Ryzen vs i7 (Mainstream); Threadripper vs i9 (HEDT); X299 vs X399/TRX40; Xeon vs Epyc

More likely it simply means nobody has bothered much with running this benchmark, since there are plenty of processors that would beat that (assuming it's heavily MT).

"What they consider state-of-the-art at roughly the same time point" is, has always been, and will always be a moving target.

 

I agree with this sentiment. But, it should be noted Intel is rumored to have gutted a couple features from Willow cover for the back port. Those will not essentially necessarily change the IPC increase overall, though (but it may, as some may effect the floating point scores or cache system, but that is all rumor right now).

If we assume IPC doesn't change and that the changes for backporting are trivial, then that is 18% for sunny cove and an additional 5-10% for Willow cove (we'll use 7% for industry average, since the rumors come to about that). That is a 25% IPC increase over skylake parts on 14nm, while possibly retaining the 5GHz speed. That is a good chip.

AMD currently has an approximate IPC over skylake of 11% (some have 15%, some 17% over Zen+, which was around 4% IPC below Skylake, but with a massive frequency deficit) with Zen 2, but still a large frequency deficit. Zen 3 is rumored for 10-12% integer and 17% average IPC.

So if Rocket was going up against Zen 3 with an approximate 28% IPC increase, rocket would win on frequency, even if we give AMD the benefit of the doubt on 100MHz more all core boost.

But the rumor is Rocket is a 2021 CPU. So, even though it might start out competing against Zen 3, by the end of the year, it is going against a 5nm Zen 4 with another IPC increase. Assuming similar mid-4GHz frequency to Zen 3, give or take a couple hundred MHz, they would need to keep with the over 12% IPC increase to stay on top with an outright win. Considering their recent track record, it is possible. But this fight will be a close one.

I don't care if you prefer Intel or AMD, that is a fight worth watching because it will be so close. If AMD gets closer to 17% IPC, Intel has an issue (still close, but they will have an "L" on their hands).

 

LoL;

 

Here's an early view of the Intel 10900x:

Dead On Arrival: Intel i9-10900X CPU Review & Benchmarks vs. 3900X, 3950X, 9900K
Dec 15, 2019
Gamers Nexus
Intel's Core i9-10900X 10-core CPU lands on the HEDT X299 platform and competes in price with the R9 3900X, R9 3950X, and Intel i9-9900K CPUs. The Intel Core i9-10900X is Intel's 10-core, 20-thread CPU that was supposed to be a big step-up from the Intel i9-9900K. It's close in frequency -- sort of -- and higher in core count, but it's also entering a field heavily dominated by the AMD R9 3900X and R9 3950X CPUs. All three of these, of course, have been largely paper launches, but both AMD and Intel are working toward getting shelves more consistently stocked. In the meantime, the 10900X's launch is nothing short of a bloodbath, and it's not in favor of the new Intel 10-core CPU. PCI-e lane count is perhaps the only defensible position, but even that is challenged by other chips on the market.


A couple of Intel / AMD Fanboy's discuss the state of Intel / AMD, and the Hardware Unboxed World.

Hardware Unboxed | RX 5500XT, Zen 3, Ray Tracing | Broken Silicon 26
Dec 13, 2019
Moore's Law Is Dead
Steve Walton joins Tom for a long form discussion regarding 2019’s hardware releases, upcoming products, & HWUB’s origins. [Steve’s audio capture crashed. Nothing lost, but audio changes at ~48 min mark]
1) 0:49 Steve Hides his RTX Money
2) 4:50 Can Intel even be recommended anymore?
3) 6:20 Threadripper and Zen 3
4) 15:15 Hardware Unboxed’s Origins
5) 27:00 How much work it actually takes to make good content…
6) 34:00 Why is Tim the Monitor guy?
7) 37:10 What were your expectations of Zen in 2016?
8) 45:00 Are PC prices “back to normal”?
9) 52:30 Comparing this era to the “28nm Era.”
10) 57:55 Turing, and the value of Ray Tracing.
11) 1:08:55 RX 5500XT & RX 5600XT
12) 1:17:20 FCLK Overclocking Capabilities, IPC Comparisons
13) 1:23:35 Should gamers turn off SMT on Ryzen 3000?
14) 1:29:35 What is Steve thankful for from 2019?
15) 1:43:00 What is a big performance increase?


Ditching Intel for Ryzen, Storage Server Upgrade With Huge Performance Gains
Dec 16, 2019
Hardware Unboxed


So much *cool* new stuff to cover...

Water Cooling is DEAD. Meet the THERMOSIPHON!
Dec 15, 2019
Linus Tech Tips
9.75M subscribers
The startup IceGiant is trying to change the computer cooling landscape – and according to our testing they’re off to a good start.
Pre-Order the ProSiphon Elite: https://www.icegiantcooling.com/

 

Will disparate Intel Pie in the Sky concepts allow their currently inferior architecture CPU's get the jump on AMD / ARM innovations?

