Best price I have seen for this CPU yet, and from a reliable retailer.
Cost is €299.90 + €5 shipping.
Dispatched & Sold by Amazon. Nine left in stock at time of posting.
Paid for in € with a fee free card, including shipping it comes in at €305.47, which translates to £275-276 depending on the card used, and exchange rate when you check out.
The R7 1700 is almost the same cost, but comes with a cooler, so technically could be considered a better deal, but if you are not interested in Overclocking, and what the higher XFR speeds then this is a nice option.
43 comments
michaeljb
12 Sep 17#2
Good price, the 1700 costs 275 itself on UK Amazon so your getting a faster chip at same price, although worth noting the x models don't ship with coolers so you'll need to factor that in if purchasing
Uncommon.Sense to michaeljb
12 Sep 17#3
I already said all that at the bottom of the OP. :stuck_out_tongue:
michaeljb to Uncommon.Sense
12 Sep 17#6
I never read all that :blush: :raised_hand:
cr7hg
12 Sep 17#4
Is it really as good as Intel?
BetaRomeo to cr7hg
12 Sep 17#1
Not for gaming.
Otherwise, pretty much, yes.
Edit: apparently starting to match, see Uncommon.Sense's post here.
Uncommon.Sense to BetaRomeo
12 Sep 17#7
he missed "at 1080p 144 Hz" on the end of that. If you are at 1440p or above then it's pretty much matched these days, until GPU's get faster and stop being the bottle neck.
Nate1492 to Uncommon.Sense
12 Sep 17#9
Well, this isn't true.
It holds firm on some games, but on many other games, the Ryzen (even the 1800x OC'd to the teeth) gets destroyed by the I7 7700k.
Yes, not all the games..... but 30 FPS in a few games? And this is a non-OC'd i7 7700k
If you do rendering, there is a reasonable chance the Ryzen 1700 will be grand, but you'd know...
I just can't see how this price point should be anything but the 7700k or the 8700k in a month.... Unless you render.
I mean, sure, if you just prefer team Red and want to support them, fine, but most people are being led astray by comments like 'oh they are very close'.
That's not close, that's 30 difference using a Geforce 1080. That's bonkers.
Uncommon.Sense to Nate1492
12 Sep 17#10
I'm just gonna post a quote from the end of that Techpowerup review you linked with all those graphs from March, which say pretty much what I said
"Our second set of tests at 1440p shows performance losses getting smaller because the GPU can not drive such high framerates anymore; it is busy processing more pixels due to the higher resolution. This trend will continue at 4K resolution."
Here's a 16 game average graph, that will show you an AVERAGE, not a single game with issues from 6 months ago.
You always seem to get on the defensive when some posts a Ryzen deal, and start feeling the need to be aggressive towards anyone who suggests they may play games 'almost' as well as Intel in certain conditions.
Feel free to take a blanced viewpoint in future, or actually read what is written, anyone can use selective reporting to their benefit, in fact you could probably work for the Daily Mail with the posts you put up.
MysticalUndies to Uncommon.Sense
12 Sep 17#11
I would just like to state that "small" difference is an 18% performance increase on minimum framerate from the top rated ryzen in that graph to an i7. I think the performance difference between a 1070 and 1080 is around that figure. I personally wouldnt want an 18% performance defecit if I was building a gaming rig, my own opinion. But if you do loads of heavily multithreaded stuff it's worth the trade off.
Nate1492 to Uncommon.Sense
12 Sep 17#14
Defensive? Nah, just look at that graph and really consider what you are trying to say.
'SMT off' suggests that the CPU handles it's extra 8 logical cores poorly, since the performance is actually better without those 8 extra threads.
And really, you are just proving my point. Your graph shows the VERY best Ryzen CPUs 10% slower at max FPS and 18% slower at minimum FPS.
And that's without considerations for OC'd hardware, which the I7 outpaces the Ryzen 1800x quite well at.
Typical OC's for both look like: Ryzen 1800x base 3.6, top 4.0ghz. I7 7700k base 4.2, top 5.0ghz.
So, we start with 10% / 18% lead, and it only increases after OC. That's pretty signfiincant.
BetaRomeo to Uncommon.Sense
13 Sep 17#23
Err... I'm looking at this comment of yours, and not only does it clearly show the 7700K far, far ahead at 1440P (not "matched", as you claimed) at what amounts to a ~£30 price premium, but you haven't mentioned the 8700K coming out in three weeks that gives the best of both worlds - high IPC for gaming with 6c12t.
For pure gaming, the Ryzen 7 is still behind the 7700K, and that's not going to change for years, even at 1440P. You need the IPC to drive the throughput in the system. And the 7700K is not worth buying with the 8700K just days away.
Ryzen has been an amazing launch for AMD. For them to have such a great comeback, only really being behind Intel in one area (well, two if you count WinRAR - anyone here putting together a WinRAR build?), is a fantastic achievement. It's a little bit sad to see you cheapen their success with your oddly defensive comments and poor advice on your Ryzen submissions (and yes, I did notice you accuse someone else of being defensive in every Ryzen deal - pot, kettle, much? :D).
