Gadgetory

welcome back to harbor unbox today we are looking into revisiting the original 9900 doing a bit of a retest quite a few people over the last few weeks have been requesting that we do that so I put together a video outlining the issue and then handing over to you guys and asking if you think we should revisit the testing and if so how and basically the most popular voted option there was to retest the well retest the 9 ROK and 95 watt TDP limit and then compare those results to the results from our original review so that's what we're going to do in this video if you missed the previous video which led up to this one and outlines in a bit more detail why we're doing this test then you can check that out it's called do we need to re review the core I $9.99 or okay I'll provide a link for that in the video description anyway I won't go over all the details again but in short motherboard manufacturers are currently getting the blame for running the 90 100 K AB spec when in reality we strongly believe it's Intel who's cheating their own spec and pushing board partners to run the 99er okay at the default clock multiplier table rather than the official power spec whatever the case out of the box the 900k isn't running at the Intel spec it's essentially overclocked and this has caused power and thermal results to go through the roof so in today's retest we will be showing how the coronoid 9rk performs when adhering to the Intel specification and comparing that data to the current out-of-the-box experience doesn't really matter where you stand on this issue having a resource that shows how these configurations compare under the same conditions is useful information in our opinion for the unlimited testing the msi Meg Zee 390 godlike has been used and for the 95 watt limited testing I use the Isuzu ROG Maximus 11 hero I loaded up the extreme memory profile and opted to use Intel settings which enforces a 95 watt TDP limit so let's get into the results first up let's look at the Cinebench r15 multi-threaded scores previously we found the 9900 K breaking the tooth Point barrier however with the TDP limit in place the score is reduced by fourteen percent down to seventeen hundred and sixty three points and that place is at roughly on par with the core i7 78 20x and crucially this meant it was a few percent slower than the 2700 X already you might be getting a sense of why Intel is happy for board partners to run this CPU out of spec next up we have our blender short run test and here the TDP limited configuration can only short bursts up to 119 watts for roughly 10 seconds this means for half the test the 900 K is still close to fully Unleashed this is why we're only seeing a 9% reduction in performance but still a reasonable drop off but it's not the full story almost every professional looking to invest in our rendering rig will be running workloads that take much longer than 20 to 30 seconds generally we're talking well hours of rendering work well we saw a 9% reduction in the short run test here we're seeing a 14% reduction in the more realistic rendering workload that's a pretty big drop-off and it means the 900k is now only keeping pace with the much cheaper rise in 7:20 700x the corona benchmark runs for over a minute and here we see a 13% decrease in performance when power limited this means whereas the 99er ok it was 25% fast in the 2700 x1 allowed to run without a power limit with the 95 watt TDP enforced it's just 9% faster it's still faster but for those of you who aren't interested in overclocking and spending big on cooling that margin isn't really impressive particularly given the price here we see a 15% reduction performance for the 9900 can using the 95 watt limit and this minute was just 4% fast in the 2,700 X whereas we've found it to be 23 percent faster previously running the 7-zip compression test we see that performance isn't really impacted that heavily by the 95 watt limit a 3% decrease is almost not worth discussing so let's move on to the decompression results here the 99 her okay was 7% slower with the TDP limit enforced and that was enough to make it slower than the horizon 720 700 X so it's not a significant difference but the fact that the 99er okay now appears to be slow on the rise and 7 processor isn't a good look for Intel moving on Excel is the perfect example of a short workload at well under 10 seconds the 99er okay isn't really impacted by the official Intel spec and we see much the same performance with and without the TDP in place testing with handbrake we see another 14% reduction performance with the TDP limit enforced and this meant that the 9900 K was now just 4% faster than the 87 era K and 13% faster than the horizon 720 700 X that said the first and second generation rise and CPUs don't do that well with a VX workload so let's check the margins again with the h.264 test where we previously saw a 14% reduction performance running h.265 encoding we only see half that hit with a non AVX encoding workload using h.264 we saw a 7% reduction but given how much better Rison does here this meant the 900k was now slower than the 2700 x it also wasn't much faster than the 87 okay for those of you wondering the 99er okay with the TDP limit in place dropped down to an all core frequency of 4 gigahertz in the h.265 test but it did sustain 4.2 gigahertz in the h.264 test moving on here we see another example where the 95 watt TDP sees the 9 RK come in behind the 2700 X or bit by a small margin it's also interesting to note that in this Premiere Pro CC export the unlimited 99 hurricane configuration matched the 78 20x a 140 watt part on the same process I've said previously that the 9900 could have at least a 140 watt TDP rating and that does seem to fit with what we're seeing here the premier warp stabiliser test doesn't max out all caused all the time it's a pretty