Wednesday, August 9th 2017

AMD Ryzen Threadripper 1950X Overclocked to 4.1 GHz With Liquid Cooling

Redditor "callingthewolf" has posted what is an awe-inspiring result for AMD's Ryzen Threadripper 1950X (that's an interesting username for sure; let's hope that's the only similarity to the boy who cried wolf.) The 16-core, 32-thread processor stands as the likely taker for the HEDT performance crown (at least until Intel's 14-core plus HEDT CPUs make their debut on the X299 platform.) With that many cores, highly thread-aware applications naturally look to see tremendous increases in performance from any frequency increase. In this case, the 1950X's base 3.4 GHz were upped to a whopping 4.0 GHz (@ 1.25 V core) and 4.1 GHz (at 1.4 V core; personally, I'd stick with the 4.0 GHz and call it a day.)

The feat was achieved under a Thermaltake Water 3.0 liquid cooler, on a non-specified ASRock motherboard with all DIMM channels populated with 8 x 8 GB 3066 MHz DIMMs. At 4.0 GHz, the Threadripper 1950X achieves a 3337 points score on Cinebench R15. And at 4.1GHz, the big chip that can (we can't really call it small now can we?) manages to score 58391 points in Geekbench 3. While those scores are certainly impressive, I would just like to point out the fact that this is a 16-core CPU that overclocks as well as (and in some cases, even better than) AMD's 8-core Ryzen 7 CPUs. The frequency potential of this Threadripper part is in the same ballpark of AMD's 8-core dies, which speaks to either an architecture limit or a manufacturing one at around 4 GHz. The Threadripper 1950X is, by all measurements, an impressively "glued together" piece of silicon.
Sources: Reddit user @ callingthewolf, via WCCFTech
Add your own comment

188 Comments on AMD Ryzen Threadripper 1950X Overclocked to 4.1 GHz With Liquid Cooling

#1
PowerPC
HopelesslyFaithful said:
who care what out of the box stock is? No one buys an unlocked 1000 dollar chip and runs stock unless they are an idiot, Your point is moot.
The real question is what idiot buys a $1000 CPU and immediately voids the warranty by delidding it?

HopelesslyFaithful said:
Again if you seriously game ryzen sucks at it. Look at the benchmarks, look at single thread. If you VR or 1440p 120hz ULMB game like I do even a 4.8 Ghz 6700K can't run many single thread games at 120hz consistently due to single thread limits. Stop your spamming of patently false statements on its ability to game on VR and 120hz. I wouldnt even use an Intel HEDT for it because of the IPC hit and thats faster than Ryzen.
Again, "serious gaming"... what does that mean? Do you even know how pretentious you sounds?

You didn't mention anything about "VR and 120Hz" or "1440p 120hz ULMB" in your original post. But still you had to make statemtns such as "Ryzen is non-gaming" and "Intel is top tier at gaming". Are you kidding me? You represent less than 1% of actual gamers who don't have Titan X in SLI for your "VR and 120Hz". Your statements don't make sense for over 99% of actual gamers who aren't hopelessly lost in your VR clouds. Everybody knows VR and 120Hz is years away for most people. Get this in your head: Not everybody shits money like you. Don't assume that everybody needs or wants the most power at all costs, just because you print money.

Fact is, if you're looking for a gaming PC from $500 to $1500, like the vast majority of people, Ryzen is the "gaming chip" you should buy! And this is the last time I will use this extremely douchy term that you just made up.

HopelesslyFaithful said:
Your just proving why i think humanity is a joke lol
LOL?
Posted on Reply
#2
Durvelle27
HopelesslyFaithful said:
yep the fastest CPU in the world at gaming coming in at a terrible 5.2GHz on water is terrible at gaming compared to an amazing 4.1GHz Ryzen with lower IPC......

Your just proving why i think humanity is a joke lol
Hmmm 1000MHz higher , high power draw, heat box and still just barely performs better

Yep that screams great deal

Intel forever screwing consumers and brainwashing
Posted on Reply
#3
HopelesslyFaithful
Durvelle27 said:
Hmmm 1000MHz higher , high power draw, heat box and still just barely performs better

Yep that screams great deal

Intel forever screwing consumers and brainwashing
yes somehow 30% faster in single thread is barely better. You are a troll on a grand level to ignore basic mathematical facts.

