**UPDATED: 3DMARK RESULTS** Intel Core i9-9900K, Core i7-9700K and Core i5-9600K CPU Performance Leaks *+Release date rumour*

[porina]

9 hours ago, syn2112 said:

note to self: never trust Geekbench for any results

Assuming the other stuff you wrote is correct (as in, I haven't bothered to check) then all it means is only look at Geekbench for non-HT performance. Not everything benefits from HT.

 

5 hours ago, NumLock21 said:

Geekbench is not a reliable benchmark, wait for Cinebench.

Cinebench (assuming R15) may be well known and predictable, but it is also one possible use case. We know it scales well with cores, gets a nice boost with HT/SMT, and isn't very ram sensitive. Still, not representative of all use cases.

 

5 hours ago, NowakVulpix said:

Don't get why the i7 couldn't stay as a 6c12t SKU, a la Coffee Lake

On average, an 8 core no HT part would probably perform better most of the time. 8 cores are 8 cores. 6c12t could be equivalent to 6 to 9 cores, depending on the workload. Cinebench R15 would be equivalent to around 8 cores and it would be uncommon to see much more gain than that. 

[yian88]

Sorry but no HT means im buying AMD ryzen 5 2600 its really cheap and then going to upgrade to 7nm ryzen 7 8 core next year.

With all their locked mobos and cpu's and lack of HT and forced selling of iGPU's im done with Intel, maybe 2600 has 10% less performance because of lower clock speeds but its going to be worth it for a 200$ cheaper build(where i live) versus i5 9600k/8600k fuck that.

HT is just a matter or fliping a microcode switch on/off, yet intel sells garbage cpu's without HT be gone assholes.

[PianoPlayer88Key]

10 hours ago, 2Buck said:

… I'm honestly hoping for much higher clock speeds, if Zen 2 really is THAT more efficient than Zen+, than that'd be great to see that translate into better clocks, but who knows. …

At this point, I'd like to see better performance PER clock, PER thread.  I think we're long overdue for a significant jump there, like Core 2 was over Pentium 4, or Sandy Bridge over Westmere.  I'd like to see at least 200-250 Cinebench at sub-4 GHz, single-threaded, for example.  (I wonder where we would be now if every generation in the last 10 or 12 years or so had the same jump as C2 over P4 ...)

 

Around 2021 or 2022 or so, when the time may come for me to replace my desktop with its i7-4790K, I'd like the ~$250 (if bought at MicroCenter on Black Friday) CPU to be as much faster single-threaded over the 4790K's multi-threading, as the 4790K single-threaded is over the CPU it essentially replaced.  (Technically it replaced an Athlon 64 X2 4000+ as that was my CPU.  However, from March 2012 to January 2015, I was using a Core 2 Duo T7250 (that my dad bought later the same year as I bought the Athlon), as my socket AM2 motherboard had died & I wasn't able to replace it for a while.)

 

For example, in one test, my 4790K did 850 Multi-Threaded and 176 Single-Threaded in Cinebench R15.  My dad's T7250 did 95 Single-Threaded, and a HWBot submission for an Athlon 64 X2 4000+ running at its stock 2.1 GHz turned up a score of 79.

 

So ... my new CPU purchased in about 3 or 4 years would need to score at least 1894 in Cinebench R15 single-threaded.  (Math: 850*(176/79)=1893.67 - comparing 4790K vs Athlon, ignoring the Core 2 Duo.)

 

For multi-threaded performance, I want to be able to encode UHD 4K 2160p H.265 lossless in Handbrake at least at 60+ fps.  My 4790K only does about 0.5 fps.

 

 

I was just thinking of another jump to compare my desired CPU performace uplift to...

 

Back in January 1989, my dad bought a 286-10, paying $940 for case, board, CPU, 1 MB RAM, and 200W PSU.  (The CPU itself isn't listed separately on the invoice which we still have, but I'd guess it would have probably been around $300 or so.)  According to the Wikipedia Instructions per Second article, the 286-12 does 1.28 MIPS, so extrapolation would put the 286-10 at 1.067 MIPS.

