AMD Ryzen 7 1800X CPU Review

[liquidaim]

how about a comparison of the power comparison between the new ryzen chips and intel's chips? this was standard practice in bulldozer/piledriver reviews. did intel ask you not to include that comparison?

 

[Elenelt]

Still, for a first stepping, Ryzen is very beefy CPU indeed
Next steppings absolutely should be even more 'baked'

 

[-Fran-]

Thanks for the review Toms Team!

I look forward for the updates or a potential second part to it. Nice set of benchies to see how good it really is.

My personal impression is "good, but...", since they need a lot of extra hertz to catch Intel in the gaming arena as things stand now (I stress that last part again: as they stand now) since Intel has a very generous advantage there with the mainstream market. I do hope the 4C and 6C parts will bump the clocks a bit so that disparity closes to reach parity at least.

Overclocking wise, I expect no miracles and I already feel a tad disapointed, but the CPUs are still good to send a straight forward message to Intel: "AMD is back into the game and prepare yourselves for Naples".

Cheers!

 

[madmatt30]

still cant see shelling out 300+$$$ for any AMD CPU . too easy to end up another let down right now for that kind of investment risk in them [opinion]

first release off just whats said in hype / promo reviews ?? look how well bulldozer was reviewed at first , same thing here , and only time will tell from real end users if the proof in in there pudding or the pie is in there faces . after Am3+ I'm pretty gun shy on this .

I do hope its all they claim if not just close to will be a plus for AMD , if 6 months or a year from now guys with these get that ''hay wait a minute ,WTF '' it will be bulldozer all over , Déjà vu .

None of this makes any sense at all to me.

 

[junkeymonkey]

''It will be interesting to see what their 6 and 4 core chips will do on the gaming front in the future, most likely great gaming value for the money''

??

thing is how much of that I on the GPU ? where it don't matter id its AMD or intel your running I could see that if the ryzen was not 350$$

even the 8350 or a i5 4670 non k with the same cards score the same but when you need core power like I my forst post that's where the money spent on intell pays off twice as fast

then here look at a 980ti and a 8350 that's overclock to death scores with a i5 4670 non ''k'' box stock ?

http://www.3dmark.com/search#/?mode=advanced&url=/proxycon/ajax/search/cpugpu/fs/P/1541/1033/500000?minScore=0&cpuName=AMD FX-8350&gpuName=NVIDIA GeForce GTX 980 Ti

http://www.3dmark.com/search#/?mode=advanced&url=/proxycon/ajax/search/cpugpu/fs/P/1614/1033/500000?minScore=0&cpuName=Intel Core i5-4670&gpuName=NVIDIA GeForce GTX 980 Ti

http://www.3dmark.com/search#/?mode=advanced&url=/proxycon/ajax/search/cpugpu/fs/P/1614/1085/500000?minScore=0&cpuName=Intel Core i5-4670&gpuName=NVIDIA GeForce GTX 1080

http://www.3dmark.com/search#/?mode=advanced&url=/proxycon/ajax/search/cpugpu/fs/P/1541/1085/500000?minScore=0&cpuName=AMD FX-8350&gpuName=NVIDIA GeForce GTX 1080


now with what I showed above on the PI runs and this would I buy a 8350 at $ 189 or that i5 at the time was $ 225-189 ?

wheres that AMD value ?? now there asking $350 for something that's unproven out side of prelease hype reviews ?? hard to be a first release sucker bait on this deal . then I guess that's why when AMD is refered to as all ways ''good enough for ''

a Yugo was ''good enough'' to got you in to town if you needed a car and so are your legs I guess ?? would not spend Cadillac $$ for a Yugo either


matt- ''None of this makes any sense at all to me. '

maybe because you fall for it all hook line and sinker , and are gullible ? then its your money not mine to spend . like they say '' a fool and there money are soon parted '' and I think a few are going to learn that here as did with AM3+

sorry , and good luck

not looking to worth it from here ?

http://www.pcworld.com/article/3176191/computers/ryzen-review-amd-is-back.html?page=3

 

[gdmaclew]

If the average power user (or non power user) uses only 2 sticks of RAM in their rig and hence uses only 2 channel memory then why did Tom's test all their Intel systems with 4 channel memory.
Why give Intel an unwarranted advantage over AMD?
I would like an answer to this point.