Don't worry i wont feed this troll anymore. Just want readers to notice how much a troll and how basic facts fail him. Minimum frame rate is most important metric and causes the most issue in terms of quality of experience especially in FPS. When I play NS2 when other players shoot at my skulk I get massive FPS drops and makes it harder to bite marines. This is a fact in any game especially in single thread games.

Ryzen fails badly at 1000s of games that are still single thread and or people who game with VR or 120hz.

HopelesslyFaithful said:
on a 6 year old cpu lol



yep and that makes VR and 120hz ULMB 100% unplayable :slap:
PowerPC said:
The real question is what idiot buys a $1000 CPU and immediately voids the warranty by delidding it?


Again, "serious gaming"... what does that mean? Do you even know how pretentious you sounds?

You didn't mention anything about "VR and 120Hz" or "1440p 120hz ULMB" in your original post. But still you had to make statemtns such as "Ryzen is non-gaming" and "Intel is top tier at gaming". Are you kidding me? You represent less than 1% of actual gamers who don't have Titan X in SLI for your "VR and 120Hz". Your statements don't make sense for over 99% of actual gamers who aren't hopelessly lost in your VR clouds. Everybody knows VR and 120Hz is years away for most people. Get this in your head: Not everybody shits money like you. Don't assume that everybody needs or wants the most power at all costs, just because you print money.

Fact is, if you're looking for a gaming PC from $500 to $1500, like the vast majority of people, Ryzen is the "gaming chip" you should buy. And this is the last time I will use this extremely douchy term that you just made up.


LOL?
yea you obviously dont read peoples posts and just spam drivel. I stated it in my first post in this thread and in basically every post since.

Again most games and OS is single thread limited. Even older games that dont require a 980TI at 1500mhz like I have are still single thread limited on my rig and have been for over a decade (since game release). RCT3, Total War, Star Wars Empire at War, WAr Thunder, and 100s-1000s of others are still single thread limits and scale nearly perfectly so top tier OC 7700K vs a top tier ryzen OC is a solid 30% faster in FPS.

Fun Fact HL2 was single thread limited until 4-5GHz SB/IB systems. So drops below 120hz and as low as 60hz was quite common on my 3.33ghz nahelam.
It wasn't until my 4.2GHz IB that I no longer had major issues with HL2.

Again minimum frame rate is the most important thing for VR and 120hz ULMB and i even have stutters with my rig. You expect me or others to throw away 20-30% performance for unbearable stutters and frame drop?

jeez. Logic and facts fail you hard huh?

(this refers to single thread limits across all programs and OS so ryzen vs an OC Kaby has major differences in snappiness)
This should be required reading for everyone because for some reason 35 years later people still don't understand this and this has been known since 1982! This is again why I don't use my Server as my main rig. It is substantially slower in day to day tasks and I can feel it.

You might want to read this about OS responsiveness since you seem deeply uninformed.

https://www.techpowerup.com/forums/threads/intel-coffee-lake-cpus-will-require-new-motherboards.235789/page-2#post-3704281

Read this post til the end of thread if you actually want to learn something.
Posted on Reply
#4
PowerPC
HopelesslyFaithful said:
yea you obviously dont read peoples posts and just spam drivel. I stated it in my first post in this thread and in basically every post since.

Again most games and OS is single thread limited. Even older games that dont require a 980TI at 1500mhz like I have are still single thread limited on my rig and have been for over a decade (since game release). RCT3, Total War, Star Wars Empire at War, WAr Thunder, and 100s-1000s of others are still single thread limits and scale nearly perfectly so top tier OC 7700K vs a top tier ryzen OC is a solid 30% faster in FPS.

Fun Fact HL2 was single thread limited until 4-5GHz SB/IB systems. So drops below 120hz and as low as 60hz was quite common on my 3.33ghz nahelam.
It wasn't until my 4.2GHz IB that I no longer had major issues with HL2.

Again minimum frame rate is the most important thing for VR and 120hz ULMB and i even have stutters with my rig. You expect me or others to throw away 20-30% performance for unbearable stutters and frame drop?

jeez. Logic and facts fail you hard huh?