 

In October 1995 (6 years, 9 months later), my dad bought a 486 DX4-120 for $102. (In this case the invoice is just for the CPU; other invoices including case, PSU, mobo & RAM add up to around $300 or so.)  The same Wiki article lists the DX4-100 at 70 MIPS.  Extrapolation puts the DX4-120 at 84 MIPS.

(One thing that might skew things a bit: the Wiki article mentions Intel DX4; my dad's CPU was an AMD.)

 

Performance uplift:  84 / 1.067 ~ 78.7x

Approximate price drop: 300 / 100 ~ 1/3

 

Price/Performance boost:  78.7 * 3 ~ 236.2x (that's 236 times, not 236 percent)

 

I bought the 4790K in January 2015, so the same time interval later would be October 2021, conveniently 3 months before my Corsair AX760's warranty expires.  It would be nice if my next CPU is about 235 times (not percent) faster, for the same price, vs my 4790K.

 

I don't think the i7-9700K or i9-9900K will have that big of a jump over the 4790K, so needless to say, I will NOT be doing that upgrade.

 

 

 

Also, by then, PCI Express 5.0 and DDR5 should be out.

 

DDR5 reminds me -- I'll also want lots of RAM support.  I'd like as big of a jump in RAM support as current system vs previous one.  For example, my laptop has 40GB, to be upgraded to 64GB as soon as it drops back down near what it was in Sept/Oct 2016.  My dad's old laptop had 2GB.

 

64 / 2 = 32 * 64 = 2 TB - the minimum amount of RAM my new board would need to support before considering registered ECC / LRDIMMS or more than 4 DIMM slots.

 

 

 

 

I was hoping the 9th-gen configuration would be like:

 

Core i7 = 8c/16t

Core i5 = 6c/12t

Core i3 = 4c/8t

Pentium = 4c/4t

Celeron = 2c/4t

 

Rather than the way rumors are saying (i9 = 8c/16t, i7 = 8c/8t, i5 = 6c/6t, i3 = 4c/4t).

 

 

 

I often hear people getting excited about a tiny 5-10% or even 25% performance jump year over year, or maybe 50% over 5 years or so.

 

Am I the only one here who, although it was in my childhood and I"m having to look up benchmarks after the fact, remembers performance jumps of like 50 to 100 *TIMES* over a previous system?

 

 

 

Oh ... btw a Tom's Hardware article from several years ago (Intel's 15 most unforgettable x86 CPUs) says the 286 was about 3.6 times faster than the 8086 at the same frequency.

 

If all x86-compatible CPU generations since then had the same performance-per-clock-per-thread uplift, how fast should an Intel Core 9000 series CPU be?

 

 

 

[Rupe]

Anxious to see if developers will start to really make use of the extra cores for gaming, as a 7th generation i7 owner I kinda hope not  although I know for the bigger picture it's obviously better if they do

[Tech Enthusiast]

7 hours ago, IAmAndre said:

I haven't compared the numbers but the fact that is R7 is $100 cheaper and can be used on a $100 cheaper motherboard can make AMD very competitive.

People in the market for a top of the line chip will not shy away from $100 usually.

If that all there is for AMD to stay competitive, I doubt it will be enough. They will have to show some improvements in those benchmarks.

The hype for more cores and less price is awesome, but it does also raise questions if Intel's lower core CPUs outdo AMD in multicore tests. Given this is just Geekbench, we can hope that Cinebench will draw a different picture. But I honestly dislike having to bank on hope. :-(

 

Unlike GPUs people don't update CPUs as much. Mainly due to much smaller jumps, but also due to usually having to buy new boards and ram.

Considering some people keep the same CPU for 3-8 years, spending an extra 100 bucks to get a better chip is worth it for most, right?