Edit: updated results here show even the lower-end 1600 almost "matching" the 7700K - although it's probably still worth waiting the 20 days to see what Coffee Lake offers!
Uncommon.Sense to BetaRomeo
13 Sep 17#19
OK, lets clear this up, I don't care what people buy, as long as they are not being fed duff/old/inaccurate info. However, it's always good to read what people write, and then process the words before ending up getting it all confused. My comment was "pretty much matched, these days" So, the key there is "pretty much" and "these days". If people insist on posting single cherry picked benchmarks/games from 6+ months ago, then they clearly didn't read the comment. If you look at a great deal of the newer information, be it from actual users, or newer reviews, the gap at 1440p across a large range of titles is now a great deal smaller than those old and out of date graphs would have you believe.
As for the 8700K, I don't see the relevance, it's not out until the 5th October, it's £360+ and since there are no actual reviews, guessing it's going to be good is hardly hard evidence. Now ironically saying that, I have an i5-8400 on order because I believe it will offer astounding performance for the money at £175 for 6c/6t but with really decent clock speeds, and may in fact be the undoing of the 8700K, in the value for money terms, alongside the 8600K, but that may be £270 by the look of things. My only annoyance is the lack of cheaper H370 chipset boards, but I'll grab a cheap ASRock Z370 Pro4 and use that to do my testing in.
There are certain people who don't understand what best value for money is, and always seem to try and divert people away to their preferred option, which i find very suspect and not helping anyone. Other than patting themselves on the back an making themselves feel better since they are stuck on one system, and can't possibly use a balanced viewpoint.
Either way, it looks like we'll have an interesting run up to the end of the year, I fully expect AMD to lower their pricing further on the R5/R7 line up, hopefully that will help some buyers offset the ever increasing DDR4 RAM prices. :smile:
BetaRomeo to Uncommon.Sense
13 Sep 17#22
I know exactly what you mean. We've had a few weirdos on this board - although I do miss Ferrari100, who had £20,000 in shares in AMD and basically spammed pro-AMD comments, accusing anyone correcting him of being part of the "Nvidia mafia"! More recently, the name escapes me, but there was one complete amateur here who tried to calculate gaming price/performance based on the price of individual components, instead of the build as a whole! There really are some people here who aren't experienced or technically qualified to comment - and yet do so anyway. I can't understand it myself - I don't go around giving people advice on which breed of dog to buy, or how to do yoga, because I don't know enough to publish meaningful comments on those subjects!
Maybe I'm mistaken... didn't you say in another deal that you're not buying CPUs for gaming, or something along those lines? Or am I confusing you for someone else?
I am intrigued - I've been trying to find decent updated comparisons between the R7 line and the 7700K, and not doing too well. The only updated benchmarks I've been able to find are a handful following Ryzen-specific patches (which naturally would lead to confirmation bias if those were the only ones considered - yes, I know that's a blindingly obvious point, I'm not trying to be condescending).
I would genuinely appreciate some links to those updated tests you're referencing?
(Although it begs the question, if you have those links, why did you share a picture of benchmarks from 6+ months ago in your comment which completely disproved the point you were writing!)
Maybe you don't have access to a calendar, but October 5th is three weeks away. I'd love to see your source for that "£360+" price, though. Guessing the price is hardly good evidence! :wink: Could it be you're confusing your pounds and euros?
Guessing how it will perform, though - we know the specs, we've had a performance leak, it's a fairly known quantity to anyone with an appropriate technical background. If Ryzen was the first CPU launch you've witnessed, then I see your mistake - Ryzen was a new architecture, and it was on a new manufacturing process, so there were many variables and potential surprises, but Coffee Lake isn't particularly special in CPU terms. And as I spent the couple of months before Ryzen's launch telling people to wait for Ryzen reviews if at all possible (heck, I probably made that comment dozens, if not hundreds of times!), it certainly seems fair now to tell gamers not to buy the currently-best-for-gaming-but-limited-lifespan 4c8t chip when we'll get reviews and availability of its 6c12t replacement within 22 days.
And I'm genuinely surprised that even you think that's irrelevant to anyone putting together a PC for gaming.
BetaRomeo to Uncommon.Sense
14 Sep 17#26
Still waiting for those links to "newer reviews" - unless you think that the other benchmarks you shared showing the 7700K 20% ahead of the 1800X in 1440P count as "matching"?
Uncommon.Sense to BetaRomeo
15 Sep 17#36
Sorry, I've not been paying attention as I had a Birthday party to organise and host!
I was just about to grab the links, when I found the following (brand new) video comparison. It is worth a watch, all the way through, rather than skipping and cherry picking parts of it to post like some people do. LINK HERE!
I think that actual quote of "pretty much matched these days" sums up exactly what I was saying.
Anyhow, I'm done here, deal is over and I am sure the same debate will happen in the next deal thread, be it for the 8700K/8600K.
Oh, and before i forget, no I don't really game that much these days, but I have quite a few folks who will ask me to build them systems, and my i5 8600 is for a test system, hoping to squeeze it into a 1U chassis, with quiet fans, not sure if it's going to succeed but I have to try it to find out! :smile:
BetaRomeo to Uncommon.Sense
15 Sep 17#38
Oh, I'm sorry, it just occurred to me that you might not see my comment if I didn't quote you!