(this refers to single thread limits across all programs and OS so ryzen vs an OC Kaby has major differences in snappiness)
This should be required reading for everyone because for some reason 35 years later people still don't understand this and this has been known since 1982! This is again why I don't use my Server as my main rig. It is substantially slower in day to day tasks and I can feel it.

You might want to read this about OS responsiveness since you seem deeply uninformed.

https://www.techpowerup.com/forums/threads/intel-coffee-lake-cpus-will-require-new-motherboards.235789/page-2#post-3704281

Read this post til the end of thread if you actually want to learn something.
You obviously didn't read mine. And are you expecting me to remember your other posts among over 150 in this thread alone? Sorry I don't have a photographic memory. If you want others to remember what you meant, it's no big effort to repeat it in every post, if you want to make multiple posts in a big thread like this. And you're not that important that I would remember you among thousands of other posters.

Back to topic.

You think I need to read all that just to get from it that most games are single thread limited on the CPU side, something I already know?

First of all, I don't see such big differences as you describe in the benchmarks. In most tests I see about 10%, not 30%.

But as I said multiple times, I think this kind of discussion is stupid to the core because it only applies to people like you who can already afford the most powerful GPU money can buy. (You notice how all those benchmarks are made with two Titans, right?) Your perspective is really boring and pointless for most people. You are completely neglecting the cost factor. How is someone going to be helped by calling Intel a "gaming CPU" and Ryzen a "non-gaming CPU" when he or she should definitely go for Ryzen with a normal $500-$1500 gaming PC budget, that includes a graphics card!?

Why do you insist on these stupid "gaming CPU" or "non-gaming CPU" terms by just isolating the CPU from the whole gaming PC? Terms like this are incredibly confusing and frustrating for consumers, as people will inadvertently end up buying a worse gaming PC with your alleged 6700K "gaming CPU", when they should have bought a much better gaming PC with a Ryzen and a better GPU for gaming!

Really, stop labeling CPUs "gaming" or "non-gaming". There is no such thing as gaming on a CPU in isolation without a GPU. If you want to say something about the CPU performance in games, don't just conveniently omit what GPU you used. You can only say a CPU is good or bad for gaming, if you include the GPU in that statement. Anything less is literally just nonsense or only applies to you. You're also directly playing into the Intel vs. AMD war, or simply neglecting to prevent it.
Posted on Reply
#5
justimber
HopelesslyFaithful said:


Your just proving why i think humanity is a joke lol
don't you think you also belong to that group? unless you landed in Roswell from another planet lol

anyways, if it's price/performance ratio, AMD wins...and yeah I agree with them, we don't shit money like you do..unless you want to share that with us and we'll also make our eyes bleed with VR and 120hz :P

why is it when I read the posts here it seems that Intel fanboys are on panic mode?
Posted on Reply
#6
Totally
HopelesslyFaithful said:
Mental diarrhea and other senselessness
All your arguments are pointless to the brink of being delusional. Yes the chip you are so proud of scores 150fps in one vs AMD's chip that only gets 120fps. Does it matter at that point? That constitutes a very minor point. Justly defined as barely, the absolute difference is significant but the actual tangible benefits are next to none.

You are being the guy with a 800hp sports car bragging about how your car's estimated top speed is nearly 200mph and is getting butthurt when then other guy points out that speed limits top out at 75mph and their Corolla hit's that number just fine.

Then you go onto wax eloquent about how the chip performs on a nearly 20 year old game?
Posted on Reply
#7
HopelesslyFaithful
justimber said:
don't you think you also belong to that group? unless you landed in Roswell from another planet lol

anyways, if it's price/performance ratio, AMD wins...and yeah I agree with them, we don't shit money like you do..unless you want to share that with us and we'll also make our eyes bleed with VR and 120hz :p

why is it when I read the posts here it seems that Intel fanboys are on panic mode?
As I pointed out in the other post linking to a discussion about IBMs study from 1982. There is substantial cost savings by having a faster system in regard to time saved and increased productivity, which is why I use fast quads for my main rig so yes....it is a worth while investment. You should read the IBM study.

Also if you can't afford a good CPU...why are you buying 60 dollar titles lol.

I rather buy games on sale and own a better rig but to each their own.

