cores and the main bottleneck is memory and i/o time. In that case more cores means nothing.
@ServeTheHomeVideo Жыл бұрын
This next-generation adds more caching plus transitioning to DDR5 w/ 8-12 memory channels per CPU, so memory bandwidth will go up. On the Intel side, there will be HBM Sapphire Rapids parts.
@andfru Жыл бұрын
How will Sparc and IBM Power and Telum CPU fit into this comparison? Understanding that their Server Platforms are limited regarding Server Vendors, but I think it would be interesting to see this as well on how they compare from a strategic point of view. But in general this was very helpful to understand.
@ServeTheHomeVideo Жыл бұрын
They are mostly focused on ensuring a slow market share loss rather than being parts to gain share.
@esra_erimez2 жыл бұрын
When are we going to have ARM based workstations that we can develop on for these platforms?
@ServeTheHomeVideo2 жыл бұрын
Yes. Their idea is more that people will use RPi or cloud instances.
@ThatNerdChris2 жыл бұрын
What's that cage around you for? Looks cool lol
@ServeTheHomeVideo2 жыл бұрын
Just trussing
@jeremybarber28372 жыл бұрын
More middle more better for me!
@Jsteeeez2 жыл бұрын
Kinda liked the older video thumbnail design more. This One just blends in on my subscription feed.
@wmopp91002 жыл бұрын
in the last few years AMD was "more RAM and interconnect, more better"
@thejo63312 жыл бұрын
Interested to hear your thoughts on the proposed OCP standard MB form factors!
@__--JY-Moe--__2 жыл бұрын
Thanks for including Ampere in this roundup! That stuff is really heating up.
@__--JY-Moe--__2 жыл бұрын
Im looking forward to the cpu market being more betterer
@ServeTheHomeVideo2 жыл бұрын
Maybe more around OCP Summit 2022.
@jeffwong13102 жыл бұрын
I'm a PC DIY user but server platform and workload knowledge are totally different animal. is there a place that I can educate myself from the ground up about the idea of varies server system and the reasons of the different servers for different workload purpose? Many thanks!
@jannegrey5932 жыл бұрын
I really wonder if first leaks about Bergamo were correct? They said that it probably was to challenge ARM in cloud services - going for higher density with compromises on some types of performance. Even if they were - it seems that ARM has managed to build CPU's that will be very competitive against such attack. Not to mention they have their own market share and applications. So I wonder if such dense cores, without some of "normal" x86 (or at least associated with HPC) capabilities will have a chance? Maybe as something to compete in the middle ground? To have something that can do both things so-so, depending on occasion? But most people in Servers know what their use is going to be. We will see. AMD is slightly behind it's earlier schedule, though compared to Intel that updated us that Sapphire Rapids (and indeed most Intel CPU's) will be delayed only last month, it is at least acceptable. Though I still hope that Sapphire Rapids and Fishhawk falls (or however it is called) will come and force AMD to re-evaluate it's decision (at least according to leaks) to abandon non-pro version of Threadripper. And hopefully increase the competitiveness of all Server CPU's. I don't mind that AMD for the first time is making more money than it needs on bare bones R&D, but I don't want them to go to crazy with prices. Which they've started with Zen 3, but it looks like it will only get worse.
@ewenchan12392 жыл бұрын
Interesting presentation of your perspectives re: server CPU space.
@jedcheng55512 жыл бұрын
One of the major ideas being pushed by Ampere is that they offer predictable performance. Because of turbo boost on their x86 competitors, there is a variation in terms of performance for customers at a different time or simply who you are sharing the CPU with. I really want to see if this argument is valid in terms of the end user because it is not possible for us to test this idea. For instance, x86 CPUs are very good at doing vector computations compared to Arm. If x86 CPUs at base clock can perform better than Arm in vector, it means this argument is invalid for vector (which is what I work with every day)
@ServeTheHomeVideo2 жыл бұрын
Not just turbo. Having SMT on does impact consistency.