typical editing workload and here we see just a 6% reduction in performance with the TDP limit in place still that was enough to see the 9900 K come in behind the 8700 K okay so these results explain a lot previously we saw a few reviews that claimed the 19 on her okay consumed a less power than the 8700 K which doesn't really pass the common sense test and for good reason however if we test the 99er okay with a 95 watt TDP limit and leave the 8700 k without a TDP limit you get this that said even with the TDP limit in place the 87 or okay and at 6 cores won't be impacted nearly as much as the 9-yard okay and it's eight cause for things to remain relatively even the 99er okay would need a TDP limit of at least 125 watts anyway what we see here is a 31 percent decrease in total system consumption as all eight cores are wound down from 4.7 gigahertz to 4 gigahertz and as you can see that 15 percent reduction has a profound impact on system consumption as we're also taking a lot of voltage out of the chip at the lower clock speed in blender we see a 27% reduction in total system consumption and now the 9900 K looks like a mighty efficient CPU it was a few percent fast in the 2,700 X and here we see a reduced total system consumption by 12 percent previously it was 19 percent faster than 2700 X but it also pushed consumption 21 percent higher well given what we just saw from the total system power consumption results and these thermal numbers while shocking hot that's surprising using the Noctua NHD 15 and Corsair hydro h 100 i pro we found that nine 200k to hit temperatures in the mid 80s when fully Unleashed however using the 95 watt TDP spec the 1900s out at just 64 watts in our blender stress test and that figure was dropped to just 58 degrees with our open custom loop so when operating all cores at 4 gigahertz the 99 are ok he's as cool as a cucumber but at 4.7 gigahertz it turns the CPU socket into a fiery pit of melting silicon okay it's not that bad but it's pretty bloody hot in comparison okay so what about games well here we see when testing with Assassin's Creed Odyssey that there is a measurable performance hit in CPU intensive titles though even then it's only under unrealistic conditions so gaming at 1080p with a 90 X 20 atti so at 1080p we do see an 8% hit to frame time performance that margins reduced to three percent at 1440p playing GPU bound titles like Forza horizon 4 shows no impact and I expect this is how the 1080p results will look in most titles so please keep that in mind as we are mostly focusing on CPU bound games in this video fours are being the exception here we see a 7% hit to frame time performance when testing with hitman at turn EP that's said we still see a 6% hit at 1440p that's not until he hit the 4k resolution the margins evaporate interesting that we see no real impact in project cars too though this title is a bit odd in the sense that the 9900 K is so much faster than the 87 or ek I'm not exactly sure why this is but the game certainly doesn't require a cause but other sources have confirmed these margins so it's not some strange bug with my test system there's also no real margin to speak of when testing with Rainbow six siege any modern CPU running at around four gigahertz seems to work well here we do see a pretty hefty 15 percent performance hit in frame time performance at 1080p and shadow of the Tomb Raider though once you get to 1440p were almost entirely GPU bound then finally we have Star Wars Battlefront 2 and here we see a small performance drop-off at 1080p nothing extreme again by the Tommy at 1440p the margins close up to nothing okay so gamers need not worry about the 95 watt TDP spec it really doesn't make much difference either way and with a run at overclocked out of the box or with the 95 watt limit motherboards won't run or the 1900 K rather won't run at insane thermals or power consumption at least for gaming having said all that if you work on your PC as well as game and your work involves intensive CPU workouts they're about the only kind of workouts I seem to do these days then yeah it's a quite a different story in our long run blender workload the 1900 K temperatures really do go through the roof when not abiding by the 95 watt limit we saw an increase of about 20 degrees so pretty significant that and of course that being the case it was a similar story for power consumption that's obviously what led to the higher thermals and we saw a total system consumption climb by almost 40% so a lot more power being used there but of course it is running about 700 megahertz higher so yeah quite a difference there in clock speed but basically what this means is if the 1900 K was forced to run at Intel's TDP spec or their power spec whatever you want to call it and well abide by the power limits it would be a much more efficient eight core processor it actually be a mighty efficient eight core processor you get 2700 X light performance while saving a little 10% on power and the 2700 X is already quite an efficient processor this is actually a pretty big problem for Intel the 1900 was already a tough sell in the overclock configuration used by motherboard makers and well I suppose Intel knew that would be the case it's a $500 eight called desktop CPU campaigning with a $300 eight core desktop CPU and as we just saw with the 95 watt limit in place it's barely any faster than the rise in 727 hundreds in fact in some tests it was slower man that's a bloody awful result for a CPU that costs around 70% more so again that really is a big issue for Intel and as I say they've kind of painted themselves into a corner for the 1900 K to make an ounce of sense for anyone who isn't an extreme overclocker it really needs to run at around 70 degrees with a quality aftermarket cooler and for that the TDP can't really go any higher than about 105 watts which