Doubtful, but Coreteks delivers an interesting video that brings some range to the connectivity and communications standards that we use and how they can be re-oriented and re-conceived to deliver useful combinations.

Are we moving to MODULAR PCs?
Dec 20, 2019
Coreteks

 

Intel Tiger Lake-U Leak Reveals 15W/28W CPU With Up To 4.3GHz Boost Clock hothardware.com | Dec 23, 2019
Intel released two families of 10th generation mobile processors over the summer: Comet Lake and Ice Lake. Of the two, Ice Lake garnered the most attention as these processors are based on Intel's long-delayed 10nm+ process tech and feature an all-new graphics core. The chip giant is on track to release another round of 10nm mobile processors in 2020 under the codename Tiger Lake...

For comparison, the fastest Ice Lake SKU, the Intel Core i7-1065G7 has a single core turbo of 3.9 GHz and an all-core boost of 3.5 GHz, so Tiger Lake hitting 4.0 GHz or more would be sizable increase. Since it concerns an engineering example, final clock speeds could be even higher.

 

From the earlier rumor of 3.6 and 3.8, I've seen that this 4.0 all core boost/4.3 single core boost is allegedly an ES2 sample. It seems they did get something figured out. That is a 10%+ boost in frequency, which is awesome and will require me to reevaluate my last estimate (although tiger lake is currently only planned for mobile while rocket lake with the same core design is planned for desktop). My only question is if they can go further with frequency for 10nm++.

TSMC barely figured out last minute how to get the frequency to low to mid 4GHz.

I wonder if they are still using cobalt or not, as the cobalt being brittle and breaking was part of the issue. If they are and they solved that problem, they will be the first fab to have solved it. Using cobalt or similar to deal with electron migration is something all theoretically have to figure out at Intel 5nm/TSMC 3nm. That would put them years ahead on that front, as others focused on EUV (Intel set that to the side waiting for partner's tech to mature).

 

Well, Micro Center in Cambridge suddenly had a stock of 3900X's today when I went there, and I bought one (have been thinking about it). Early tests, which don't appear to be exercising the full power of the chip, suggest about 50% faster than the 2700X it's replacing. But those are some single shot ffmpeg runs that don't appear to be able to take advantage of the whole chip. I also spotted that the BIOS reset the memory speed to the default 2133; setting it back to 3000 netted about 7% improvement.

It will be interesting to see what my big overnight batch jobs do. I'm particularly interested in seeing how the Gutenprint test run does (most of it's embarrassingly parallel). It typically takes about 73 minutes on the 2700X. I'm hoping it will be closer to 40~45 minutes on the big boy.

Incidentally, at least for what I've done thus far the stock cooler is considerably quieter (subjectively) than it was on the 2700X. But again, that likely wasn't full load.

 

It took 46:10, which works out to about 57% improvement, or about 5% per-core.

I noted all-core boost to around 4.0~4.1 GHz (starting around 4.1, gradually falling to 4.01'ish), and one core at one point showed 4.6 GHz (by means of running cpufreq).

 

Welcome to HEDT. Personally I consider above 6 cores entry level of that. AM4 now at 16 cores is into the mid range of it, actually not too long ago this was the top end of HEDT. Computationally for the most part the 3900x meets the 1950x and the 3950x bests it.

 

I don't entirely agree with your conclusion, but I agree with the questions you're asking.

So the question is, where does desktop end and workstation begin? It's hard for me to see how a true desktop system is going to use 72 PCIe gen4 lanes. That's just a stupendous amount of I/O capacity, potentially 144 GB/sec. That's more than just storage in most cases; it would be multiple high end graphics adapters, video acquisition/distribution, things that just sound a lot more like workstation than desktop.

Intel seems to be classifying parts as server (Xeon), workstation (Xeon-W), HEDT (X), high end mainstream desktop (K), and normal desktop (non-K). With that kind of classification, Threadrippers (particularly Xen 2) are just too powerful -- and more importantly, have vast I/O and memory capacity, and support functionality such as ECC RAM -- to really be desktop; they are full-fledged workstation parts. They have a lot more in common with Xeon-W than with, say, the 10980X, which only has 48 gen3 PCIe lanes and no ECC support.

I agree with you that the 3950X and 3900X are really high end desktop. They are more powerful than most people need for desktop purposes, and have I/O capability very close to the Intel X series parts (in terms of MT/sec rather than lanes) and in a lot of cases outperform those parts in computation power. But they don't have any distinction besides number of CPU cores from the entry level and midrange CPU parts. I think looking at price that I'd have to call the 8 core parts high end mainstream, which may be a distinction without a difference.