Thanks again, those are exactly the kind of benchmarks I've been looking for month after month all year - hence my asking twice! :wink:
Uncommon.Sense to cr7hg
12 Sep 17#5
I am not sure if one processor can be as good as a whole company, but who knows these days!
Uncommon.Sense
12 Sep 17#8
Sellling fast, only three left now!
polly69
12 Sep 17#12
Ive had this since launch and its a monster, great price for such a powerful chip.
Valter84
12 Sep 17#13
Dammm, not in time... price updated to 349€. Shame...
Joshimitsu91
12 Sep 17#15
I don't think quoting an article from March is going to help anyone make a decision. The Ryzen platform was shaky at launch, which is precisely why they have "SMT off", for example.
Much better to consider more recent benchmarks if you are looking to buy now as some or all of these issues may have been resolved. Not saying they have, I haven't looked into it myself.
If I was buying (which I may be soon) then I would want to think beyond just gaming. How do these processors differ for other tasks? 10% less in games if I'm already achieving a good frame rate doesn't matter if my productivity is doubled by faster performance in other software. Plus you need to consider the trend towards multi-threaded applications and games which will surely only accelerate going forwards.
Another consideration is that AMD has supposedly committed to the AM4 platform for 4 years, meaning you could upgrade your processor in 2020 without the worry of buying a whole new motherboard and RAM (in theory). This is an exciting prospect for anyone concerned with the longevity of their system and not something that Intel are looking like they will offer any time soon. Of course the flip side to that is that you won't get any of the new motherboard technologies that may exist by that time if you stick with a 2017 board.
Nate1492 to Joshimitsu91
12 Sep 17#16
If you have found improvements since the March article, then please *link them*. Do not assume that AMD have magically fixed their issues and wave it off with 'just google it' because the thing is...
I have 'just googled it' and I have *NOT* found improved results. The fact of the matter is, if results don't change, people don't tend to post new, duplicate, results over the course of many months.
If they existed, WCCFTECH would be the best place to find positive AMD news. They are very eager to report on anything AMD related.
As I've also stated/pointed out, a lot of productivity favors single threaded speed.
Web browsing, Outlook, Excel, Office, CS Photoshop, and *MOST* productivity software relies on single threaded speed.
There are very few productivity tasks that involve multi threaded speed boosts.
If you do rendering as a high ratio of your tasks (say 20-30% or higher) then Ryzen is a great option.
I have personally rendered less than 5 times in 5 years. I've talked to colleagues, friends, gamers, work mates... No one that I talk to does any rendering. I find it crazy at how often rendering is brought up, compared to how often people *ACTUALLY* render.
Anyway, there is so much 'future potential' discussion, but this is the same stuff that came up with Bulldozer. It hasn't happened because effectively making an application that scales remotely well with more than 4 cores is *very* difficult.
Joshimitsu91 to Nate1492
13 Sep 17#17
If you have Googled it then feel free to *link* the recent results you found. My point was simply that posting results from March, soon after launch when by the articles own admission there are known issues with the new platform is worthless to a purchasing decision made in September.
I'm sorry but talking about web browsing and office applications in this discussion is frankly hilarious, these applications are not going to tax either AMD or Intel CPUs.
Rendering is just one example of a multi-threaded workload. What about compiling software (while still using the PC) or running virtual machines? Whether your social circle does a lot of rendering or not doesn't add anything to the discussion.
I also notice you simply miss my entire last paragraph which is a major selling point for anyone conscious of their spending.
Honestly as OP has said seems like you are just a blind AMD hater or Intel fanboy.
Nate1492 to Joshimitsu91
13 Sep 17#21
1) I have googled for 'newer' results and as I have already stated, I cannot find any. I drew the conclusion that if there are no changes, there is no reason for a new review. You find it, they don't exist. I've stated this twice now.
2) The 'known issues' haven't been fixed. I do not see any updated SMT results on google showing that it 'works now'. I can't link something that does not exist. Three times now.
3) Compiling software is *EXTREMELY* niche and very infrequently done, even by people who code for a living. I do not understand why you dismiss single threaded application performance, such as Excel, Web Browsing, Office, and various other common applications. If you use something WAY more frequently (Web Browsing) incremental, numerous, small savings add up.
4) Virtual Machines are a positive for Ryzen, I agree. Again, if you were running multiple VMs, you would *know this already and consider it*. There is no way you would have to tell someone who runs VMs that more cores is good.
5) There is no evidence to suggest the AM4 line will last 4 years. The history of AMD socket changes suggest changes at a much more frequent pace. From here: en.wikipedia.org/wik…ket
When AMD was competitive in the CPU game, they released much more frequent socket updates. Do we buy a random internet comment that AM4 will last 4+ years? I don't. Anyway, this portion is entirely speculation. I don't really trust them to follow up, and why would I trust any timescales from AMD?
CoeK to Nate1492
14 Sep 17#29
That is a lot of sockets and dates but isn't the progression for most people using AMD really just AM2+ to AM3 to AM3+ to AM4?