And those upgrading every year, will likely not do that to get a good deal, but to get better performance. So they will likely look for higher performance as well.

 

On the flip side, AMD has a pretty strong argument in not needing new Boards all the time. From what I hear this does not really work all that great though.

Sure, you can plug in a new AMD chip in an older Board, but missing features, less optimal support etc. are all things that seem to happen. This again raises the question: Will people in the market for the top of the line CPUs take those drawbacks, or just buy a new board anyways?

 

Not quite sure, but the market for "slightly cheaper, but also slightly less powerful" - CPUs is kinda slim in my mind. 

Just don't see who would pick that in the upper segments. This totally changes for budget builds, but the market for i7 and i9 is likely not the same market on a tight budget.

 

Basically, I am not so sure if the stuff that AMD does better than Intel really works for the upper segment. It is awesome for the low and mid-end due to better budget Options, but for the i7, i9 crowd budget does not matter as much, if at all. Can't speak for everyone obviously, but for me an added 100 bucks for a CPU that does score 20%+ higher is a no-brainer. I would not even consider the cheaper option, because 100 bucks are like 3 hours of work that will sit in my PC for at least 3-4 years (8 in the case of my wife, she really hates upgrading CPU lol).

[schwellmo92]

7 minutes ago, Tech Enthusiast said:

People in the market for a top of the line chip will not shy away from $100 usually.

If that all there is for AMD to stay competitive, I doubt it will be enough. They will have to show some improvements in those benchmarks.

The hype for more cores and less price is awesome, but it does also raise questions if Intel's lower core CPUs outdo AMD in multicore tests. Given this is just Geekbench, we can hope that Cinebench will draw a different picture. But I honestly dislike having to bank on hope. :-(

 

Unlike GPUs people don't update CPUs as much. Mainly due to much smaller jumps, but also due to usually having to buy new boards and ram.

Considering some people keep the same CPU for 3-8 years, spending an extra 100 bucks to get a better chip is worth it for most, right?

And those upgrading every year, will likely not do that to get a good deal, but to get better performance. So they will likely look for higher performance as well.

 

On the flip side, AMD has a pretty strong argument in not needing new Boards all the time. From what I hear this does not really work all that great though.

Sure, you can plug in a new AMD chip in an older Board, but missing features, less optimal support etc. are all things that seem to happen. This again raises the question: Will people in the market for the top of the line CPUs take those drawbacks, or just buy a new board anyways?

 

Not quite sure, but the market for "slightly cheaper, but also slightly less powerful" - CPUs is kinda slim in my mind. 

Just don't see who would pick that in the upper segments. This totally changes for budget builds, but the market for i7 and i9 is likely not the same market on a tight budget.

 

Basically, I am not so sure if the stuff that AMD does better than Intel really works for the upper segment. It is awesome for the low and mid-end due to better budget Options, but for the i7, i9 crowd budget does not matter as much, if at all. Can't speak for everyone obviously, but for me an added 100 bucks for a CPU that does score 20%+ higher is a no-brainer. I would not even consider the cheaper option, because 100 bucks are like 3 hours of work that will sit in my PC for at least 3-4 years (8 in the case of my wife, she really hates upgrading CPU lol).

You save money on the CPU, motherboard and guaranteed future socket/chipset support till at least next gen.

 

I would personally struggle to pick between the two camps when the 9th gen drops as Intel have addressed the few concerns I had (other than being a mega evil corp).

 

if you take a look at mindfactory the AMD and Intel sales have been very competitive, even in the high end of the market, so your personal opinion doesn’t reflect that of the masses.

[Tech Enthusiast]

3 hours ago, yian88 said:

Sorry but no HT means im buying AMD

1

Did you check how much HT is actually doing?

Not gonna lie, I was shocked at seeing no HT in an i7. Then I did some research on how much HT actually does for me and it boiled down to a 30% maximum performance, in very specific workloads that don't need the physical cores at 100%. So HT basically splits physical cores in half to make room for more parallelism. This does not increase raw horsepower at all, only splits the work a little more efficient IF and only IF the software you use actually does that. 