 

[madmatt30]

Im not seeing any kind of discrpancy over what I expected or in all honesty what AMD claimed.
IPC from those benches is anywhere between 40-55% over vishera ,so a 52% increase claim by AMD ,while a slight exhagguration maybe over a comparison of 10-15 different scenarios is within what they claimed.
Youve got the same IPC as haswell pretty much,its exactly what I expected ,Its enough so long as the prices are sensible.
The 7 series have little interest for me,theyre not a general purpose consumer chip ,definitely not a gaming chipimo but at that 50% pricepoint of a 16 threaded intel chip for 80% of the single core performance they are a viable option for people who actually have a need for that kind of use.

The 4c/8t,6c/12 thread are the ones that interest me & should interest 90% of the members on these forums.
Get thse 6 cores out at a locked i5/i7 pricepoint & there is going to be literally no reason at all to go for chips like the i5 7500 or i5 7700
vice versa the 4c/4t - If that comes in at an i3 pricepoint that should have intel very very worried indeed.

 

[BulkZerker]

"GDMACLEW" that's because the hasswell e supports quad channel. Its giving the ryzen a worst case scenario comparison. Or beat case for Intel's chip.

 

[FormatC]

how about a comparison of the power comparison between the new ryzen chips and intel's chips? this was standard practice in bulldozer/piledriver reviews. did intel ask you not to include that comparison?

Hey dude - I got the Ryzen only at TUESDAY afternoon (this week), made all workstation benchmarks (over 60) and measured the power and temps. For all this I had only 26 (in words t-w-e-n-t-y-s-i-x) hours (without sleep). What you are awaiting? AMD wasn't able to send me the CPU and mainboard last week. The power consumption of the pure CPU is very similar, depending at the clocks. If OC'ed, ist is higher than Intel, if not it depends at the job. I think to make a follow up, but for now it was simply to short.

And I had some issues with the two coolers I used, an Asus mainboard died and.... It was all other than perfect.

 

[madmatt30]

If the average power user (or non power user) uses only 2 sticks of RAM in their rig and hence uses only 2 channel memory then why did Tom's test all their Intel systems with 4 channel memory.
Why give Intel an unwarranted advantage over AMD?
I would like an answer to this point.

Those intel systems only support dual channel so there is no advantage at all in 4 sticks (the 7700k setup was only running 2 dimms anyway)

To me its fair comment using quad if a system supports quad anyway as a comparison even against a dual channel setup if thats all the platform supports - thats how 90% of power users would run it.

I dont really see any bias in the testing at all,probably one of the fairest toms reviews of amd hardware that Ive ever seen in all honesty.

 

[madmatt30]

matt- ''None of this makes any sense at all to me. '

maybe because you fall for it all hook line and sinker , and are gullible ? then its your money not mine to spend . like they say '' a fool and there money are soon parted '' and I think a few are going to learn that here as did with AM3+

sorry , and good luck

ME ?? Fell for nothing, had no intention of jumping on the day 1 bandwagon ,Ive been in the game too long, at all & as I said these 8c/16t chips have no interest for me at all.
That said I still dont find the results a disappointment at all - they are what were claimed & what I expected.

 

[InvalidError]

If the average power user (or non power user) uses only 2 sticks of RAM in their rig and hence uses only 2 channel memory then why did Tom's test all their Intel systems with 4 channel memory.

An i7-7700 (or any LGA115x CPU) only has two memory channels regardless of how many DIMMs you put in and using two DIMMs per channel makes very little performance difference on Intel's chips. Someone who buys an LGA2011 CPU which does have four memory channel would use four DIMMs since it makes no sense to cripple a $400+ CPU on a $200+ motherboard.

Ryzen on the other hand currently has very limited memory support which AMD is hoping it can fix with BIOS updates. Using two DIMMs per channels either doesn't work or substantially degrades stable memory timings and clocks. At the moment, it seems that all Ryzen chips being unlocked may be pointless since the chip has next to no overclocking heardroom on the CPU and memory controller. Hopefully this will improve with steppings and BIOS updates.