is the TDP rating for the 2700 X though Intel and AMD I arrived at their TDP ratings quite differently however even at 105 watts it's barely any faster than the uncapped 8700 cane it's only a whisker faster than the much cheaper 2700 X so you can't really have that with a core I name processor now for those of you wondering at 105 watts or a hundred and five watt limit that sees the knight RK sustain a clock speed of four point one to five gigahertz in our blender workload and it runs it up to 69 degrees using the course at age 100 I approach that's a hundred and fifty megahertz increase over the 95 watt TDP and that increase to the operating temperature by about five degrees basically the 99 or okay is I was going to say a really good overclocker suppose it's a good overclocker and if you could say it's a really good overclock if you invest big in cooling it is quite good it's certainly better than the 2700 X and that was the point I was working towards making the 2700 X is pretty much maxed out out of the box you can squeeze a little bit more out of it if you get a really good chip then there's a little bit more there to be had but most of the gains will come from tightening up the memory sometime because that really helps out the horizon processors yeah my point is you can't really overclock the snot out of the like you can with the Intel CPUs that said with motherboards technically overclocking the 9900 k2 ad the default clock multiplier table which sees 4.7 gigahertz as the old core for example under those conditions there really isn't that much left on the table for most 5 gigahertz will be the limit and going beyond that good luck keeping it cool requires a serious amount of time effort money spent on cooling and then probably a bit of risk involved with dealing it and if you really want to get serious sanding down the die so yeah not for the faint heart of that one so realistically you're talking up to maybe 6% extra left in at a 6% boost from a 5 gigahertz overclock that's really the most you're looking at there and that is that's over the unlimited results I shouldn't say so not over the 95 watt TDP but that's kind of irrelevant since none of the board's use that out of the box so not a huge amount of overclocking Headroom left with the way the 1900 K is coming technically overclocked out of the box for those of you planning on buying the 900 K because you want to take advantage of the fact that it is an unlocked part and by that I mean you plan overclocking it then you'll want to look at how much overclocking head room there is in reviews and whatnot and see how it performs there however believe it or not for at least half of our viewers that by these unlocked K type processors from Intel such as the 8700 K or the 99 her okay don't actually plan on overclocking them you might find that hard to believe but we've done some polls on the channel before and it's about 50 percent for overclocking 50 percent for not overclocking and even out of that overclocking bunch some of them admit to just enabling things like MC so enabling those kind of auto overclocking type features which is kind of what we're seeing here out of the box anyway with these ed 390 motherboards I've had a few people who've managed to buy the 1900 back and some of them have been trusted patreon members and for the most part they do appear very disappointed and the common theme that I'm seeing is that it has unmanageable thermals and that's certainly what we've found in our day 1 review and then they say enabling the Intel spec which avoids the thermal problems you then get a situation where the upgrade wasn't really worth it so I suppose you could come up with your own spec that's sort of somewhere in between there maybe a hundred and twenty watts keep it there but rather than letting it go up to a hundred and fifty plus so yeah I think Intel could do a much better job with these unlock parts of creating a spec that makes sense and just as a side note I guess you could blame the people who bought the 900 K for buying it and then finding out that it runs like a toaster on their motherboard at least with the sort of well the unlimited type conditions I suppose you can say but they would be basing their purchase on the results that we showed and most other reviewers showed so you have to admit the day one reviews the productivity results were quite impressive I know they didn't justify the price but yeah it did leave the 2700 X and the dust for quite a few tests so not a bad processor anyway moving forward I think the best course of action to do is stick mostly to the sort of typical out-of-the-box experience and if that means testing overclocked CPUs well that's what we'll do if there was just one or two motherboard brands or even one or two motherboards showing these overclocked or running the cpu out of spec something like the MSI god like for example then we would have redone our review completely because that wouldn't reflect the typical out-of-the-box experience but that's not the case as far as I can tell all of the Z 390 motherboards or just almost all of them run out of spec and the asou spawns do as soon as you load XMP and agree to the asou optimized settings so I think that's pretty much it for this one on a final note if we do find situations where we need to do alternate testing or a follow up like we've done this video then I'll create videos just like this one and hopefully that helps avoid some of the confusion anyway I think that's about enough for this one I hope you enjoyed the video if you did hit the like button subscribe for more content just like this we have a lot of really cool videos coming up on the channel over the next few days few weeks and probably the next few months that I can't see that far in advance just yet anyway thank you for watching I'm your host Steve I'll see you again next time