 

Other than the 3900X, my trip to Micro Center wound up being wasted and I have a bunch of things to return. Silly me for not doing more research:

• Radeon RX5500XT isn't yet supported well enough under Linux to be of use. It resolves the problem I'm having with Radeon cards not link training successfully when connected through a DisplayPort KVM switch, but it's just not ready for prime time.
• Asus RT-AC66U just doesn't have the capabilities I need (DNS server, custom routing rules, VLAN). I'm just going to have to get a DD-WRT router for that purpose.
• CoolerMaster Hyper 212 cooler, in addition to being a pain to install (I'll have to lift the motherboard) isn't much quieter than the Wraith Prism.

So foo on me.

 

I was tempted to go 3960x route but have decided for now against it. We are talking over $2,000 now for plenty of performance gain but not a required one. My 1950x is more than ample for anything I have to throw at it. 99% of my time is just simple stuff handled by the 2500u as well. Maybe one day down the line, we shall see.

 

LTT builds a new Epyc Storage Server to replace their Intel Server, as always - cringe worthy entertainment. 28GB/sec Read - 22GB/sec Write - as individually loaded drives, they've got some debugging to get through to make Linux / Windows work as pooled or RAID of some sort. Cutting edge, it's sharp stuff.

Where Intel is in REAL Trouble...
Dec 25, 2019
Our editing server is in dire need of an upgrade, and let's just say that this one is going to be EPYC...

 

Although I liked this one, did you see the one where he used a 32 core Epyc chip to build his home systems? He used fiber optics to run the USB and HDMI lines while doing VMs for reach different system, similar to his 7 editors one system rig.

Now, if AMD does an OCable 64 core with the memory channels and lanes available (like the upcoming workstation boards, or possibly the TRX80 or if they make a class of boards like that of the W3175 for the 64 core OCable chips so that the VRM stay cool, but with all the lanes available), that may be a preferred way over doing client pass through machines that screen mirror (like developer boards being used) in regards to latency.

The fact we are now talking about home centralized servers, a single system, being able to host all other systems that were in the home, we are looking at changing what personal computing is. This is thanks to AMD chasing the core counts for the commercial space while bringing it to the consumer. Considering it is replacing so many other CPUs (you still have dedicated GPUs per VM, so cannot discount the electrical costs of those), you should be able to do a decent OC on a 64 core chip and still be at a lower power consumption than all the systems being replaced.

I don't know if you posted that one, but it was in the past week or so of videos from LTT. Worth a peak.

Edit: here's the video-

 

400w, this be nice with 24 cores threadripper in the future. with zen4 24 cores or intel 18 cores.

 

Agreed, but I'll probably keep water for the main system for the foreseeable future. Already have 3x480mm rads. This cannot do that. But it does look like a great product, one I'd consider over AIOs depending on the finished product's performance.

 

Put 3 similar Delta fans on the water cooler and you'll see temp will drop with the tested AIO cooler as well Saved by Delta.

 

A personal cloud! Wow.

The _last_ thing I would want to do with that is overclock the CPU, though. I'd be a lot more inclined to run it at a lower TDP for long-term reliability, in addition to putting it on a big UPS, using ECC RAM, a good RAID setup, and a commercial-grade hypervisor to manage all those VMs.

 

I agree with what you said except for the overclock. You would be thermally limited and VRM limited, if not electrically limited if you don't run it on a proper circuit, possibly to itself with a 25-30A breaker dedicated, before you ever had a chance of blowing up that chip. Bearded hardware showed overclocking the 3970X on water could shut down on the VRM on the TRX40 boards before heat on the CPU or voltage were in dangerous places.

This is why I mentioned a 32 phase VRM, made so that a custom waterblock can cool them, that way you could approach a high 3GHz to 4GHz all core OC with 64 cores on package. Granted, the TDP comes from base, not boost, which is why the 7H12 increased the base, not the boost really, over the prior top end 64-core.

Basically, what I'm getting at is getting the boost speed high enough that it is a close to 4GHz machine, which sense some are meant for gaming, this would help with frame rate. As IPC continues to increase in the next couple generations, along with new games and game engines better able to scale with cores and threads, a need for higher frequency may be less relevant.

But this is why I brought that up.

Edit: reason thinking the dedicated line and breaker is you also will have 4+ graphics cards in there, so you can wind up having a lot of pull if all were under sustained load, which isn't likely depending on your household. If overclocking with a 64 core plus multiple graphics cards, that is a lot of power draw. Add anything else, like a large scale NAS of 30-96 drives, and you are looking at needing to take a lot into consideration. Add on network switch(es), router, dedicated servers aside from the one for the home cloud, and you are now talking about a good power draw. One cabinet and you would be great. And all sound is far away from regular rooms in the house.