When I disable HT in my current system I don't see any degrading performance at all. Which kinda made me sad and happy at the same time.

 

Not trying to change your mind here, just giving food for thought as I was in the same boat. I thought HT was more than it actually is.

3 hours ago, yian88 said:

HT is just a matter or fliping a microcode switch on/off

1

Also a matter of security holes that open up as far as I gathered. So just having more real cores is not only faster, but also more secure. Basically, a 8c8t chip is faster than a 6c12t chip in all but the most extreme cases. 

 

Again, not trying to change your mind here, just saw your reasoning that was 100% in line with what I thought two weeks ago.

[GoldenLag]

2 minutes ago, Tech Enthusiast said:

So HT basically splits physical cores in half to make room for more parallelism. This does not increase raw horsepower at all, only splits the work a little more efficient IF and only IF the software you use actually does that. 

When I disable HT in my current system I don't see any degrading performance at all. Which kinda made me sad and happy at the same time.

it does actually. though the performance increase is minor and situated somewhere around ~5-20% depending on workload. there is something to gain from it. comparing a 8c/8t 9700k vs a 6c/12t 8700k in a workload that will do use it, the 9700k and 8700k will be very similar. 

5 minutes ago, Tech Enthusiast said:

Did you check how much HT is actually doing?

also ht/smt is very important in things like streaming where an r5 2600 is significantly better than the i5 8400. 

7 minutes ago, Tech Enthusiast said:

Also a matter of security holes that open up as far as I gathered. So just having more real cores is not only faster, but also more secure. Basically, a 8c8t chip is faster than a 6c12t chip in all but the most extreme cases. 

what security holes. i know ht can bring a couple, though i havent heard from them in a while. 

[Tech Enthusiast]

2 minutes ago, schwellmo92 said:

if you take a look at mindfactory the AMD and Intel sales have been very competitive, even in the high end of the market, so your personal opinion doesn’t reflect that of the masses.

1

I don't really care about sales on any single shop side. Mindfactory is kinda known for being visited by AMD fans and that is totally fine, just pretty useless to base any statistics on it. Also, I don't care what the market share is, but question if high-end seeking people care for cost savings as much. There is very little about High-End that I would call "cheap" or "must have". So spending more for more performance is an easy sell.

 

But sure, it is possible that I am totally wrong in my thinking. There are plenty of people that base their purchase on points I will never understand. Like brand loyalty. If I am in the market I buy whatever fits the performance I want, for the best price. If only one company offers said performance it is easy. If both offer it, it is a matter of looking up two prices.

Other simply seem to want red or blue, no matter what the other camp offers. I fail to see the benefit for the customer, but they do it anyways. /shrug

 

These i7 and i9 numbers are pretty damn impressive if confirmed. 

I certainly would not buy any currently available AMD chip over these, since I only care for performance and not for saving a few bucks on something I will use for years.

But it will be another 1-2 years before I am actually in the market for something new. Hope to see AMD hit the high-end as well by then. Not only with high core counts and low prices, but also with performance.

[Tech Enthusiast]

4 minutes ago, GoldenLag said:

it does actually. though the performance increase is minor and situated somewhere around ~5-20% depending on workload. there is something to gain from it. comparing a 8c/8t 9700k vs a 6c/12t 8700k in a workload that will do use it, the 9700k and 8700k will be very similar. 

3

That is what I said. It can be a gain. Though the gain is not based on raw horsepower, but down to software preferring threads over clockspeeds. Which is indeed pretty rare. So it comes down to wanting that performance all the time, or only in certain workloads. Unless you focus on those workloads and it actually scores higher (which I doubt), you would always go for the option that has the performance all the time, right?

4 minutes ago, GoldenLag said:

also ht/smt is very important in things like streaming where an r5 2600 is significantly better than the i5 8400. 