 

[FormatC]

I dont really see any bias in the testing at all,probably one of the fairest toms reviews of amd hardware that Ive ever seen in all honesty.

Thx. I see, you understood us.

For all others:

AMD sent all media a list with games, that AMD prefer to see. So what?
All media, followed this recommendaion can be called biased.

 

[ttt_2017]

I dont really see any bias in the testing at all,probably one of the fairest toms reviews of amd hardware that Ive ever seen in all honesty.

Thx. I see, you understood us.

For all others:

I need to know about the Quality of the motherboards and the chipset VS intel motherboard please ..... you said one Asus Motherboard died , I am little worried now about my order.

 

[madmatt30]

^ issue regarding aftermarket cooling mate , washers are required for the majority of aftermarket coolers - issue here in my eyes are whether this will fall down to the cooler manufacturers or the board manufacturers.
Its going to be a mass of confusion for a lot of novice builders who intend to dump the stock cooler & use aftermarket.

Early adopter ?? Stick with the supplied wraith cooler & stock clocks for the time being & you won't have an issue.

 

[bigdragon]

The results of this review tell me that Ryzen is not as good as Broadwell-E but also mostly better than Kaby Lake. This means AMD is competitive again. It's a viable option against Intel. The review doesn't give us the blowout AMD was hyping before release, but Ryzen is a good enough alternative for me to switch anyway. We desperately need competition in the CPU space.

 

[TJ Hooker]

''It will be interesting to see what their 6 and 4 core chips will do on the gaming front in the future, most likely great gaming value for the money''

??

thing is how much of that I on the GPU ? where it don't matter id its AMD or intel your running I could see that if the ryzen was not 350$$

6 and 4 core Ryzens are obviously going to cost less. 4C/4T Ryzen 3s are anticipated to cost $150 or less. 4C8T Ryzens for under $200.

 

[FormatC]

^ issue regarding aftermarket cooling mate , washers are required for the majority of aftermarket coolers - issue here in my eyes are whether this will fall down to the cooler manufacturers or the board manufacturers.
Its going to be a mass of confusion for a lot of novice builders who intend to dump the stock cooler & use aftermarket.
Early adopter ?? Stick with the supplied wraith cooler & stock clocks for the time being & you won't have an issue.

AMD wasn't able to send Wraith to media....

 

[madmatt30]

I dont really see any bias in the testing at all,probably one of the fairest toms reviews of amd hardware that Ive ever seen in all honesty.

Thx. I see, you understood us.

For all others:

I need to know about the Quality of the motherboards and the chipset VS intel motherboard please ..... you said one Asus Motherboard died , I am little worried now about my order.

^ issue regarding aftermarket cooling mate , washers are required for the majority of aftermarket coolers - issue here in my eyes are whether this will fall down to the cooler manufacturers or the board manufacturers.
Its going to be a mass of confusion for a lot of novice builders who intend to dump the stock cooler & use aftermarket.
Early adopter ?? Stick with the supplied wraith cooler & stock clocks for the time being & you won't have an issue.

AMD wasn't able to send Wraith to media....

yes mate I know,reply was aimed at tt-2017 above me who was concerned about a board dying.

I had the opportunity to buy the 1700 + asus prime 350 board as a spare media sample at a knockdown price without a cooler a week or so ago ,tempted but declined & am glad I did or It may have also been dead by now.

 

[InvalidError]

Early adopter ?? Stick with the supplied wraith cooler & stock clocks

If you buy a 1700X or 1800X, there is no supplied cooler, you have to source your own just like you do with Intel's K-chips.

 

[gdmaclew]

If the average power user (or non power user) uses only 2 sticks of RAM in their rig and hence uses only 2 channel memory then why did Tom's test all their Intel systems with 4 channel memory.
Why give Intel an unwarranted advantage over AMD?
I would like an answer to this point.