He probably said single threaded applications like excel etc don't really matter because there isn't much in it in real life, as far as excel which I use everyday it can't really get any faster for me or anyone else in my office.
Nate1492 to CoeK
14 Sep 17#30
What? How are you defending someone who said something provably wrong? He didn't say they didn't matter, he said it was 'slower'.
Also, please, list the applications you use on a daily basis. Just go, see what you do per day.
My guess is 'Rendering 2+ hours of video' isn't one of your use cases.
CoeK to Nate1492
14 Sep 17#31
No rendering isn't one of my use cases.
I think you are confused about what comment I am referring to, I'm referring to someone saying:
"I'm sorry but talking about web browsing and office applications in this discussion is frankly hilarious, these applications are not going to tax either AMD or Intel CPUs."
You said you don't understand why he dismissed this. I am saying it was probably dismissed because to go from calculating in excel in .0002 seconds to doing it in .0001 seconds is pretty much irrelevant. I don't actually know anyone that only deals with excel anyway, I generally have at least 5 applications on the go at once, none of them will tax a cpu enough to make single thread performance matter.
You ignored what I said about sockets btw.
Nate1492 to CoeK
15 Sep 17#34
The links showed a much longer response time in excel. You are pretty much just handwaving at the idea that it could be faster...
And about the cores, what's the difference? Sure, AM1, AM2, AM2+, AM3, AM3+, AM4.
If you want.
And did they mean when they first discussed the concept? Who knows, I don't trust AMD's timelines.
CoeK to Nate1492
15 Sep 17#35
I'm telling you in real life excel performs calculations as close to instant as it is possible to perceive. So who cares which one is closer to instant?
So AM3+ out in 2011 to AM4 out in 2017 is 6 years, it is actually closer to 6.5. You don't have to trust their timelines and fair enough if you don't but adding in various sockets from all over the place doesn't really come across as trustworthy either. You can make your point with the gap between AM3 and AM3+ being only 2 years for example.
MRP
13 Sep 17#18
What is it with these oddly viral marketing/ultra fanboy posters defending high priced Intel chips.
This is hotukdeals not Coldestukdeals.
The cores are similar. Ryzen offers more of these cores for less. Unlocked. Intel have faster cores in certain K chips through higher frequencies but these also cost much much more for what they are. The 7700k is in its own ultra high priced quad core category of £310. For a quad core, in 2017. In the month a 6 core replacement by intel is also due...
Reviewers cant recommend the Core i5 or i7 as most fall below Ryzen in single core too. When you go to higher clock speeds the prices increase far too much. The biggest issue for intel is pricing.
Worth pointing out the i9 is £1000 for 10 cores. This AMD CPU is a total bargain in comparison.
Nate1492 to MRP
13 Sep 17#20
This is a straight up piece of untruth.
MRP to Nate1492
13 Sep 17#24
I am not going to waste time arguing with you. You seem to link reviews that dont support your 'opinion' yet you link them anyway. You constantly make apples to oranges comparisons.
Look at the prices and products. Why the hell are cheap Ryzen 3's for example being mentioned alongside the 1080 ti and 7700k's?
Nate1492 to MRP
13 Sep 17#25
You said Ryzen beats the I7 in single threaded performance. This is an outright lie that no one is suggesting is true, except you.
MRP to Nate1492
14 Sep 17#27
Again. I never claimed as such. You said a pentium was better.
Nate1492 to MRP
14 Sep 17#28
Read your post man! This is, word for word, what YOU typed. You claimed exactly this:
"Reviewers cant recommend the Core i5 or i7 as most fall below Ryzen in single core too." From this comment
There is absolutely nothing unclear about that sentence. And it is wrong.
MRP to Nate1492
14 Sep 17#32
Wow. Someone's comprehension skills are lacking. Note the word 'most'.
Nate1492 to MRP
15 Sep 17#33
The word "most" is still an outright lie. There is no saving grace here. You either fall on your sword, or you pretend you weren't wrong.
MRP to Nate1492
15 Sep 17#39
I wouldn't go around recommending the expensive lower clocked and locked chips. Which is what most of Intel's portfolio is. From what we are hearing about Coffe Lake that will continue.
You you will notice reviewers and benchmarkers focus on other elements besides 7700k single core performance.
BetaRomeo
15 Sep 17#37
Ooh, a brand-new article, and I hadn't seen it - thanks! (Here's a proper link for people who don't need things to be read aloud to them: techspot.com/rev…80/ )
It certainly does show the Ryzen 1600 to be roughly matching the 7700K! While they did mention that a couple of the 1440P results were "inexplicable" wins for the 1600, that doesn't really take away from an excellent overall result:
The testbeds look fine, too (4.0GHz for a 1600 isn't much more optimistic than 4.9GHz for a 7700K, and both used sensible DDR4-3200). Brilliant stuff - I can absolutely see why you said "matched" now!
It's still worth noting the reason that 1080P is primarily used for CPU benchmarks - it's a better indication of what future performance will be like with a GPU upgrade.