2

This is one of those workloads I guess. The question is: How many threads are enough to not slow down? As I said, for a quad core, it is no question at all. It needs HT. But 8+ cores is a different picture. Seeing Intel is doing this now, I hope to see some in-depth tests that go over this. All I am doing here is speculating. 

4 minutes ago, GoldenLag said:

what security holes. i know ht can bring a couple, though i havent heard from them in a while. 

Unsure, as I did not follow those security holes at all. Just saw some mentioning of HT is one of the security issues that are annoying to fix. So I kinda expected that to be a reason why Intel removed it on the i7 model.

[GoldenLag]

16 minutes ago, Tech Enthusiast said:

That is what I said. It can be a gain. Though the gain is not based on raw horsepower, but down to software preferring threads over clockspeeds. Which is indeed pretty rare. So it comes down to wanting that performance all the time, or only in certain workloads. Unless you focus on those workloads and it actually scores higher (which I doubt), you would always go for the option that has the performance all the time, right?

it is not only dependant on if the software preffers multiple threads, it also depends on things running in the background. one could allways go with the one that allways perform highest, but it also means a higher cost. a ht lower core will be cheaper than a high core no ht. i know YOU dont care about cost, but when comparing "equivelent" products it needs to be taken into an account. 

20 minutes ago, Tech Enthusiast said:

Unsure, as I did not follow those security holes at all. Just saw some mentioning of HT is one of the security issues that are annoying to fix. So I kinda expected that to be a reason why Intel removed it on the i7 model.

intel probably removed it to seperate skues. it seems they just moved everything up a tier when they strapped another 2 cores to the ringbus. it doesnt have anything to do with security. 

21 minutes ago, Tech Enthusiast said:

This is one of those workloads I guess. The question is: How many threads are enough to not slow down? As I said, for a quad core, it is no question at all. It needs HT. But 8+ cores is a different picture. Seeing Intel is doing this now, I hope to see some in-depth tests that go over this. All I am doing here is speculating. 

a lot of workloads and gaming are moving to multicore setups. games now allready use 6 cores/threads leaving the rest to manage OS and background tasks (that when it starts affecting performance), ht and smt negate some of this by doing 2 tasks simultatiosly. yes 8 cores will be nice, ut it not being ht is going to hurt it when things get to use 8 cores. (and as things are moving that seems to be likely). 

 

yes the 9900k is expencive to the point when you can grab a 12 core 24 thread instead. what is the better choice?

[Tech Enthusiast]

3 minutes ago, GoldenLag said:

yes the 9900k is expencive to the point when you can grab a 12 core 24 thread instead. what is the better choice?

1

The one having more performance I would say. According to Geekbench (yeah, I know), the i9-9700 almost equals 32 thread Threadripper 2950x. And that is with only 8 cores and no HT while being almost half the price and also doing A LOT better in single core.

It sure is hard to understand how that is possible tbh. But I would rather believe benchmarks showing numbers than people claiming "it will be better for [something] anyways!". People keep claiming stuff that is flat out false all over the place, even if facts are presented and available. So I kinda shy away from trusting people for stuff I don't fully understand (like CPUs). So Benchmarks > all for me in this case.

 

So to answer your question: I don't know. For me, it is an obvious pick, but others may choose differently based on their preferences. Some don't care about real performance and just want that price/performance beast, no matter which performance class it actually falls into.

[porina]

1 hour ago, Tech Enthusiast said:

Did you check how much HT is actually doing?

Not gonna lie, I was shocked at seeing no HT in an i7. Then I did some research on how much HT actually does for me and it boiled down to a 30% maximum performance, in very specific workloads that don't need the physical cores at 100%. So HT basically splits physical cores in half to make room for more parallelism. This does not increase raw horsepower at all, only splits the work a little more efficient IF and only IF the software you use actually does that. 