@jedcheng55512 жыл бұрын
@@superkoopatrooper4879 of course I can do anything to my CPU in the bios. But I can't enter the bios of my VM on public cloud
@anno59362 жыл бұрын
I'm currently not much into dedicated servers, but what gets me hyped is the rumor AMD will include FPGAs in their chiplets as well. Pair that with maybe 8 cores and 8 CUs in an APU and that will be my workstation. ARM is also interesting; something like an Apple M1pro for half the price and upgradeable RAM would be nice. And then software needs to follow up, time to cut all that decades old tech
@YouTubeGlobalAdminstrator2 жыл бұрын
When are we getting a HP Microserver Gen11? The gen 10 Plus is ancient now.
@ServeTheHomeVideo2 жыл бұрын
Yea, I know. Hoping to get the Altra HPE soon.
@andre-le-bone-aparte2 жыл бұрын
Possible Podcast Content? -- Would love to hear your thoughts on the business side of the China Ban effecting Nvida
@juanjosepariciolago27762 жыл бұрын
Can you link the setuper channel, soft and track, so you don't have to color and rena 3 tis ? Btw aweso video!
@gearboxworks Жыл бұрын
Really love how you boiled it down into a pithy slogan for each chip vendor. It really helps those like me who only have ~1/100th of your knowledge to be able to grok your video, at least somewhat. 😊
@blackmennewstyle2 жыл бұрын
Dang, he was pretty excited all along the video lol More cores is definitely the kind of stuff which turns him on lol
@electrocyper2 жыл бұрын
First, thank you STH 👍
@ServeTheHomeVideo2 жыл бұрын
Wow! You are fast.
@Ivan-pr7ku2 жыл бұрын
Intel's new smash hit: Domain specific acceleration killed the general purpose star.
@andrewnorris54152 жыл бұрын
ARM may have its eye on the Desktop x86 market. Comparing it may be about giving some street cred in that market. Like how car makers produce the ultimate version of family cars in rallies. It's all about image. ARM previously had an image of Raspberry PI in the Desktop buyer's mind. Of course, they know they are for the server market.
@m5a1stuart832 жыл бұрын
I am still waiting for Core 2 Quad and Core 2 Extreme at low prices. They prices still high here in my country. Above $120 second hand. Anyway is Xeon X3470 any good for a CPU?
@matthewc9942 жыл бұрын
Im looking forward to the cpu market being more betterer
@ServeTheHomeVideo2 жыл бұрын
Maybe one day the More Bestest too. :-)
@__--JY-Moe--__2 жыл бұрын
hell yea!🥳
@jamesvincentcarrollII2 жыл бұрын
I was doing something else and listening to this in the background 👍
@tommihommi12 жыл бұрын
Lots of exciting stuff happening these days
@christopherjackson21572 жыл бұрын
Server market analysis videos are the best ones. Whatever happened to xeon phi lol
@ServeTheHomeVideo2 жыл бұрын
Xeon Phi is basically what you are seeing as NVIDIA Grace, AMD MI300, Intel Sapphire Rapids HBM
@poofygoof2 жыл бұрын
@@ServeTheHomeVideo I suspect E-core is related, too
@LiLBitsDK2 жыл бұрын
Hey STH, what about a video about "netgear GS105Ev2 vs. Prosafe GS105 vs. SG1005p vs. GS305E" WHICH one to get, WHICH one NOT to get, WHY to get "x" and so on? They have so many 5 port models, it quickly gets confusing and well not much comparison info on their website on why one is better than the other
@ServeTheHomeVideo2 жыл бұрын
That is a cool idea.
@LiLBitsDK2 жыл бұрын
@@ServeTheHomeVideo thank you :D sitting in the dilemma now and thought "hey getting one shouldn't be hard" oh boy I was wrong, they have like 4-5 different models on the market that practically look the same some managed, some unmanged, some "smart" some "prosafe", think they even have some PoE model but not sure if the switch is just powered by another PoE switch or if the switch itself delivers power to 4 of the 5 ports, but you can't even compare them on their site to get a list of features one has that the other doesn't have and so on. So you atleast got 1 viewer for such a video :D and I hope you folks can help us "simpler" folks figure out what is up and down with all these models.