 

Definitely agree about the mains supply.

If you're after frequency, though, might you do better with a smaller chip? How many VMs do you want hanging off this, anyway? Is it going to be more than 4-8?

 

I'm thinking 7. That covers bedrooms, living room, office, and game room. I've also considered 2x32-core systems, but once again, for the server chips it tops around 3.4GHz boost. The TR would possibly be able to handle that, but...

 

your chiller setup will beat this anyday. just that i dont wish to deal with any possibility of spill/leaks and this seems to be the best option.

 

That's the benefit, but even Linus says it's unknown whether the 1/3 re-sized final product will be as efficient or as effective after the design changes.

Next-level Cooling has Arrived for Gamers and Creators
Introducing the IceGiant ProSiphon Elite
Industrial-grade cooling for your CPU
Engineered in Austin, Texas
Available Spring 2020
https://www.icegiantcooling.com/

Let's see how it pans out over time...I signed up for updates (bottom of that link page)

 

i honestly dont mind the thickness and imo higher capacity is a good thing that can give less audible fans. i wish they can make it shorter with smaller fans, like 130mm max height which seems to be too much to ask.

anyway this area gives a big headache for me cause it seems only way to TR is either go AIO, or wait for super efficient node, which would kill my joy in overclocking of having high frequency. but with this new cooler, 400w at that temp is amazing and i think a 24c TR with new node will probably only use ~300w that will give more room for OC, or less audible fan whichever i choose and i like to have choices.

 

If you are setting up a TR then most likely the 24c is the way to go. That unless you have the need of the 32c over the 24c.

 

I've said it before, but I think all the emphasis on overclocking is misplaced. Overclocking simply means running the processor at a higher frequency than the manufacturer states. If the manufacturer rates the chip for a fast enough clock rate, why does it matter whether it's overclocked or not? Would you prefer a chip rated 4 GHz that you OC to 4.5, or a chip that starts out at 4.5? Because what AMD is doing is essentially dynamically overclocking the chip on the fly to whatever frequency can be supported by the thermals and power delivery.

I also think the emphasis on specific clock frequency, as opposed to instructions/second or instructions/second/core, is misplaced, but we'll leave that go for now.

 

Dude, Epyc is CAPPED. It doesn't dynamically go over 3.45GHz on ANY CPU. So unless using the TR series, you are missing a huge point, which part of the appeal of Epyc is the lanes and memory channels. Until we know if the workstation boards overclock, you would need to wait for TR80, and that isn't guaranteed the same support.

Further, Epyc is capped not be heat but by TDP envelope. By capping it to 250-285W, you slow it way down. If I'm pulling 325W+ on my 16-core doing 4.2, then creating a platform that can handle 1KW is not out of the question, so long as heat is taken care of.

So you ignoring limitations of the server chips, instead saying it dynamically clocks, then using figures that NO EPYC CPU IS DESIGNED FOR, could also be said to be misplaced or misleading.

 

I don't see what that has to do with what I said, namely that at most it's actual clock speed, not overclocking, that matters. In reality, I think IPS or IPS/core is more important (and even that's complicated, with different instruction set extensions, instructions aren't all the same number of cycles).

 

When you are using the chip for replacing gaming PCs around the house, single threaded performance matters. Because IPC*frequency=IPS, having low frequency CAN AND WILL effect gaming performance. Now if playing 4K, the impact may be mitigated, but most still game at 1080p or 1440p for higher frame rates.

So, going from 3.4 to 4GHz is an almost 20% jump in performance. Going to 4.3GHz, closer to what is seen on mainstream Ryzen, is 26.5% more performance. In other words, closer to what Zen 4 would have at these lower frequencies (Zen 4 may even be ahead of that figure). As such, even with Zen 2 being as good as it is, the CPU can act as a bottleneck for gaming purposes, and having all systems running on that one CPU, that matters.

If we were purely talking compute workloads, encode and decode, etc., then it doesn't matter as much. Hell, the 64 core Epyc was shown to do near 8K in real-time on the frame rates it can achieve for h.265. That's incredible! But we are NOT talking about those workloads, now are we? Simple browsing tasks, office, etc., it is more than enough. But, because you are replacing all other systems, gaming does matter, and a 26.5% frequency reduction negates the IPC gain seen, if looking at moving individual gaming machines to a centralized system.

As said, in 2 years, that won't matter. That alone is a crazy thought. And as software better utilizes cores and is less serialized, it also will make it matter less. But, for the moment, it matters.

Do you yet understand what I'm saying, or do I need to go into more depth?

Edit: mainstream chips are 26.5% faster, but it is 21% slower than mainstream chips. Not sure I used the correct ratio in all instances, so appending my comment here.