But a 1600 system would come in ~£100 cheaper than a 7700K. As part of a ~£1000 system, one would be spending an extra 10% to get an initial 2-8% performance boost depending on resolution and card (based purely on the figures in that article). Obviously gamers expect diminishing returns on increased performance (otherwise we'd all have 1050Ti GPUs!), but it does knock a serious dent in the 7700K's advantage now that it no longer has the maths behind it.
Just a shame that these figures have come up just 20 days out from Coffee Lake, when it's simply bad advice to tell anyone to buy a CPU today. Still, I've got two gaming builds waiting for the Coffee Lake launch, and with more results like these it looks like one of them may end up Ryzen after all!
Nate1492 to BetaRomeo
15 Sep 17#40
These results, on the face of it, look good and 'matched' until you realize all the reviewer did was remove the worst performing AMD game from the list.
In his first article he says: "Some readers will undoubtedly declare that I'm biased towards a certain
company, but if that was the case, there's no way you would see Deus Ex:
Mankind Divided and Gears of War 4 in this list. Meanwhile, if I were
biased towards Intel, I'd drop Total War: Warhammer and Civilization IV."
So, in the re-review, he drops Gears of War 4. Guess what?
Also, this reviewer came to this conclusion back in July.
The dude massages his results big time, look at the variable lists of games.
Why didn't he use the same 30 games this time?
BetaRomeo to Nate1492
15 Sep 17#42
I take your point about the slight leaning of the choice of games, but in my opinion it really is very slight. It's a shame that Fallout 4 wasn't in the mix, as it shows scaling with both cores and IPC, but at the end of the day, no reviewer can cover every game, and no choice of games is going to satisfy everybody.
Results like these from a non-disreputable site like Techspot certainly justify Uncommon.Sense's use of the word "matched", at any rate.
And, of course, it's all pretty much moot. No gamer would seriously buy a CPU twenty days out from a significant CPU launch (and it does speak to the integrity of people advising them to do so), and so in a couple of weeks we'll have Coffee Lake reviews from multiple reputable sites covering plenty of games - including updated Ryzen results, which will either confirm or undermine these results today. So it's really no problem either way.
Uncommon.Sense
15 Sep 17#41
No, you just can't read.
"Please note that because there was a large difference in performance between running Deus Ex: Mankind Divided with DX11 and DX12 with the Ryzen CPU, I included both the results in this graph. However, I only included DX12 results for Total War: Warhammer and not the DX11 results since the R5 1600 was much faster in both tests. I favored DX12 as it's the newer API."
So in fact he was giving Intel a free boost, by excluding one of the results, then including both DX11//DX12 results for Deus Ex.
Also, game lists will change, as the new games that come out over time change, is 20 games not enough to create an average? How many other reviews do more that 8-10?
Ask him why he didn't use 30 games (tweet him)if you actually care that much, 'cause you seem obsessed with trying to nit pick. I am sure it will have something to do with testing two different graphics cards, on two different platforms, with twenty times, with three runs each, can you imagine how much longer that would take with thirty? What would it add, probably 1-05.-1% difference if that, since some of the games would have favoured the Ryzen, I am sure much to your disappointment.
Nate1492 to Uncommon.Sense
16 Sep 17#43
The problem arises when you 'revisit' a result and directly compare it.
The guy linked to his 3 month old review where he picked a set of games, and justified his game set there and showed he wasn't biased by including 'bad games for each platform'.
Then he re-reviews it and removes the worst game on the Ryzen test bed, Gears of War 4.
He's chosen a bunch of GPU bound games that show the Ryzen in a favorable light by averaging the performance of '20 games' but only really comparing 6 games that do anything with the CPU.
Who really thinks adding DOOM's results of 200/200 for literally every processor is beneficial?
I don't see 'TechSpot' as a beacon of testing integrity, this guy links to 'AdoredTV' as some trusted source. AdoredTV might as well just be a giant paid advertisement for all things AMD.
This list of games just looks cherry picked to allow the conclusion to be 'AMD has closed the gap, #finewine' when we're not focusing in on the fact that many titles are still 20% slower.
Because how on earth does this only suggest that Ryzen is '5% slower'?
Like really, we've got some pretty old games on this list, yet GOW4 is off? I think this graph would just look bad if you plopped in GOW4 being 30% slower. Where is FO4? GTA5?
You can easily massage results, and trying to suggest these games 'take a long time to test' is just outright untrue.
This guy is really good, from GamerNexus.
Anyway, techpowerup does a great job, and they don't adjust their testbed without strong thought. The idea that the set of games is randomly chosen for each review should absolutely flag techspot up in your mind as something weird.
![Fresh Breakfast Meat Box (Using code) @ Campbells Meat [Haggis / Lorne Sausage / Bacon / Black Pudding & More]](/images/topics/208684_1.jpg)
![[Xbox One] Q. u. b. e: Director's Cut on Deals with Gold](/images/topics/208716_1.jpg)
Opening post
Cost is €299.90 + €5 shipping.
The R7 1700 is almost the same cost, but comes with a cooler, so technically could be considered a better deal, but if you are not interested in Overclocking, and what the higher XFR speeds then this is a nice option.
43 comments
Otherwise, pretty much, yes.
Edit: apparently starting to match, see Uncommon.Sense's post here.