Totally said:
All your arguments are pointless to the brink of being delusional. Yes the chip you are so proud of scores 150fps in one vs AMD's chip that only gets 120fps. Does it matter at point? That constitutes a very minor point. Justly defined as barely, the absolute difference is significant but the actual tangible benefits are next to none.

You are being the guy with a 800hp sports car bragging about how your car's estimated top speed is nearly 200mph and is getting butthurt when then other guy points out that speed limits top out at 75mph and their Corolla hit's that number just fine.

Then you go onto wax eloquent about how the chip performs on a nearly 20 year old game?
as I said before minimum frame rate is really the only metric that matters. Average FPS is not important. They also test only AAA games that are actually threaded. I dont see one single thread game tested.

Also it cant even average 120 fps let alone keep 99% or 99.9% of frames above it. lol (going from tweaktown picture in first thread)

There is a 36% difference in minimum frame rate with the 7740x vs 1800x. Lets see how much worse that looks when Kaby is at 5.2GHz vs 1800x at 4.1GHz. I bet more like 40-45% difference....yep....totally not noticeable.:slap:

That is also GTA:5 which is a very well threaded and optimized game. Try playing older games that arent even coded that well. Not happening.

If you only play AAA threaded games sure those ryzen maybe acceptable in that niche area of gaming. If you don't game with VR or 120hz sure....it may work. But if you VR or play at 120hz or want to play a wide range of games old and new at a smooth experience ryzen does not cut it....That is a fact.

Also your analogies are not even relevant and patently stupid but whatever.
Posted on Reply
#8
PowerPC
HopelesslyFaithful said:
As I pointed out in the other post linking to a discussion about IBMs study from 1982. There is substantial cost savings by having a faster system in regard to time saved and increased productivity, which is why I use fast quads for my main rig so yes....it is a worth while investment. You should read the IBM study.
But... but... Ryzen Octacore wipes the floor with Intel Quadcore in productivity and multitasking. It's the same as having two monitors instead of one for a huge productivity boost. I won't even get into an argument over it because anything trying to contradict this could only be seen as trolling.

Totally's car analogy is actually perfect for that "gaming CPU" drivel.
Posted on Reply
#9
HopelesslyFaithful
PowerPC said:
But... but... Ryzen Octacore wipes the floor with Intel Quadcore in productivity and multitasking. It's the same as having two monitors instead of one for a huge productivity boost. I won't even get into an argument over this because anything trying to contradict this could only be seen as trolling.

Totally's car analogy is actually perfect for that "gaming CPU" drivel.
Again as for win 7 is concerned most of the OS is single thread limited. Win10 is a bit better threaded but still single in most tasks.

Web browsing and PDF and most office programs and day to day tasks are not threaded and single thread is more important.

Go read IBMs study on responsiveness.

You still are wrong. A fast single thread is still better than many cores for day to day tasks. If it wasn't I wouldnt buy binned quad core CPUs.
Posted on Reply
#10
PowerPC
HopelesslyFaithful said:
Again as for win 7 is concerned most of the OS is single thread limited. Win10 is a bit better threaded but still single in most tasks.

Web browsing and PDF and most office programs and day to day tasks are not threaded and single thread is more important.

Go read IBMs study on responsiveness.

You still are wrong. A fast single thread is still better than many cores for day to day tasks. If it wasn't I wouldnt buy binned CPUs.
Are you serious? You obviously have no clue about the benefits of having multiple cores.

Separate programs, even if they are single core limited will still always run in separate threads on different cores! Yes, an isolated PDF reader or office program may run just on a single core, but again, that's if you just isolate one of them. Each of those programs running in parallel still will be assigned to its own core. Even each tab in Chrome runs in it's own thread. Just look in your task manager. It means it will use multiple CPU cores.
Posted on Reply
#11
Durvelle27
HopelesslyFaithful said:
Again as for win 7 is concerned most of the OS is single thread limited. Win10 is a bit better threaded but still single in most tasks.

Web browsing and PDF and most office programs and day to day tasks are not threaded and single thread is more important.

Go read IBMs study on responsiveness.

You still are wrong. A fast single thread is still better than many cores for day to day tasks. If it wasn't I wouldnt buy binned quad core CPUs.
You might as well have a P4 than
Posted on Reply
#12
HopelesslyFaithful
PowerPC said:
Are you serious? You obviously have no clue about the benefits of having multiple cores.