@WolfgangBrehm2 жыл бұрын
As a programmer I can confirm that more cores = more better
@beatadalhagen2 жыл бұрын
(Not serious, but still) You haven't talked about what IBM is doing. I assume they're still in the Power game?
@ServeTheHomeVideo2 жыл бұрын
I did a Power10 video some time back. Actually, we used one of IBM's last systems to target this type of market the LC921 (Power9 PCIe Gen4) in a SSD review last weekend where we used 10 different PCIe Gen4 architectures to test a drive: www.servethehome.com/solidigm-d7-p5520-7-68tb-pcie-gen4-nvme-ssd-review-on-x86-arm-and-power9/2/ IBM is not really a competitor in this market anymore. Instead, IBM is in the market of providing upgrades to those customers running Power.
@user-hj8rn5wp8z2 жыл бұрын
riskV ? I know this is outside integrations still. But any rumors for using it in next 2 years?
@ServeTheHomeVideo2 жыл бұрын
Probably not in mainstream sockets unless it is part of an accelerator tile.
@denvera1g12 жыл бұрын
Speaking on more cores=more better. Extrapolating from Buildzoid's video from Actually Hardware Overclocking on Intel Alderlake E core and P core performance scaling. If Intel had made an all E core 12900E, it would have the performance of around a Threadripper 3960x(probably more but i'm going to assume the performance doesnt scale well) All while using only about 165w instead off the 241w of the 12900K or 280w of the 3960x What i cant wait for is to see the performance of purely E core embedded server/IOT SKUs to verify these extrapolated estimations.
@LunarLaker2 жыл бұрын
I don't feel like an overclocker is the person to look to here. He only tested Cinebench (small overlap with cloud apps) and he also calls E cores ERROR cores and has insisted they're a waste of silicon.
@denvera1g12 жыл бұрын
@@LunarLaker Thats the thing, he tries to paint them in the worst light, yet his own data shows that they are FAR more efficient than the P cores, and a 40E core 12930E, if scaled perfectly, would be better than a 3960x, and use as little as 140w, but more likely 165w
@jimatperfromix27592 жыл бұрын
I like your point since it's something I've given a bit of thought to myself. Not only are the performance questions centered around P-cores versus E-cores interesting, but the whole use of P/E cores per se (by both Intel gen 12/13 as well as Apple M1/M2) raises a plethora of what, for lack of a better word, you might call philosophical questions. Sometimes I like the P/E cores idea and sometimes I hate it. A reason to like it would be when you have a mix of work that can run multithreaded but the needs of one type of thread are more intense than the other type of thread. Often the reason I hate it is because I just want a uniform core like AMD gives you. Whereas on Intel 12th gen, I ask the scheduler for 12 equally fast cores, and instead it gives me 4 cores plus 8 pieces of crap. If you wanted to take the most cynical attitude possible toward Intel, you could use the cheeseburger under the hot lights analogy. If Intel were making burgers like MacDonalds, it's like they realized they were way behind AMD (Burger King) in technology and couldn't match their sale specials of 8 Whoppers for $10. So instead, Intel (MacDonalds-like) offers you (as a $9.95 special) 4 fresh Big Macs plus 8 free cheeseburgers that have been sitting under the hot lights for five years. Yum! 5 years - that's about how old the technology is on those E cores. Intel is counting on non-technical customers who might think 4 Big Macs plus 8 stale cheeseburgs is a better deal than 8 fresh Whoppers. Yet (as you point out) for some apps that are conducive to multithreading solutions, an ostensible all-E-core chip might be a nice thing to have. If you search a bit, you might be able to find a recent podcast by one of the major Mac review channels in which they note (quite amazingly to them at the time) that you can take a brand new Apple M2 Air with 4 P cores and 4 E cores, and if you set the macOS bit to run at low power, it actually runs on the 4 E cores plus the 4 P cores only downclocked to the E-core clock rate, and for normal usage that doubles your battery life to absolutely insane numbers with no noticable reduction in humanly perceived performance (until you go export some video, that is). Here's another good one for ya - no tests to prove it yet, but I conjecture there are some model-number-versus-performance inversions within the P series of 12th gen Intel CPU chips. That is, I think some of the lower part numbers may be faster (or more horsepower actually) than some of the higher part numbers with more P cores. I'll just say, choose wisely among 1240P, 1260P, 1280P etc.