It holds firm on some games, but on many other games, the Ryzen (even the 1800x OC'd to the teeth) gets destroyed by the I7 7700k.
Yes, not all the games..... but 30 FPS in a few games? And this is a non-OC'd i7 7700k
techpowerup.com/rev…tml
techpowerup.com/rev…tml
techpowerup.com/rev…tml
Also, let's not forget, most programs you run are single threaded beasts.
techpowerup.com/rev…tml
If you do rendering, there is a reasonable chance the Ryzen 1700 will be grand, but you'd know...
I just can't see how this price point should be anything but the 7700k or the 8700k in a month.... Unless you render.
I mean, sure, if you just prefer team Red and want to support them, fine, but most people are being led astray by comments like 'oh they are very close'.
imgur.com/a2C…I7p
That's not close, that's 30 difference using a Geforce 1080. That's bonkers.
"Our second set of tests at 1440p shows performance losses getting smaller because the GPU can not drive such high framerates anymore; it is busy processing more pixels due to the higher resolution. This trend will continue at 4K resolution."
Here's a 16 game average graph, that will show you an AVERAGE, not a single game with issues from 6 months ago.
You always seem to get on the defensive when some posts a Ryzen deal, and start feeling the need to be aggressive towards anyone who suggests they may play games 'almost' as well as Intel in certain conditions.
Feel free to take a blanced viewpoint in future, or actually read what is written, anyone can use selective reporting to their benefit, in fact you could probably work for the Daily Mail with the posts you put up.
'SMT off' suggests that the CPU handles it's extra 8 logical cores poorly, since the performance is actually better without those 8 extra threads.
And really, you are just proving my point. Your graph shows the VERY best Ryzen CPUs 10% slower at max FPS and 18% slower at minimum FPS.
And that's without considerations for OC'd hardware, which the I7 outpaces the Ryzen 1800x quite well at.
Typical OC's for both look like: Ryzen 1800x base 3.6, top 4.0ghz. I7 7700k base 4.2, top 5.0ghz.
So, we start with 10% / 18% lead, and it only increases after OC. That's pretty signfiincant.
For pure gaming, the Ryzen 7 is still behind the 7700K, and that's not going to change for years, even at 1440P. You need the IPC to drive the throughput in the system. And the 7700K is not worth buying with the 8700K just days away.
Ryzen has been an amazing launch for AMD. For them to have such a great comeback, only really being behind Intel in one area (well, two if you count WinRAR - anyone here putting together a WinRAR build?), is a fantastic achievement. It's a little bit sad to see you cheapen their success with your oddly defensive comments and poor advice on your Ryzen submissions (and yes, I did notice you accuse someone else of being defensive in every Ryzen deal - pot, kettle, much? :D).
Edit: updated results here show even the lower-end 1600 almost "matching" the 7700K - although it's probably still worth waiting the 20 days to see what Coffee Lake offers!
As for the 8700K, I don't see the relevance, it's not out until the 5th October, it's £360+ and since there are no actual reviews, guessing it's going to be good is hardly hard evidence. Now ironically saying that, I have an i5-8400 on order because I believe it will offer astounding performance for the money at £175 for 6c/6t but with really decent clock speeds, and may in fact be the undoing of the 8700K, in the value for money terms, alongside the 8600K, but that may be £270 by the look of things. My only annoyance is the lack of cheaper H370 chipset boards, but I'll grab a cheap ASRock Z370 Pro4 and use that to do my testing in.
There are certain people who don't understand what best value for money is, and always seem to try and divert people away to their preferred option, which i find very suspect and not helping anyone. Other than patting themselves on the back an making themselves feel better since they are stuck on one system, and can't possibly use a balanced viewpoint.
Either way, it looks like we'll have an interesting run up to the end of the year, I fully expect AMD to lower their pricing further on the R5/R7 line up, hopefully that will help some buyers offset the ever increasing DDR4 RAM prices. :smile:
Maybe I'm mistaken... didn't you say in another deal that you're not buying CPUs for gaming, or something along those lines? Or am I confusing you for someone else?
I am intrigued - I've been trying to find decent updated comparisons between the R7 line and the 7700K, and not doing too well. The only updated benchmarks I've been able to find are a handful following Ryzen-specific patches (which naturally would lead to confirmation bias if those were the only ones considered - yes, I know that's a blindingly obvious point, I'm not trying to be condescending).
I would genuinely appreciate some links to those updated tests you're referencing?
(Although it begs the question, if you have those links, why did you share a picture of benchmarks from 6+ months ago in your comment which completely disproved the point you were writing!)
Maybe you don't have access to a calendar, but October 5th is three weeks away. I'd love to see your source for that "£360+" price, though. Guessing the price is hardly good evidence! :wink: Could it be you're confusing your pounds and euros?
Guessing how it will perform, though - we know the specs, we've had a performance leak, it's a fairly known quantity to anyone with an appropriate technical background. If Ryzen was the first CPU launch you've witnessed, then I see your mistake - Ryzen was a new architecture, and it was on a new manufacturing process, so there were many variables and potential surprises, but Coffee Lake isn't particularly special in CPU terms. And as I spent the couple of months before Ryzen's launch telling people to wait for Ryzen reviews if at all possible (heck, I probably made that comment dozens, if not hundreds of times!), it certainly seems fair now to tell gamers not to buy the currently-best-for-gaming-but-limited-lifespan 4c8t chip when we'll get reviews and availability of its 6c12t replacement within 22 days.