Separate programs, even if they are single core limited will still always run in separate threads on different cores! Yes, an isolated PDF reader or office program may run just on a single core, but again, that's if you just isolate one of them. Each of those programs running in parallel still will be assigned to its own core. Even each tab in Chrome runs in it's own thread. Just look in your task manager. It means it will use multiple CPU cores.
and my CPU almost never maxes out in said work loads and those select cases of maxing out is not worth the cost of single thread performance. I watch the load and log the load on my own system and know the added single thread performance is better hence why i dont use HEDT as main rig even though I have a 1650v3 (at 4.2-4.4 GHz) next to me.

Go read the IBMs study and educate yourself and do some math and log your system and see where you hit the wall the most. I single thread max out far more than multithread max.

The time saved with having a better single thread is substantial hence why i use quads and will get the 8700K when it comes out since that should overclock comparable and give me an extra 2 cores. If a quad exists with 10% or more better single thread than a 6/8 core CPU it is overall better day to day CPU and better overall gaming CPU for those tasks.

Durvelle27 said:
You might as well have a P4 than
and no surpise you post something meaningless devoid of facts.
Posted on Reply
#13
Hood
INSTG8R said:
Not sure why people keep trotting out gaming performance...If you bought this for gaming you're doing it wrong...:wtf:
I think it's because AMD and their board partners are marketing it towards gamers in a big way, probably because the legitimate workstation market is so small. Bling it up with LEDs and call it ROG, Aorus Gaming, Gaming professional, or Gaming Pro carbon, and you trick a lot of people into wanting one for their gaming rig. Too bad they'll have to find out it's FX all over again, except with a much higher price tag this time. Well, maybe not as bad as FX, but the same concept - more cores = better (pay no attention to the abysmal single-thread performance - it has moar cores!).
Posted on Reply
#14
PowerPC
HopelesslyFaithful said:
and my CPU almost never maxes out in said work loads and those select cases of maxing out is not worth the cost of single thread performance. I watch the load and log the load on my own system and know the added single thread performance is better hence why i dont use HEDT as main rig even though I have a 1650v3 (at 4.2-4.4 GHz) next to me.

Go read the IBMs study and educate yourself and do some math and log your system and see where you hit the wall the most. I single thread max out far more than multithread max.

The time saved with having a better single thread is substantial hence why i use quads and will get the 8700K when it comes out since that should overclock comparable and give me an extra 2 cores. If a quad exists with 10% or more better single thread than a 6/8 core CPU it is overall better day to day CPU and better overall gaming CPU for those tasks.


and no surpise you post something meaningless devoid of facts.
I have seen a video where they opened a couple of programs, games and videos at the same time and the Ryzen 8-core didn't flinch like you could throw anything at it, while the Intel 4-core started having massive stutter in the game. That's the kind of real world testing you need to do, not math with pen and paper.

Chrome for example runs multiple tabs in different processes each with multiple threads. This means it will use multiple CPU cores just nicely, as many as you throw at it. You may say that single threaded performance is more important for productivity, but how exactly did you come to that conclusion except from a study that was conducted 40 years ago and personal "logging", as you say?

Maybe that your server configuration is more different to your desktop 4-core than just the CPU? (RAM, drives etc.)
Posted on Reply
#15
HopelesslyFaithful
PowerPC said:
I have seen a video where they opened a couple of programs, games and videos at the same time and the Ryzen 8-core didn't flinch like you could throw anything at it, while the Intel 4-core started having massive stutter in the game. That's the kind of real world testing you need to do, not math with pen and paper.

Chrome for example runs multiple tabs in different processes each with multiple threads. This means it will use multiple CPU cores just nicely, as many as you throw at it. You may say that single threaded performance is more important for productivity, but how exactly did you come to that conclusion except from a study that was conducted 40 years ago and personal "logging", as you say?
again as i keep saying and you ignoring. Go read the study and understand how human minds and productivity works and come back after you read the study and read the facts on how the human mind has and will always work.

each tab is single thread limited. fact.

where do most transactions occur? With in a single tab/loading 1 tab at a time.