 

And my point is that if IPC is high enough that that will compensate for clock rate. Different architectures will get different IPC; if the rumors are correct that Zen3 will get a big IPC boost over Zen2 (which admittedly is a mighty big if), you might not need such a high clock rate. 4 GHz on one architecture is not the same as 4 GHz on another.

And even if we assume that GHz = GHz, overclocking is not the issue. Does it matter if you get (say) 4.5 GHz from a chip rated for 4.5 GHz or from one rated at 4 GHz that you overclock to 4.5 GHz?

 

Stop trying to obfuscate. The estimate on Zen 3 is 17% IPC. That does not equal the mainstream being 26% faster. Zen 2 is the first time game performance between AMD and Intel became negligible, even though Intel still has the crown. And no, I would not recommend an Intel platform on this because of inadequate PCIe and power draw for using a dual CPU board.

Frequency is a measure of cycles per second. So, when you multiple the instructions per cycle by the cycles per second, you get the instructions per second because the cycles cancel out.

Because of this, to get the same performance, any reduction in the cycles per second (frequency) must be made up for in instructions per cycle to get the same performance. Add to this that due to APIs, a single thread takes the bulwark of instructions for gaming, instead of it being spread out more, along with crap schedulers involved, and you are looking at not being there yet.

Maybe in 2021. Or having it able to reach 3.7-3.8GHz zen 3 would be close to equivalent for that purpose, as you slapped 10% more frequency on, but that is next year's line.

So I get what you are saying. I do. But math is a fickle mistress and happens to love me today. Math is on my side here.

Further, you are confusing my arguments with those of overclockers that just want headroom to feel special. Why bring that up? We are talking enterprise hardware and ability to get comparable performance with a single home server to desktop. That means comparing it to desktop performance is apt. As explained, due to frequency and IPC calculations, the current Zen 2 Epycs are running too slow without being overclocked, the server boards don't have sufficient VRM to overclock, and to my knowledge, only an all core overclock is currently possible with Zen 2 Epyc anyways.

This is why I hope the TRX80 is a robust overclocking monster with the same accouterment of features from EPYC but with overclocking fully supported. That allows servers to remain tailored for their purpose, while also giving home enthusiasts a platform that can meet these measurements.

 

We're talking about two different things. My original response was to this:

My point, which I stand by, is that it does not matter whether the performance -- single thread or otherwise -- comes from overclocking, stock frequency, or improved IPC. There's a lot of emphasis on overclocking here that I consider to be misplaced. AMD could easily make their chips more "overclockable" by dialing back the stock frequency and having less aggressive boost algorithms, but that would only reduce performance for most people.

I agree that it's not likely that core for core a 3.7~3.8 GHz Zen3 would generally match even a 9900KS -- that would require something like a 30% IPC increase. But that's not the comment I was originally responding to that sparked this.

But if you're really building a gaming cloud-in-a-box, you need more than single thread performance anyway. Sure, if each VM were isolated on its own chiplet or at least CCX there would be less thermal or power cross-interference, but there would likely be some, and if your gaming mix can take advantage of even 4 threads, you're going to find yourself looking at something rather close to all-core performance. Anyway, one of these days the game programmers are going to have to learn how to write proper MT code, because that's where the big improvements are going to come from.

 

You are missing my point, and annoyingly so. Literally, I am saying at 3.45GHz, the top frequency an Epyc CPU can reach on Zen 2, you will bottleneck the GPU you are passing through to the VM. I'm not saying it has to be 9900K performance, but at 4.3GHz, AMD 3000 CPUs are now beating the 6700k and 7700K in gaming. That is quad core performance.

Now, to achieve similar PERFORMANCE, since IPC is FIXED, and because AMD does not utilize their boost algorithm to allow two, 4, and 8 cores to boost into the 4GHz range when thermals allow and the CPU as a whole is NOT under load on Epyc CPUs, and noting that household servers will NOT be fully loaded at the same time, then increasing frequency is the ONLY way to get rid of the bottleneck.

This isn't about their boost algorithm, it isn't about if frequency is left on the table or not, etc. Server chips are built to operate at a SPECIFIC TDP. They are not meant to overclock (even though you can), nor are they often of need to have the highest single threaded performance.

BUT, we are not talking about using those chips in their ordinary use case for which they are designed. We are talking about a specific use case that WAS NOT ENVISIONED when the chips were designed. That means that we need to modify the behavior of the chip to fit OUR NEEDS. And that means increasing frequency in this instance. EVEN LINUS MENTIONED IN PASSING OVERCLOCKING THE EPYC CPU! He didn't cover it, but sure did mention it, and that means a frequency overclock.

Moreover, people WILL use 8-cores per machine. Even Linus did that, except one was 6 cores to leave 2 cores for the host.