@denvera1g12 жыл бұрын
@@jimatperfromix2759 This is my whole problem with comparing Apple M1 and Windows laptops. The other part that bothers me is that they never take into account Apple's use of 5nm, while windows laptops use a mix of 14, 10, 8(which is just a hot 10nm), 7, and 7++(6nm). But more to your point, Apple has designed their OS to run on the E cores at the lowest speed possible, only ever using the P cores when the E cores are so slow, or so congested with other programs, that it would be more efficient to use the P cores. Windows on the other hand has a schedule for 1995 that struggles managing more than 4 cores, let alone many differnet kinds of cores as well as power and cache states. The Steamdeck is an oddly good example of how efficient a properly tuned OS is on X86 . I honestly expected far less performance because, on paper, based on windows performance of the equivelant to this hardware, it should have performed about as well as a 4750u with the GPU clocked at 2.1Ghz. Instead i would put it on par with a theoretical 4750u with 12 compute units instead of 8. On paper, based on windows numbers, RDNA2 should be about 30% faster than the VEGA2 found in the 4000G/5000G processors. Given they both have the same CU cound and clock speeds. But the steam deck is running at 1.6Ghz max, where as the 4000G/5000G can often run at 2.4Ghz. Based on windows numbers, the steam deck should be equivelant to the VEGAii 8 running at 2.08Ghz Yesterday, i was playing batman arkham origins a game from i think 2013 so its not like its new, but still at 800p30, the whole SOC was only using 1.4w. If i had a way to turn off the display and use an external display i'd probably have a battery life that approaches 20 hours with things like haptics and wifi ect off. Its just too bad i cant test because this thing seems to just turn the screen black when using an external display, and there is no 'off' backlight. Then if i turn it up to max, the thing is able to push 4K30(limited by my dock or the USB-C port) on Elite dangerous, using only ~16w from the wall, which probably means the SOC is using ~12w Now imagine you were not playing a game, and instead doing something highly parallelized that scales effectively perfectly with core/CU count, say, video editing, or blender, or a specialized server application. Sure it would be "efficient" with this 15w 4 core. But it would be a heck of alot more efficient if you had a lower clocked 16 core that still ran at 15w. Heck with 16 cores, you could run it at 30w and it would still probably be 2-4x as efficient.
@drifter4training2 жыл бұрын
So is it a foreseeable future to have a mobile CPU with the igpu performance @ 1080p medium settings ?
@denvera1g12 жыл бұрын
I'd like to know the efficiency of X86 with that many cores at a lower frequency. Would a 128 core EPYC running at 200w be significantly faster/more efficient than 64 core 280w EPYC on the same node?
@chengcao4182 жыл бұрын
I think that's the same idea as AMD zen4c skus with more cores, less cache, and lower frequency for each
@popcorny0072 жыл бұрын
@@chengcao418 Exactly right. As for being better/worse, it depends on the workload
@mustafasahin27062 жыл бұрын
you are the best dude
@SP-ny1fk2 жыл бұрын
Is Proxmox on Arm a thing?
@Cardroid2 жыл бұрын
Your videos are very helpful.. can you do one on RISC v CISC in server or that doesn’t even matter.. thanks.
@cambrown57772 жыл бұрын
Ampere (using ARM64) has incredibly low power draw compared to x86 based Intel and AMD
@tgsparkyoriginal Жыл бұрын
I’m looking at getting a Dell Micro 7000 series to run a Laser Cutter but I’m trying to work out the best CPU as the software I’m using works on a ingle core
@PhillioDoede2 жыл бұрын
No mention of IBM? Also no consideration of the power draw of these different platforms?
@Technovir2 жыл бұрын
I wonder if Intel would abandon ego and look at Phi again for customers who just need max cores
@ServeTheHomeVideo2 жыл бұрын
Sierra Forest mentioned in this video
@SmokeytheBeer2 жыл бұрын
[deletes rage comment about AMD]
@Wasmachineman2 жыл бұрын
Speaking of podcasts, you ever thought of going on Broken Silicon to talk about the server side of hardware with Tom of Moores Law is Dead?