Which hurts productivity more?
100 600ms responses vs 100 300ms responses?
30 seconds of total system delay
or
20 10 second responses vs 20 5 second responses?
100 seconds of total system delay.
Posted on Reply
#16
RejZoR
HopelesslyFaithful said:
who care what out of the box stock is? No one buys an unlocked 1000 dollar chip and runs stock unless they are an idiot, Your point is moot.



Who spends that much money and does not delid and overclock? A fool maybe? It is stupid not to delid...your throwing away perforamnce.

that is not true. Silicon lottery delids his 7700K for single phase change because it makes a notable difference. He runs his at 5.5-5.7GHz. There is a thread in his forum where i was asking him about it.



Again if you seriously game ryzen sucks at it. Look at the benchmarks, look at single thread. If you VR or 1440p 120hz ULMB game like I do even a 4.8 Ghz 6700K can't run many single thread games at 120hz consistently due to single thread limits. Stop your spamming of patently false statements on its ability to game on VR and 120hz. I wouldnt even use an Intel HEDT for it because of the IPC hit and thats faster than Ryzen.
People who actually work on them for a living don't overclock them. And your point is silly, AMD has all CPUs unlocked anyway. Even bottom ones.
Posted on Reply
#17
Khonjel
Man, I don't understand people sometimes. I can understand people bashing Intel for stifling improvements and features but bashing AMD for finally introducing competition? What the hell!

And look at some guys bragging like "oohhh my $450k Ferrari or Lamborghini is still faster than a $100k GT-R35 lol"

People are gonna buy what they want. Hell, we might see people buying that 8-core TR just for the heck of it. No amount of loud-mouthing in the forums or YouTube comments will change that.
Posted on Reply
#18
PowerPC
HopelesslyFaithful said:

Which hurts productivity more?
100 600ms responses vs 100 300ms responses?
30 seconds of total system delay
or
20 10 second responses vs 20 5 second responses?
100 seconds of total system delay.
What is the difference here? Of course in both cases more delay hurts productivity.

I did read your other thread.

But what was shown in that video I described doesn't bother you? That's not just 100ms delay, if you can even prove that with something like Chrome. That's noticeable stutter in a game that makes it unplayable while running other programs in the background on a 4-core compared to completely smooth gameplay with the same programs running on the 8-core.

Did you actually capture those responses on video or can it maybe just be placebo with something like a word processor? Can't it be the monitor, lighting or so many other things at this point?
Posted on Reply
#19
Melvis
How the heck did this thread get 7 pages?

TR OC's to 4.1GHz.....Cool? unlike all the other Ryzen CPU's do just with more cores?
Posted on Reply
#20
HopelesslyFaithful
PowerPC said:
What is the difference here? Of course in both cases more delay hurts productivity.

I did read your other thread.

But what was shown in that video I described doesn't bother you? That's not just 100ms delay, if you can even prove that with something like Chrome. That's noticeable stutter in a game that makes it unplayable while running other programs in the background on a 4-core compared to completely smooth gameplay with the same programs running on the 8-core.

Did you actually capture those responses on video or can it maybe just be placebo with something like a word processor? Can't it be the monitor, lighting or so many other things at this point?
If you read the thread and IBMs study you would realize that there is 3 seconds of time saved ever every transaction between 600ms and 300ms delay. Granted, different tasks have different human response delays but it is universal in regards to the concept (humans see a larger time saved at lower delays). Go read the study.

There is no notable added human delay in a 10 vs 5 second delay.

According to IBMs study there is roughly 3 seconds human delay difference between 600 ms vs 300ms. You would know this if you actually read it.....but you obviously didn't :slap:

So the 600ms vs 300ms scenario is more damaging to productivity and enjoyably of using a system than large delays. So many small delays are more damaging than a few large delays. Small delays in system responsiveness are more common in day to day tasks which is why I build my system the way I do. These delays are almost always linked to single thread.

You can log and gauge your system using many tools. There is throttlestop, logging mouse actions (seeing how many transaction you do per hour..not a perfect method but gives usable statistics), windows tool I forget the name of it but it lets you look at all the threads and see where throttling is according. You can see this in throttlestop also to an extent.

I wish I know of a program that logged all single core/thread maxes and counted them and reported the delay for each instance so you didn't have to figure it out in a drawn out not perfectly accurate manner.

Those statistics would be amazing to read and to see how to optimize a system on a user by user and program by program basis.