And it is fine to say games need to become more MT. I've screamed in this forum about software optimizations FOR YEARS, and even made that point above. But until we have larger DX12 and Vulkan adoption, and for occasions when people STILL WANT TO PLAY THEIR OLD GAME LIBRARY, they need the ST performance, and to get that, you need frequency.

Now, because EPYC doesn't boost in the same way Ryzen does and just limit it by temp limits, which would allow for a server chip of this type to not need overclocking, you have to include the overclocking in the discussion. I even covered this previously.

Because of that, having a board designed with 32-phases, similar to the Intel $1000-2000 boards, but made as TRX80 with all the PCIe lanes of Epyc and with 8-channel memory, that would allow for it to boost quite high, so long as you had cooling. I am assuming AMD is going to have boost similar to Ryzen on their 64-core TR. So, this is why I'm saying what I would like to see that platform be/become.

Instead, all you see it as is complaining about overclockability instead of going beyond that to see what features are being requested and WHY they are being requested.

I even pointed out that a 3.7-3.8GHz speed is acceptable for Zen 3 based CPUs, and 3.4 will likely be acceptable for Zen 4 CPUs. Instead of acknowledging why I said those per generation, basing it off of IPS, you immediately complain about people focusing on overclocking without examining WHY!

That is the problem I have with you. I see your point, but you are ignoring mine COMPLETELY, even when I say how your points are already incorporated into my analysis.

Instead, try asking questions, as it seems I've given this more thought than you have. After finding out the WHY and the FACTORS that I weighed into coming to the conclusions I did on frequency and overclocking, then you can make better responses.

You can even go back and re-read my posts. Also, I'm and enthusiast that does overclocking for a hobby. I'm the first person that was telling people that boost, although killing my hobby, is the way things are going. I've given a fair amount of credit to AMD's boost algorithm. But that ignores the context I am discussing now, which is EPYC SERVER CPUs. They function differently.

Hell, I'm one of the people able to run my 1950X at [email protected]. I have a cherry chip. It made the 2950X not worth it to me and the higher core count chips, due to architecture and scheduler optimization issues, are not worth it. So when I am talking about using a chip for a non-stated purpose that is outside the norms, there might be a reason that you need to change parameters to make it useful for one's needs.

Is this getting through to you yet?

 

The Hardware Unboxed guys Q&A, some interesting subscriber questions and answers:

December Q&A [Part 3] Would Cheap Intel 10th-Gen CPUs Be Worth Buying vs Ryzen?
Jan 3, 2020
Hardware Unboxed
02:31 Do you think Intel has anything left to offer?
04:39 Do you think you might switch Intel cpu with AMD for gpu testing?
06:49 Do you feel Intel can price their cpu's to gain recommendations in the desktop user space equal to 3600?
16:29 Do you think in the current time space that Intel having it's own fabs is a blessing or curse?
16:29 Do you think Intel's dive into GPU and memory market is a sign of wanting to better diversify their portfolio?
18:14 How much CPU threads and memory is enough in 2020 for playing back videos?
19:42 Do you think we can expect AMD to finally update their encoder and get better with their drivers?
26:19 Should I wait for new GPUs or upgrade now (from 1060 6gb to 5700X)?

Spoiler: Whole Index:

Leo 3 hours ago (edited)
00:36 Front or top mounted radiator? Which has the lower temps?
02:31 Do you think Intel has anything left to offer?
04:39 Do you think you might switch Intel cpu with AMD for gpu testing?
06:49 Do you feel Intel can price their cpu's to gain recommendations in the desktop userspace equal to 3600?
08:38 Why is it that Steve is already rocking 20+ different AIB models of 5800xt while Tim is using a humble reference model?
09:51 Does every new gpu launch bring this much criticism from the respective brands fanboys?
13:10 Will we ever see a game that use DLSS 2X, the one that is supposed to improve images at native res.?
16:00 With the end of the Skywalker franchise would you guys be a Jedi or Sith?
16:29 Do you think in the current time space that Intel having it's own fabs is a blessing or curse?
16:29 Do you think Intel's dive into GPU and memory market is a sign of wanting to better diversify their portfolio?
18:14 How much CPU threads and memory is enough in 2020 for playing back videos?
19:42 Do you think we can expect AMD to finally update their encoder and get better with their drivers?
21:25 Tim, are there any curved color accurate monitors? Or are they all gamer focused?
23:42 Are there any free / affordable monitor display calibration?
24:08 How much of an impact on SSD (M.2) thermals does a thick layer of insulating plastic have?
26:19 Should I wait for new GPUs or upgrade now (from 1060 6gb to 5700X)?
28:39 Your Christmas and NY plans?

 

time to gave up my crappy years old adobe premier. start tapping into handbrake and x265 looking good. amd cpu looking more and more promising. workload change with hardware and vice versa.