@Speak_Out_and_Remove_All_Doubt2 жыл бұрын
AMD = more 'balanced', more better? Turin (Zen 5 server) from AMD is meant to double the count again going upto 256c/512t for a single socket in part because of the move to TSMC's 3nm node, plus the Xilinx acquisition will be well and truly in full flow by then so we should see a lot Xilinx influence in the chip design and accelerators present.
@ServeTheHomeVideo2 жыл бұрын
Sure, but both Intel and Arm folks will be adding cores by then too. By 2025 a 128 core per socket system will seem small.
@Speak_Out_and_Remove_All_Doubt2 жыл бұрын
@@ServeTheHomeVideo The way Intel is executing at the moment we might be lucky to have Sapphire Rapids by then :-) Whilst I have you Patrick, I would love for you to do a 'deep dive' on Submer. Their cooling solutions look really interesting.
@AdeptSoul2 жыл бұрын
If an AMD EPYC thread has similar performance to an ARM core, and AMD EPYC Bergamo will have up to 128 cores and 256 threads... The same number of cores as ARM and double the threads, more better? 🤔 Time will tell...
@LunarLaker2 жыл бұрын
worth noting Zen 4c is smaller than Zen 4, and the V1 in Graviton3 is much larger than the N1 in Altra...
@jimatperfromix27592 жыл бұрын
One needs to remember some basic facts - (pretty much up til now anyway) a core is a core is a core, and a thread is a thread is a thread (hyperthreaded or not), and a thread must run on a core, that is, a core must be allocated to a thread for the thread to run. The advantage of hyperthreading (two-way, say, for simplicity) is that it seriously speeds up the context switch from one hyperthread to the other. It's as if (an oversimplification) the OS kernel allocates 2 threads to be brother/sister hyperthreads on some given core, and then those two hyperthreads get scheduled round-robin on that core. But hyperthreading only minimizes what is always an overhead, namely context switching between sibling hyperthreads. You can think of the optimal context-switch overhead reduction as simply replacing a single pointer word in the core to point to the thread-context register file of one hyperthread or its sister. Fast switch at the expense of one level of indirection. (As far as I know) up til now, no hyperthreading architecture actually runs both threads at the same time. The latter is possible in some future architecture. After all, we have instruction lookahead with multi-issue instructions within a single thread; What's to prevent one from having multi-issue across two hyperthreads? However, there is no free lunch. One can sanely argue that if you're gonna do that, your chip real estate would be better spent just doubling the number of cores (and while you're at it, why not quadruple the multithreading level too). So as far as your question, no, same core count but double hyperthread count does not, in general (all other things equal) give you a 2X boost. It might give you a boost of 1.y where y is some small fraction probably not greater than 0.1. As pointed out, the true decider of which chip is better is both a function of the application (or application mix in the case of timesharing use or virtual sharing of cloud resources) as well as the "fit" of the instruction-set architecture of a given CPU to the mix of instuction needs of the application. For a given app, a better fit might be provided by an Epyc chip or a RISC chip of some sort, depending on chip architecture.
@LunarLaker2 жыл бұрын
@@jimatperfromix2759 on the "both threads at once" point, I think VISC was aimed at doing that (they've since entered the shadow realm of course.) They didn't necessarily mean the performance was doubled with SMT for twice the threads. other than that, 🧱 📖
@cambrown57772 жыл бұрын
being a cloud service provider is also about cost of vCPUs to your customers. Ampere provides much more efficient chips, thus lowering costs for general cloud tasks
@bridgetrobertson71342 жыл бұрын
It's definitely a piece.
@ozdemirsalik Жыл бұрын
More collab with Jeff!
@MrZiemwit2 жыл бұрын
what is better for cheap small home virtual lab? A8 PRO-8600B 28nm 4 core /ddr3 i3-6100t 14nm 2core +2ht /ddr4 ? what has beter low power states managment for cheaper usage what is easier for virtualisation does amd bad for virt
@jyvben15202 жыл бұрын
intro "this is Patrick", android ? where is the real Patrick, does he even exist ?