From my work experience and watching, higher single thread is better if we are considering a 10% difference in single thread. Again depends on what you do but if you dont have a specific load where 8+ cores is needed your doing a disservice to yourself.

I dont get your last statement in post. It makes no sense.

And go read the study. I am no longer going to bottle feed you because you want to be lazy and not take the chance to actually learn something
Posted on Reply
#21
PowerPC
HopelesslyFaithful said:
I wish I know of a program that logged all single core/thread maxes and counted them and reported the delay for each instance so you didn't have to figure it out in a drawn out not perfectly accurate manner.

Those statistics would be amazing to read and to see how to optimize a system on a user by user and program by program basis.
This is exactly why I think this video would clear some things up. I found it btw. The test starts exactly 5 minutes in:


You can see there is a much greater hit in overall performance to all programs on Intel 6700K when having multiple programs open compared to the Ryzen 1700X. It shows that you maybe are underrating the effect somewhat.

Still, this seems like a far better argument for Intel than the tired and shameless "gaming CPU" argument. Responsiveness in desktop applications would actually depend purely on the CPU. I didn't think I'll be redefining my understanding of productivity using computers in this thread lol. Even though I feel like most people would probably find this kind of thing somewhat OCDish and would rather want to future proof their PC with double the cores rather than obsess about 100 ms more or less delay in responsiveness. Doesn't matter because the research is solid. At least for now, you probably will have a more responsive desktop experience with Intel, although what I meant was it's still something to be proven for me personally. Have you done the actual responsiveness measurements? Because you wrote somewhere that it just "feels" more responsive to you, but it still might actually be placebo or some other factor with your server compared to the 4-core PC. Have you ruled that out? Don't programs have some kind of maximum responsiveness anyway that you can't go beyond? I have never actually cared enough to think about this until 10 minutes ago. It's not exactly something that stands out when you use a computer with an already decent CPU.
Posted on Reply
#22
HopelesslyFaithful
PowerPC said:
This is exactly why I think this video would clear some things up. I found it btw. The test starts exactly 5 minutes in:


You can see there is a much greater hit in overall performance to all programs on Intel 6700K when having multiple programs open compared to the Ryzen 1700X. It shows that you maybe are underrating the effect somewhat.

Still, this seems like a far better argument for Intel than the tired and shameless "gaming CPU" argument. Responsiveness in desktop applications would actually depend purely on the CPU. I didn't think I'll be redefining my understanding of productivity using computers in this thread lol. Even though I feel like most people would probably find this kind of thing somewhat OCDish and would rather want to future proof their PC with double the cores rather than obsess about 100 ms more or less delay in responsiveness. Doesn't matter because the research is solid. At least for now, you probably will have a more responsive desktop experience with Intel, although what I meant was it's still something to be proven for me personally. Have you done the actual responsiveness measurements? Because you wrote somewhere that it just "feels" more responsive to you, but it still might actually be placebo or some other factor with your server compared to the 4-core PC. Have you ruled that out? Don't programs have some kind of maximum responsiveness anyway that you can't go beyond? I have never actually cared enough to think about this until 10 minutes ago. It's not exactly something that stands out when you use a computer with an already decent CPU.
It was proven in 1982 and the data exists. Go read it. How the human brain works does not change.

Again I have analyzed my system and studied the logs and concluded single thread is the most hindrance to me for day to day tasks. If I am doing anything that maxes my CPU. I move it to my server to run the task.

Except it does stand out if you actually edit win 7 OS settings like remove fixed animations and other settings which is why I turned my S8 animations to .5x too.

What you just stated is the same though process that IBM disproved 35 years ago. :banghead::slap:

Additionally, ryzen and intel mesh both have bad penalties switching between cores/clusters.

The 7700K and 8700K do not have those penalties either.

Do you have a ryzen chip? Run kraken on it for me.


and that video is a no duh....he is doing an extremely power hungry task in background derp derp...of course.

No where did I say running an encode or F@H in the background would be faster on a quad core chip lol.

Your example is stupid.
Posted on Reply
#23
PowerPC
HopelesslyFaithful said:
It was proven in 1982 and the data exists. Go read it. How the human brain works does not change.

Again I have analyzed my system and studied the logs and concluded single thread is the most hindrance to me for day to day tasks. If I am doing anything that maxes my CPU. I move it to my server to run the task.