Those intel systems only support dual channel so there is no advantage at all in 4 sticks (the 7700k setup was only running 2 dimms anyway)

To me its fair comment using quad if a system supports quad anyway as a comparison even against a dual channel setup if thats all the platform supports - thats how 90% of power users would run it.

I dont really see any bias in the testing at all,probably one of the fairest toms reviews of amd hardware that Ive ever seen in all honesty.

I stand corrected.

 

[anbello262]

I dont really see any bias in the testing at all,probably one of the fairest toms reviews of amd hardware that Ive ever seen in all honesty.

Thx. I see, you understood us.

For all others:

AMD sent all media a list with games, that AMD prefer to see. So what?
All media, followed this recommendaion can be called biased.

I strongly feel that most other review sites are pretty biased, mainly because of the "AMD is back! Let's rejoice!" stuff. Being happy about something that was bad and has come back is no reason to ignore its shortcomings.
Just because you say "this works like this, is good at this and is bad at this" people complain that you are saying "it's bad", only because they feel AMD deserve a handicap for their comeback.

 

[jimmysmitty]

This appears to me to be a Intel-slanted review based on the reading of numerous other articles, and doesn't really go into detail about why the gaming benchmarks are so different. AMD deserves Kudos for taking the risk with Ryzen and achieving spectacular results. As AMD works with game mfgrs over the next few months to achieve the optimization Intel already has achieved in these games the differences will disappear.

I am not sure that catching up deserves kudos. When Intel launched Core 2, they didn't catch up to Athlon 64 they pounded Athlon 64. When Athlon 64 came out it pounded NetBurst. That deserves proper praise.

Catching up and matching does not merit that in my opinion.

Thanks for the review Toms Team!

I look forward for the updates or a potential second part to it. Nice set of benchies to see how good it really is.

My personal impression is "good, but...", since they need a lot of extra hertz to catch Intel in the gaming arena as things stand now (I stress that last part again: as they stand now) since Intel has a very generous advantage there with the mainstream market. I do hope the 4C and 6C parts will bump the clocks a bit so that disparity closes to reach parity at least.

Overclocking wise, I expect no miracles and I already feel a tad disapointed, but the CPUs are still good to send a straight forward message to Intel: "AMD is back into the game and prepare yourselves for Naples".

Cheers!

Hopefully AMD gets a better process to handle higher clocks.

 

[FormatC]

who was concerned about a board dying.

The board died and we don't know why. I sent it back to Asus for investigation. It doesn't react on power button.
But the other boards are working well, also Asus. No idea what happened.

After NDA it is allowed:
We got in the last days a lot of calls or mails, how to test Ryzen right (from their sight). Mostly with recommendations, to switch something on or off - or to use this and that. We decided to test all this possible options - seen in the gaming benchmarks. Paul does a good job to benchmark with all this suggestions and for all CPUs, not only for AMD. And to be honest: this method is not biased, it is soimply fair.

I personally like Ryzen, because it is a hard worker. But it is (as so often) too early. The amount of games with extensive CPU utilization on all possible threads is very, very small. On the one hand tey are pushing GPGPU and on the other side they are forcing multihreading for Gaming? This is perfect for MMOs an big maps with a lot of players. But if I had a wish: better torbo mode for one or two cores to compensate the disadvantage in non-scaling apps. XFR doesn't work for me as expected. I tried to cool down the CPU also with ice-water - not one MHz more.

 

[jimmysmitty]

I dont really see any bias in the testing at all,probably one of the fairest toms reviews of amd hardware that Ive ever seen in all honesty.

Thx. I see, you understood us.

For all others:

AMD sent all media a list with games, that AMD prefer to see. So what?
All media, followed this recommendaion can be called biased.

I strongly feel that most other review sites are pretty biased, mainly because of the "AMD is back! Let's rejoice!" stuff. Being happy about something that was bad and has come back is no reason to ignore its shortcomings.
Just because you say "this works like this, is good at this and is bad at this" people complain that you are saying "it's bad", only because they feel AMD deserve a handicap for their comeback.

That... is spot on.

The problem I see is people want reviews to be all sunshine and puppies but I would rather a review be honest and wiling to disclose the short comings.