@cranebird1467 Жыл бұрын
most softwares aren't compatible with Arm CPU
@Speak_Out_and_Remove_All_Doubt2 жыл бұрын
In all seriousness I am very worried for Intel, even if you believe Intel's roadmaps things aren't looking good but when you take into account their very poor execution record of late things are looking very worrying. Sapphire Rapids was originally due late 2021 to early 2022, it's now 'hoped' to be out by March 2023. This is a massive head start AMD will have with Genoa and Genoa X, plus these have 96 cores Vs Sapphire Rapids 56-60 cores, not to mention much lower power usage. AMD also has Bergamo coming not long after this which is 128-160 Zen 4c cores. But the most worrying thing is AMD looks on track to release Turin (Zen 5) in 2024, by all accounts this will go up to 256 cores per socket and if Intel executes their roadmap well they might have Emerald Rapids with just 64 cores! Intel is still claiming Granite Rapids and Sierra Forest are 'due 2024' but customers have lost all confidence in anything Intel says so I don't believe many clients who need to plan years into the future will be putting their eggs into Intel's basket! I just really hope we aren't looking at another IBM.
@mikebruzzone95702 жыл бұрын
Ampere high core count r not effective on cost : price in relation the lower core count Q Ampere selling the high core count would have put themselves out of business. No, more core count is not necessarily better as in cost effective for the design producer Patrick think about it. mb
@cambrown57772 жыл бұрын
high core count IS effective on cost, especially when you're running them on 40% less power than the x86 competition
@andrec72462 жыл бұрын
Is this arm processor, Chinese?
@MagnumCarta2 жыл бұрын
Ampere is an American company. They are headquartered in Santa Clara, California.
@catchnkill Жыл бұрын
There is similar Chinese ARM server chip but they may not be sold outside China. Alibaba's chip designing department has designed an ARM v9 server chip Yitan 710. It runs neck and neck with Amphere Altra Max. The chips are being used in Aliyun's cloud servers.
@matthewgarland5944 Жыл бұрын
seeing a lot of multi astounding ... bill says nighty right now but can imagine a lot of cores working on mother nature or the spectral scene workstation ... i liked more the fairly older tech you described accelerator connections i was hoping we would call up the pete for advise at some point .. jump in the lobby for grid iron ..we at home are still dorita.... bill said nighty makes for a good modem ... like hes not scrambling a lot of eggs to say phi could work in other dimensions kinda like i get out of it as off loading blowfishy dark web overheads ...the oldest cleek is if each core has a bank of ram devastation could result AI AI AI ... were a little off here for the consumer for instance phi stable wasnt pointed correctly ... what about the general store of proc some fast some background hes right there too again more ram is more ram centralizing ram is a parody of Chinese checkers but maybe we should use it to interact otherwise im having diificulty with anaconda trunk spaces running between buildings in real time... finite analysis
@alexfolsom39102 жыл бұрын
Love this guy.
@movax20h2 жыл бұрын
nginx is extremely bad example to be talking about here. nginx on just few cores can saturate 100Gbps. And nobody, literaly nobody is going to be using nginx on 128 cores. For performance and reliability you will run instead haproxy, nginx, or something similar, on 128 machines instead. A way better examples would be databases (Postregresql, Mysql), data analytics pipelines (i.e. Spark, Flink), application front ends (talking to databases and storage, possibly doing a bit of internal caching), queueing and pubsub systems (Kafka, RabbitMQ), some basic data ingestion transformations (i.e. resizing images from user uploads). They are also great for integer heavy batch workloads, like compilations, container image building, testing infrastructure, CI/CD workflows, and hosting various general productivity systems (i.e. gitlab, teamcity, airflow). On bare metal they can also be useful for storage systems (i.e. Ceph), but storage and memory are primary cost there, so Arm doesn't actually have big advantage in terms of saving any total cost. They are also good for generally cheap density. If you can fit 128 small customers (each just hosting a small website, DB and few utilities, some maybe a dns or email server, using 1 core each shared between many of them), that is more cost effective than going with smaller and more expensive chip from Intel or AMD, which would not be fully utilized anyway, and use more power (i.e. when you have many small customers, you do not care about core-to-core latency for example, or ultimate FP performance). There is nothing stopping Intel or AMD doing similar with just cramming a lot of eCores into their chips. And AMD is already designing 96 core CPU for just that, chiplets, and each core will most likely be faster than Altra cores in many dimensions.