In applications, I've seen from 0 to 50%, and in one obscure synthetic benchmark a subscore increases by 70-80%. Cinebench R15 is around 30%. The AMD Ryzen Blender benchmark, with the software version at the time, was closer to 50%. Compute workloads similar to Prime95: 0%.

 

Personally I view HT (and SMT) and enabling more of the CPU to be used than it would otherwise. It does duplicate some parts to allow it to appear as another front end, but many of the resources are shared. Best gains are obtained when a single thread doesn't make good use of all the core's resources, the second thread could use some of what's left over.

 

The above is also why HT/SMT can appear to lower performance in some cases. If the critical path code were to run on a dedicated core, you get best possible performance for it. With HT/SMT on, anything else running on the core's other thread will slow it down.

[BuckGup]

13 hours ago, TetraSky said:

Interesting... Too bad they decided to do away with HT on i7.

Now they can charge 2X the price for HT on an i9

[schwellmo92]

1 hour ago, Tech Enthusiast said:

I don't really care about sales on any single shop side. Mindfactory is kinda known for being visited by AMD fans and that is totally fine, just pretty useless to base any statistics on it. Also, I don't care what the market share is, but question if high-end seeking people care for cost savings as much. There is very little about High-End that I would call "cheap" or "must have". So spending more for more performance is an easy sell.

 

But sure, it is possible that I am totally wrong in my thinking. There are plenty of people that base their purchase on points I will never understand. Like brand loyalty. If I am in the market I buy whatever fits the performance I want, for the best price. If only one company offers said performance it is easy. If both offer it, it is a matter of looking up two prices.

Other simply seem to want red or blue, no matter what the other camp offers. I fail to see the benefit for the customer, but they do it anyways. /shrug

 

These i7 and i9 numbers are pretty damn impressive if confirmed. 

I certainly would not buy any currently available AMD chip over these, since I only care for performance and not for saving a few bucks on something I will use for years.

But it will be another 1-2 years before I am actually in the market for something new. Hope to see AMD hit the high-end as well by then. Not only with high core counts and low prices, but also with performance.

So we're supposed to trust your "thinking" over actual statistics? A lot of people in the higher end of the market still care about value for money, and future performance (i.e. threads > outright core speed). The Intel equivalent might be 20% faster in non heavily-threaded workloads, but at more than 20% cost, and then in heavily threaded workloads it loses, so you would be paying more for a worse processor in certain scenarios. And then looking at gaming benchmarks on their own is also inaccurate, how often is the only thing you have running on your PC the game? For me and everyone I know, almost never. It's not as clear cat as you are making it sound.

[pas008]

11 minutes ago, schwellmo92 said:

So we're supposed to trust your "thinking" over actual statistics? A lot of people in the higher end of the market still care about value for money, and future performance (i.e. threads > outright core speed). The Intel equivalent might be 20% faster in non heavily-threaded workloads, but at more than 20% cost, and then in heavily threaded workloads it loses, so you would be paying more for a worse processor in certain scenarios. And then looking at gaming benchmarks on their own is also inaccurate, how often is the only thing you have running on your PC the game? For me and everyone I know, almost never. It's not as clear cat as you are making it sound.

I agree with this but we dont get these benchmarks anymore we use to get multitasking benches with different scenarios from each reviewer too

I know this is another area intel shined past bulldozer but nothing new on this for ryzen and current intel

[SupremeGOAT]

15 hours ago, JediFragger said:

Flog thyself 100 times as penance.

 

Me? My 8700k@5Ghz is gonna last a fucking long time and will probably be single digit frames between it and a 9900k at the same frequency at 1440p. I'm happy.

Nay, flog THYSELF fool! Thou canst not begin to perceive pre-eminence. Where once the 8700k, perchance, hath shone, the numbered days have then since grown. From INTEL, behold the only truth and tremble, for within the belly of the beast lies the one true path. MURDER! MURDER I say! Only death awaits the 8700k! Why dost thou not get hard for the 9900k? Perhaps... thou art? _ _ _?