 

Here's a user review of their experience OC'ing their 3950x vs their old Intel system, 2nd video with OC first:

Ryzen 9 3950X Max System Overclock (Memory, Fabric, CPU, GPU)
Jan 2, 2020
Nick Muir
Continuation of my 3950X real world video


It's here! RYZEN 3950X - How does it perform in the real world vs i7-4770k?
You be the judge!
Premiered Dec 21, 2019
Nick Muir
New RYZEN 3950X vs old i7-4770k Skip ahead links in the first comment. I sped up the computing footage for the i7 and the 3950X by the same amount in each group of comparisons. That way you can see the difference in real time - without watching hours of video. The 100 photos that were used were Nikon D750 .NEF (RAW) files (~25MB each). The video that was used was GH5 4k 24fps 10 bit ALL-I (~54 GB).
Vaguely scientific real world tests: Photo editing in Adobe Lightroom 4k and 1080p video editing in Adobe Premiere Pro ASUS RealBench Red Dead Redemption 2 benchmark Borderlands 3 benchmark Link to build: https://kit.co/NickMuir/diy-pc-build-...

 

Hardware Unboxed covers AMD and Intel news + fills in AMD news not covered during the presentation - not enough time for AMD to share it all on stage:

News Corner | AMD Talks Big Navi & Zen 3 (Briefly), Nvidia 360Hz G-Sync, Intel Comet Lake Teaser
Jan 8, 2020
Hardware Unboxed
[spoiler = News Topics Index and Sources:]
00:00 - Lisa Su on High-End Navi, Zen 3, Desktop APUs
05:04 - Intel Teases Comet Lake H and Tiger Lake
08:02 - Intel Ghost Canyon NUC
10:06 - Nvidia and Asus Partner on 360Hz G-Sync
11:19 - Nvidia Game Ready Driver for CES
12:26 - What is Thunderbolt 4?
Sources:
https://www.anandtech.com/show/15344/...
https://videocardz.com/newz/amds-lisa...
https://www.tomshardware.com/news/int...
https://koolshare.cn/thread-168913-1-...
https://www.theverge.com/2020/1/7/210...
https://www.gizmodo.com.au/2020/01/ra...
https://www.nvidia.com/en-us/geforce/...
https://www.nvidia.com/en-us/geforce/...
https://www.tomshardware.com/news/wha...
[/spoiler]


AdoredTV weigh's in on Intel and AMD at CES 2020

Intel and AMD at CES 2020
Jan 8, 2020
AdoredTV
My recap and analysis of the big two at CES.
0:00 - More Dodgy Intel Marketing
5:39 - Spence Leaks Xbox Series X APU
6:17 - AMD at CES
24:18 - Intel at CES
30:06 - Nvidia at CES
30:11 - Summary

 

So much for all the insight, we got from all the analysis above, huh?

See:
https://www.anandtech.com/show/1535...t-exist-neither-does-the-intel-lga1159-socket

Keep it real guys, keep it real.

 

Do you know the origin of the TRX80 and WRX80 rumor?

It was from the official filing regarding a specific generation of USB support.

Now not all filings, such as patents, result in products that are made and sold to consumers.

Personally, I took the filing, comments from AMD clearly stating that memory bandwidth is an issue for the highest core count chips, part of the reason why, most likely, the 64-core chip only shows 50% more scaling over the 32-core chip shown at CES, the other factors being software optimization, frequency difference on boost, etc.

When you examine these together, you wind up with the logical conclusion that something would need done to resolve the memory bandwidth issue.

In comes increasing channels. This is the logical extension, with the other logical recourse being to wait for DDR5 in 2021. Even with that increased bandwidth, there is a possibility of more cores with Zen 4 on 5nm.

But adding extra channels also comes with challenges, from signal integrity on motherboard traces, which will be even more rigorous for DDR5, to the design of the I/O die, including size constraints (assuming they would also want to increase the channels for server boards).

Another option marries waiting for DDR5 with placing HBM on package, which acts as a massive first line with much higher bandwidth versus either DDR standard, so long as you can keep it feed from the DDR efficiently (which shouldn't be much of an issue).

So, are you going to say the filing was false now? Or rather wait for confirmation that a product will see the light of day? The latter is sound, the first is not.

 

The USB filings has resulted in multiple rumors. This included with the x399 and Zen 2 compliance. I think a lot of misinformation, intentional or otherwise, has extended from that. The lower scores to me are mathematically linked to 3.8 GHz all core on the 3970x on cb20 where only the base of 2.9 GHz was on the 3990x. further higher boosts from better silicon and cooling may yield higher results.

 

Did you read the article I linked? It explains it perfectly.

So, instead of letting the rumors die, you continue with more analysis and conjecture. Huh.

Wonder how the rumors got started? lol...