And I'm genuinely surprised that even you think that's irrelevant to anyone putting together a PC for gaming.
I was just about to grab the links, when I found the following (brand new) video comparison. It is worth a watch, all the way through, rather than skipping and cherry picking parts of it to post like some people do. LINK HERE!
I think that actual quote of "pretty much matched these days" sums up exactly what I was saying.
Anyhow, I'm done here, deal is over and I am sure the same debate will happen in the next deal thread, be it for the 8700K/8600K.
Oh, and before i forget, no I don't really game that much these days, but I have quite a few folks who will ask me to build them systems, and my i5 8600 is for a test system, hoping to squeeze it into a 1U chassis, with quiet fans, not sure if it's going to succeed but I have to try it to find out! :smile:
Thanks again, those are exactly the kind of benchmarks I've been looking for month after month all year - hence my asking twice! :wink:
Much better to consider more recent benchmarks if you are looking to buy now as some or all of these issues may have been resolved. Not saying they have, I haven't looked into it myself.
If I was buying (which I may be soon) then I would want to think beyond just gaming. How do these processors differ for other tasks? 10% less in games if I'm already achieving a good frame rate doesn't matter if my productivity is doubled by faster performance in other software. Plus you need to consider the trend towards multi-threaded applications and games which will surely only accelerate going forwards.
Another consideration is that AMD has supposedly committed to the AM4 platform for 4 years, meaning you could upgrade your processor in 2020 without the worry of buying a whole new motherboard and RAM (in theory). This is an exciting prospect for anyone concerned with the longevity of their system and not something that Intel are looking like they will offer any time soon. Of course the flip side to that is that you won't get any of the new motherboard technologies that may exist by that time if you stick with a 2017 board.
I have 'just googled it' and I have *NOT* found improved results. The fact of the matter is, if results don't change, people don't tend to post new, duplicate, results over the course of many months.
If they existed, WCCFTECH would be the best place to find positive AMD news. They are very eager to report on anything AMD related.
As I've also stated/pointed out, a lot of productivity favors single threaded speed.
Web browsing, Outlook, Excel, Office, CS Photoshop, and *MOST* productivity software relies on single threaded speed.
There are very few productivity tasks that involve multi threaded speed boosts.
If you do rendering as a high ratio of your tasks (say 20-30% or higher) then Ryzen is a great option.
I have personally rendered less than 5 times in 5 years. I've talked to colleagues, friends, gamers, work mates... No one that I talk to does any rendering. I find it crazy at how often rendering is brought up, compared to how often people *ACTUALLY* render.
Anyway, there is so much 'future potential' discussion, but this is the same stuff that came up with Bulldozer. It hasn't happened because effectively making an application that scales remotely well with more than 4 cores is *very* difficult.
I'm sorry but talking about web browsing and office applications in this discussion is frankly hilarious, these applications are not going to tax either AMD or Intel CPUs.
Rendering is just one example of a multi-threaded workload. What about compiling software (while still using the PC) or running virtual machines? Whether your social circle does a lot of rendering or not doesn't add anything to the discussion.
I also notice you simply miss my entire last paragraph which is a major selling point for anyone conscious of their spending.
Honestly as OP has said seems like you are just a blind AMD hater or Intel fanboy.
2) The 'known issues' haven't been fixed. I do not see any updated SMT results on google showing that it 'works now'. I can't link something that does not exist. Three times now.
3) Compiling software is *EXTREMELY* niche and very infrequently done, even by people who code for a living. I do not understand why you dismiss single threaded application performance, such as Excel, Web Browsing, Office, and various other common applications. If you use something WAY more frequently (Web Browsing) incremental, numerous, small savings add up.
4) Virtual Machines are a positive for Ryzen, I agree. Again, if you were running multiple VMs, you would *know this already and consider it*. There is no way you would have to tell someone who runs VMs that more cores is good.
5) There is no evidence to suggest the AM4 line will last 4 years. The history of AMD socket changes suggest changes at a much more frequent pace. From here: en.wikipedia.org/wik…ket
Socket 754 came in 2003 (Athlon 64). Socket 939 2004, Socket AM2 2006, Socket AM2+ 2007, Socket AM3 2009, Socket Am3+ 2011, Socket FM1, 2011, Socket FM2 2012, Socket FM2+ 2014, Socket AM1 2014, Socket Am4 2017.
When AMD was competitive in the CPU game, they released much more frequent socket updates. Do we buy a random internet comment that AM4 will last 4+ years? I don't. Anyway, this portion is entirely speculation. I don't really trust them to follow up, and why would I trust any timescales from AMD?
He probably said single threaded applications like excel etc don't really matter because there isn't much in it in real life, as far as excel which I use everyday it can't really get any faster for me or anyone else in my office.
Also, please, list the applications you use on a daily basis. Just go, see what you do per day.