 

 

AMD i was about to say "Perhaps... thou art? AMD?" yeah...

Edited September 10, 2018 by SupremeGOAT

[GoldenLag]

2 minutes ago, pas008 said:

I agree with this but we dont get these benchmarks anymore we use to get multitasking benches with different scenarios from each reviewer too

I know this is another area intel shined past bulldozer but nothing new on this for ryzen and current intel

Personally would love to see spmething for hardwareunboxed.

 

Something somewhat standardized

 

Example of things in the background while gaming:

1. Song playing using your chosen mediaplayer

2. Discord with a couple of servers.

 

3. A couple of webpages in chrome, varying from a wiki page to forum page.

 

Something lile this i would call fairly standard for the average user. Though non of these things consumer a lot of CPU power it could show a weakness in the windows sqedualler or in CPUs where all the threads are being utilized.

[pas008]

11 minutes ago, GoldenLag said:

Personally would love to see spmething for hardwareunboxed.

 

Something somewhat standardized

 

Example of things in the background while gaming:

1. Song playing using your chosen mediaplayer

2. Discord with a couple of servers.

 

3. A couple of webpages in chrome, varying from a wiki page to forum page.

 

Something lile this i would call fairly standard for the average user. Though non of these things consumer a lot of CPU power it could show a weakness in the windows sqedualler or in CPUs where all the threads are being utilized.

yes I agree

atleast something to have to compare

and not just a song too a movie

and with browser stuff youtube video but they would have to make sure this isnt being compromised from network or isp etc etc etc

and compressing files

just something with apples to apples comparison on multitasking with a better rough idea

believe they could create a batch file just to make them all open and run similar on each test too

 

[GoldenLag]

1 minute ago, pas008 said:

yes I agree

atleast something to have to compare

and not just a song too a movie

and with browser stuff youtube video but they would have to make sure this isnt being compromised from network or isp etc etc etc

and compressing files

just something with apples to apples comparison on multitasking with a better rough idea

believe they could create a batch file just to make them all open and run similar on each test too

 

Could be interesting to use a benchmarking tool that launches different .exe and have them run a repeated pass of set tasks in different windows. 

 

Would be reasonably easy to set up a script to do something, but to make it standardized would require a up to date program.

 

Not to mention we need to take into account hardware acceleration. 

[porina]

12 minutes ago, GoldenLag said:

3. A couple of webpages in chrome, varying from a wiki page to forum page.

This alone was becoming a problem for me on a 4c8t CPU, specifically when Chrome was playing Youtube content it uses a not insignificant amount of CPU. For more grindy game content (MMO) I do tend to have YouTube, or Amazon prime video on 2nd monitor. I brute forced my way out of it and my gaming and video playback are now separate systems.

 

The difficulty with having multiple loads is the repeatability of the scenarios. 

[pas008]

Just now, GoldenLag said:

 

Could be interesting to use a benchmarking tool that launches different .exe and have them run a repeated pass of set tasks in different windows. 

 

Would be reasonably easy to set up a script to do something, but to make it standardized would require a up to date program.

 

Not to mention we need to take into account hardware acceleration. 

can have it disabled one run and enabled another

 

I do miss multitasking reviews

havent seen them since sandy/ivy?

 

and work/office related shit too considering many multitask at work with email/word/spreadsheets/along with many others that are pretty much standard

[Misanthrope]

It's kind of impressive how Quickly has intel scaled from 4/8 chips to 8/8 and 8/16 chips.

 

Like, imagine how much fucking faster and accessible Nvidia GPUs would be if the Radeon part of AMD was able to stop eating glue with their toes and got their shit together.

[Silentprototipe]

Yet another thing I cant afford  

[suits]

I'm worried that the reason they soldered ther chip is because of thermals and lack of overclocking headroom. It's a very "Intel" thing to do. The 8700k and 7700k have had pretty high thermals with 4 and 6 cores, hopefully they didn't move to soldered to make 8 cores just "do-able".