Discussion in 'Frontpage news' started by Hilbert Hagedoorn, Oct 5, 2017.
Socket pinout is different. Another site actually posted the pinouts of 270 and 370.
You really need to get it into your head that not everyone places as much emphasis as you do on price. For many enthusiasts, price is not the primary consideration.
I see you banging on in every thread about price/performance but this is an enthusiast PC hardware forum/site, not a used car lot.
I'm saying that a 1700X has a similar performance-per-dollar to the 8700K and therefore makes a good comparison. Meanwhile the 1600X does not make for a good comparison when considering performance-per-dollar. It also doesn't fit in terms of performance-per-watt. The article lists performance-per-dollar (which is great), but the 1600X is the main competing CPU that was compared to. What this does is it makes the 1600X look worse than it really is.
The review as a whole was unbiased, but if the 1600X is the highlighted competing product, that will just drive people's biases, because it (to me) doesn't make for a suitable competitor, overall.
Maybe you should consider lowering the resolution for processor benchmarks, these cards are such a bottleneck even at 1080p.
Thanks for the answer!
Both of them can be considered competitors to the 8700K, one competes on price (1700) and the other competes with the same number of cores (1600X). Just a question of perpective.
On coffeelake some of the unused pins of lga 1151 are now used as power pins, and 1 auxillary sense pin is no longer used, nothing that would prevent intercompatibility considering an overlocked 7700k will draw as much or more power than a stock 8700k. Only reason i can see intel blocking z270/z170 is to prevent people from blowing up budget boards with weak vrms, which is fair. z370 however, there is no excuse.
There is more to it. WCCF has the pinout in their review.
The 1700X competes on both price and performance. I don't understand the logic in just core/thread count being sufficient to make a product a direct comparison. Keep in mind that the 1600(X) was intended by AMD to compete against the 7600K. The 8700K compared to the 7600K has 3x the thread count, much higher clocks, and is marketed as a whole performance tier higher.
To reiterate, I'm not complaining about the 1600X being highlighted, I just think the 1700X is more apples to apples and should also be highlighted in the graphs.
ive looked at the pin out, there are no significant differences, beyond the vsense and extra power pins
one more pin change is AC38 and ab35, a socket detect pin (sktocc), but even so, the analogous pin on each platform is reserved accordingly, which either means they aren't used or have a function not specified, which leads me to believe either the sockets are intercompatible or they did a switcheroo inorder to prevent it on purpose.
would be hilarious if a lga 775 xeon sticker type adapter is possible
Oh yeah this is a very Interesting offering from Intel I like it Alot however....Zen being only a (few) fps behind I just cannot justify spending double the price for a few fps gain. I like the motherboards as well they seem to be way way better then Amd offerings ( typical )
So the top-dog 6/12 i7 will be $380?
the i5 6/6 Might be interesting buy if they can price it at around $200-$250.
Loving the battle between Amd and Intel haha!
seems like it performs closer to AMD 1700X- why so damn $$$
Wow, these 8700K really need delidding, the review shows that a wPrime1024 run when overclocked resulted in a max temperature of 89 degC while outputting a system Wattage of 147 Watts. My system strangely enough also outputs almost exactly the same system wattage (148W) when running wPrime1024 even though it's only a 6700K @4.7Ghz @1.4V - and my max temps are only 58 degC using a simple air heatpipe NH-D14 cooler at minimum fan RPM (using ULV adapter - silient effectively) - mine is delidded, so those 8700K's really need delidding because my CPU is pulling the same wattage but running 30 degC cooler. (My system idle wattage is also lower than the test system Hilbert used, so even plausible that my CPU is using even more Watts than the 8700K.). Intel gotta improve that paste or use solder for their CPUs - it's a real shame!
8700K seems like a good gaming CPU though, with quite a bit of future proofing.
£360 at Scan, for 4k gaming its not great value when the cheaper AMD chips are so close in performance. Even the 1440p scores arnt that great compared to the 1600x.
That's because of GPU bottleneck. What happens when faster graphic cards come out like Volta next year? Ryzen's "good enuff" won't be good enough.
Nothing happens, i will still have a 1080Ti therefore whatever else comes out is irrelevant, certainly at 4k as that will still be very GPU dependant for a few generations yet.
So the real conclusion is: if you are gaming at 1440p and above buy a Ryzen because it's cheaper and with the same performance?
Also for 1080p still buy a Ryzen unless you game at 144hz or higher?
I just wanted to point out that Hardware.fr, a European test/review site has posted extensive testing of all the new Intel CPUs, so clearly the Intel supply issue is not universal. Might be worth asking them if there is anything that needs to be done to resolve your issue.
Also, the CPU prices vary a lot with the US especially since now the dollar and Euro are no longer so similar in value (1 Euro = 1.18 dollars). In the US it is pretty normal to find the Ryzen 1600x for US$199, which is a huge difference with Europe. 229 Euros is US$270.
Edit: I just looked, and Amazon.fr has the 1600x for 204 Euros.
Volta will *not* come out next year.
you are looking at pascal until 2019. what you will get is a pascal refresh at 10-12nm which will improve power efficiency and potentially tweaking. they are working hard on cooling solutions (smaller processes have higher, more localized temps) and reworking volta to a more manageable yield from the wafers.
pascal at a smaller process also will put AMD to the test with the highly scalable navi processes... will navi at 7nm be able to handle the big pascal on a smaller node?
For gamers, the 8600K will likely be the best all-around option: not too expensive, great clock speeds, good latency, and sufficient threads for any modern game. I think a Ryzen 5/7 or a 8700K would have more longevity and will be better values, but the 8600K will probably be good enough for anyone who doesn't expect to keep their PC longer than several years.
In every test where the top-performing Intel chip reached 60FPS, Ryzen was also there to reach 60FPS. How is that not "good enough" at higher resolutions? Most of the time, 90FPS was achievable at 1440p. As long as you're not gaming at 120Hz+, Ryzen is perfectly fine for gamers and will remain that way for a while. I know Volta seems to be promising, but Nvidia doesn't have voodoo magic to bring all of these games to 120Hz+ on Intel.
What happens when 12nm Ryzen+ comes out in Q1 2018 with higher clock speeds? What happens when Intel's new lineup comes out later next year with eight cores? What happens when Ryzen 2 arrives in 2019 on a 7nm process and hitting 5GHz? If you keep thinking about the future, you'll end up never buying anything. There's always something better just around the corner, so you might as well just not buy anything and keep waiting. Forever.
I guess at least those buying a Ryzen build today know that their motherboard will be good for an upgrade path for the next few years, so they can just drop in Ryzen+ and Ryzen 2 when they arrive. I wonder if whatever chips end up in Z390 boards next year will work on Z370 ones?