 

I am not sure the final of the 3990x but TBH it should blow about anything from Intl out of the water even with just the TRX40. About the only advantage of the TRX80 would have been the higher ram amounts from 8 channels. In aqbout a month or so more we should find out more.

 

No, I gave CONTEXT to the rumor NOT GIVEN in the article. There is a difference. I DID NOT dispute his findings. There is a difference!

 

Yeah, AMD binned their best chips for their most expencive Sku's (+$1399). Of course the remaining chips wasn't enougt to justify launch of 3950X as scheduled. And of course you couldn't find many 3900X after launch as well. Yeah, why let the best binned chips go to their cheaper High end mainstream chips as 3950X? Much better throw it's customers over to the gready wolfs who don't bother with MSR prices.

"These new CPUs have only been available in small quantities since launch and have been selling out almost immediately, leading to price gouging on eBay"

https://www.tomshardware.com/news/amd-threadripper-3970x-overclocking-record
"Another thing I noticed is that AMD appears to save the best chiplets for the 32- and 64-core parts. This is obvious seeing that all four 3970X CPUs I have tried beat my best 3950x and 3900x in max clock speed on water cooling! There's nothing wrong with tight binning, but for overclockers, finding a chip that you can just overclock to be competitive with the next SKU doesn't seem possible on the current-gen AMD chips. If you want to overclock your processor, pay the extra cash and get the X-series, as the non-X series are the chips that don't hit the higher boost bins"

 

Your post makes no sense and add's nothing to the discussion. Collecting a list of long past transient glitches that no longer are relevant and waffling on about nothing helpful in a random outburst, why?

Kinda reminds me of the last historical "burning of an empire", While Cesar Fiddles - Rome Burns, @Papusan mumbles incoherently as Intel turns into a smoking hole in the ground.

How is this binning of silicon different than AMD and Intel have done throughout history?

Higher binned silicon makes it's way up the ladder to higher performance parts - where they can be appreciated by their owners - that's why they are paying more, for exceptional performance.

The AMD lower end components meet their specifications and provide great price / performance, and x components also OC as I've posted - and sooo many others have shared as well.

The 3950x single core outperforms the 3900x, and draws less power for the same tasks, so it's a great upgrade for only $250 more than the 3900x and also adds 2c/4t's more. It's really the sweet spot in the AM4 line for highest performance / throughput.

But, the 3700x at only $300 gives 8c/16t vs Intel's overpriced parts and the 3700x @ all core 4343mhz outperforms Intel's 9900k's overclocked to 5ghz. Save a bunch of $ and use the CPU savings to get 5GB/sec SSD's, couple that with a nice $150 Asus x570-P (not Prime) or the new MSI MAG Tomahawk x570 for $189, and you've no need for Intel's no-undervolt / overpatched insecure CPU's.

Take those potential 3700x configurations and sub in the 3600x and save $100 for just as good gaming performance when you don't intend on streaming or running loads capable of needing 8c/16t's, and save $60 more and sub in a 3600 and do manual OC'ing, owners have gotten good results doing that too.

Why buy old broken technology like Intel? Can't get unstuck from the past? Try looking up and out past the wet paper sack, then tear it off and emerge into the light.

Zen 3 is coming, but no need to wait with such great AMD AM4 CPU's.

 

Of course it makes sense. AMD screwed their 3900X and 3950X buyers. Only Good know what they had to pay getting an chips while AMD collected most of the chips for coming TR.

 

You are being silly.

AMD have blessed their AMD 3950x owners, and blessed their 3900x owners - they are thrilled with their purchases.

You are blowing raspberries of jealousy, nothing more.

When are you gonna admit Intel has lost and buy yourself some AMD joy? It's inevitable... I hope soon, I feel sad for you and all the Intel curmudgeons... sigh.

I'm sure you'd enjoy benchmarking on a ThreadRipper 3, why not follow your own advice and buy the best binned silicon from AMD

 

Nope. "finding a chip that you can just overclock to be competitive with the next SKU doesn't seem possible on the current-gen AMD chips". I feel just sad for the AMD customers seeing this stated. @Mr. Fox

Of course... You still have SL. And for the time they have some aviable https://siliconlottery.com/collections/matisse/products/3950x41p

 

That is only partially true, and it's not silly to everyone. It depends on the customer. Not all AMD Ryzen owners are happy with the lackluster overclocking abilities short of going to extreme cooling measures like LN2 or phase change. Now, if you don't care about having a great overclocking experience, then sure... what's not to love if you are content with stock, or near-stock, performance? If you do care about overclocking, it's not all sunshine and roses. We have at least two forum members in the desktop overclocking threads that are not particularly thrilled with that situation. One is selling his new AMD mobo and CPU because it's too boring for someone that enjoys overclocking. Wrong tool for the job if the job is overclocking. It's powerful and all that, but it's a belly-button solution.