Except it does stand out if you actually edit win 7 OS settings like remove fixed animations and other settings which is why I turned my S8 animations to .5x too.

What you just stated is the same though process that IBM disproved 35 years ago. :banghead::slap:

Additionally, ryzen and intel mesh both have bad penalties switching between cores/clusters.

The 7700K and 8700K do not have those penalties either.

Do you have a ryzen chip? Run kraken on it for me.


and that video is a no duh....he is doing an extremely power hungry task in background derp derp...of course.

No where did I say running an encode or F@H in the background would be faster on a quad core chip lol.

Your example is stupid.
I don't have a Ryzen but I will buy it if the 8700K isn't going to come out soon and isn't cheap. I'm not going to go 4-core in 2017, no way. I generally use a CPU for way over 5 years and I'm sure multi-core applications will pick up, if they aren't already gaining momentum. Intel 6-core would be the only compromise, but there is not even a release date for it yet, or other information for that matter.

It's Ryzen 1700 or wait for me right now, those are the two only choices. Intel is sleeping behind the wheel.
Posted on Reply
#24
Vayra86
HopelesslyFaithful said:

Again most games and OS is single thread limited. Even older games that dont require a 980TI at 1500mhz like I have are still single thread limited on my rig and have been for over a decade (since game release). RCT3, Total War, Star Wars Empire at War, WAr Thunder, and 100s-1000s of others are still single thread limits and scale nearly perfectly so top tier OC 7700K vs a top tier ryzen OC is a solid 30% faster in FPS.

Fun Fact HL2 was single thread limited until 4-5GHz SB/IB systems. So drops below 120hz and as low as 60hz was quite common on my 3.33ghz nahelam.
It wasn't until my 4.2GHz IB that I no longer had major issues with HL2.

Again minimum frame rate is the most important thing for VR and 120hz ULMB and i even have stutters with my rig. You expect me or others to throw away 20-30% performance for unbearable stutters and frame drop?

jeez. Logic and facts fail you hard huh?


You might want to read this about OS responsiveness since you seem deeply uninformed.

https://www.techpowerup.com/forums/threads/intel-coffee-lake-cpus-will-require-new-motherboards.235789/page-2#post-3704281

Read this post til the end of thread if you actually want to learn something.
Speaking of logic - why are you considering VR and 120hz a great investment if EVEN with that 30% higher single thread performance and top end gear, you STILL experience stutter?

Would it not be much more effective to lock at 90 fps, and scale a system on that, avoid stutter because you don't have that high a frame time variance altogether, AND reduce the total cost of the system so you can invest more towards GPU / future upgrades / whatever else in peripherals?

The only real logic here, is that high FPS/high refresh gaming AND VR simply aren't ready for market yet, because the performance level and optimization of the technology is far beyond the spec of the hardware it is used on. You're basically pointing that out yourself. Its on an even worse position right now than 4K, which is achievable but not comfortable on most hardware.

The fact still is, that the market for high refresh gaming is purely competitive of nature and VR and competitive don't mix at all, for now and the foreseeable future. The only failure in logic lies with you and your view of what is feasible, and what is possible with todays' tech.

The only real high refresh gaming that happens realistically is playing CS:GO and MOBAs at the lowest detail settings and at a high competitive level. Everything else is just a personal preference of which the vast majority of 'hardcore gamers' are just overinflated ego's like yourself that achieved nothing in life but a rank on some silly video game. You then apply that ego to everyone's buying decisions with regards to CPUs while the ONLY niche that the 7700K excels in, is high refresh gaming. For everyone else, Ryzen is the go-to CPU right now.

Get a life and some sense in your gut. Shit. Speaking of faith in humanity...
Posted on Reply
#25
Hood
PowerPC said:
Intel 6-core would be the only compromise, but there is not even a release date for it yet, or other information for that matter.
I guess you missed this from Tuesday - https://www.techpowerup.com/forums/threads/intel-to-debut-8th-generation-core-family-on-august-21.235943/ - this is just a paper launch, but I think Intel usually follows with the retail launch fairly quickly, and "before the holidays" is implied in the press release, so probably Oct/Nov for pre-built systems and Dec. for CPUs.
Posted on Reply
Add your own comment