@ServeTheHomeVideo2 жыл бұрын
The Altra Max is not really designed to be 128x cores running nginx. It is designed to be 128 cores running 16x 8 vCPU VMs running nginx as an example. AMD will be at 128C with Bergamo and Intel will be (much) higher when Sierra Forest arrives with E-cores. By then this will be an entire industry segment.
@movax20h2 жыл бұрын
@@ServeTheHomeVideo Then I do not see really a point of mentioning "per-core" licensing of some software. It is a real niche of a niche.
@0eroOverride Жыл бұрын
Sighhhhh Just before 6:05.. you glossed over literally everything and went into no details about how lacklustre the e cores are. No ai/machine learning or AES extensions. No mmx.. no vmmi
@spuchoa2 жыл бұрын
According to Igor'sLab Sapphire Rapids will launch in 2023 Q1, bad news to Intel and us customers.
@KingTheRat2 жыл бұрын
You should change your channel name to ServeTheDataCenter hahahaha
@ServeTheHomeVideo2 жыл бұрын
The STH main site (non YT) is ~10x as big as the YT channel, so it has to match. Home is the /home/ directory in linux.
@mikebruzzone95702 жыл бұрын
Intel lead AMX 512? No, AMD lead Intel AVX 512 at AMD 128 bit floating point and NO, both had 256 matrix math and Intel did not lead they took everything they have from AMD developments enslaved as a design house. mb
@mikebruzzone95702 жыл бұрын
Listening in the background [?] who r u kidding. mb
@Decki7772 жыл бұрын
More cores more better not in gaming
@LunarLaker2 жыл бұрын
sir, this is a Wendy's
@TonyLee_windsurf Жыл бұрын
More Cores Competition, More Better
@jolness12 жыл бұрын
Y U NO AMD YET! 😡
@jannegrey5932 жыл бұрын
I stopped watching something else when I saw there was video from you. I will post again, once I've watched it.
@ServeTheHomeVideo2 жыл бұрын
Ha! Thank you.
@__--JY-Moe--__2 жыл бұрын
Im looking forward to the cpu market being more betterer
@jannegrey5932 жыл бұрын
@@__--JY-Moe--__ One hopes that will be a case. I still haven't even managed to start building my 5950X system.
@octapc2 жыл бұрын
Do you think you could talk any faster?
@ServeTheHomeVideo2 жыл бұрын
I can try to film after I have coffee in the morning if that will help speed things up
@ama-tu-an-ki2 жыл бұрын
Intel has become the IBM of 2000s: they are trying to build a total stranglehold ecosystem, where you buy INtel everything and you can't mix and match. Why? To keep out the competitors, to build a moat and NOT having to compete head-to-head on all individual compute parts (CPU, GPU, AI, etc). Intel is losing on all individual fronts, so all they have left is vertical deep integration and trying to ram everything down their corporate clients throats. Usually, on the long time scale, this will fail. Ask IBM.
@everyhandletaken2 жыл бұрын
excellent analogy
@grtitann74252 жыл бұрын
Don't forget to add Intel's favorite tactic, bribe OEMs to only carry their crappy CPUs. Just ask Dell...
@chasonsnotes Жыл бұрын
Talking too fast
@__--JY-Moe--__2 жыл бұрын
I hope ARM knocks it home! but don't forget? isn't optical transfer going 2 make a comeback?? feeling old? I am!! U know the world U live in, is sagging. when Japan doesn't have any ''new'' optical transfer devices? right!! just keep in mind, I came here 2 check U'r reflexes! ha..ha..
@__--JY-Moe--__2 жыл бұрын
will this cut down on : U'v lost that loving feeling?😁