My guess is 'Rendering 2+ hours of video' isn't one of your use cases.
I think you are confused about what comment I am referring to, I'm referring to someone saying:
"I'm sorry but talking about web browsing and office applications in this discussion is frankly hilarious, these applications are not going to tax either AMD or Intel CPUs."
You said you don't understand why he dismissed this. I am saying it was probably dismissed because to go from calculating in excel in .0002 seconds to doing it in .0001 seconds is pretty much irrelevant. I don't actually know anyone that only deals with excel anyway, I generally have at least 5 applications on the go at once, none of them will tax a cpu enough to make single thread performance matter.
You ignored what I said about sockets btw.
And about the cores, what's the difference? Sure, AM1, AM2, AM2+, AM3, AM3+, AM4.
If you want.
And did they mean when they first discussed the concept? Who knows, I don't trust AMD's timelines.
So AM3+ out in 2011 to AM4 out in 2017 is 6 years, it is actually closer to 6.5. You don't have to trust their timelines and fair enough if you don't but adding in various sockets from all over the place doesn't really come across as trustworthy either. You can make your point with the gap between AM3 and AM3+ being only 2 years for example.
This is hotukdeals not Coldestukdeals.
The cores are similar. Ryzen offers more of these cores for less. Unlocked. Intel have faster cores in certain K chips through higher frequencies but these also cost much much more for what they are. The 7700k is in its own ultra high priced quad core category of £310. For a quad core, in 2017. In the month a 6 core replacement by intel is also due...
Reviewers cant recommend the Core i5 or i7 as most fall below Ryzen in single core too. When you go to higher clock speeds the prices increase far too much. The biggest issue for intel is pricing.
Worth pointing out the i9 is £1000 for 10 cores. This AMD CPU is a total bargain in comparison.
Look at the prices and products. Why the hell are cheap Ryzen 3's for example being mentioned alongside the 1080 ti and 7700k's?
"Reviewers cant recommend the Core i5 or i7 as most fall below Ryzen in single core too."
From this comment
There is absolutely nothing unclear about that sentence. And it is wrong.
You you will notice reviewers and benchmarkers focus on other elements besides 7700k single core performance.
It certainly does show the Ryzen 1600 to be roughly matching the 7700K! While they did mention that a couple of the 1440P results were "inexplicable" wins for the 1600, that doesn't really take away from an excellent overall result:
The testbeds look fine, too (4.0GHz for a 1600 isn't much more optimistic than 4.9GHz for a 7700K, and both used sensible DDR4-3200). Brilliant stuff - I can absolutely see why you said "matched" now!
It's still worth noting the reason that 1080P is primarily used for CPU benchmarks - it's a better indication of what future performance will be like with a GPU upgrade.
But a 1600 system would come in ~£100 cheaper than a 7700K. As part of a ~£1000 system, one would be spending an extra 10% to get an initial 2-8% performance boost depending on resolution and card (based purely on the figures in that article). Obviously gamers expect diminishing returns on increased performance (otherwise we'd all have 1050Ti GPUs!), but it does knock a serious dent in the 7700K's advantage now that it no longer has the maths behind it.
Just a shame that these figures have come up just 20 days out from Coffee Lake, when it's simply bad advice to tell anyone to buy a CPU today. Still, I've got two gaming builds waiting for the Coffee Lake launch, and with more results like these it looks like one of them may end up Ryzen after all!
These results, on the face of it, look good and 'matched' until you realize all the reviewer did was remove the worst performing AMD game from the list.
In his first article he says:
"Some readers will undoubtedly declare that I'm biased towards a certain company, but if that was the case, there's no way you would see Deus Ex: Mankind Divided and Gears of War 4 in this list. Meanwhile, if I were biased towards Intel, I'd drop Total War: Warhammer and Civilization IV."
So, in the re-review, he drops Gears of War 4. Guess what?
Also, this reviewer came to this conclusion back in July.
techspot.com/rev…tml
The dude massages his results big time, look at the variable lists of games.
Why didn't he use the same 30 games this time?
Results like these from a non-disreputable site like Techspot certainly justify Uncommon.Sense's use of the word "matched", at any rate.
And, of course, it's all pretty much moot. No gamer would seriously buy a CPU twenty days out from a significant CPU launch (and it does speak to the integrity of people advising them to do so), and so in a couple of weeks we'll have Coffee Lake reviews from multiple reputable sites covering plenty of games - including updated Ryzen results, which will either confirm or undermine these results today. So it's really no problem either way.
"Please note that because there was a large difference in performance between running Deus Ex: Mankind Divided with DX11 and DX12 with the Ryzen CPU, I included both the results in this graph. However, I only included DX12 results for Total War: Warhammer and not the DX11 results since the R5 1600 was much faster in both tests. I favored DX12 as it's the newer API."
So in fact he was giving Intel a free boost, by excluding one of the results, then including both DX11//DX12 results for Deus Ex.
Also, game lists will change, as the new games that come out over time change, is 20 games not enough to create an average? How many other reviews do more that 8-10?
Ask him why he didn't use 30 games (tweet him)if you actually care that much, 'cause you seem obsessed with trying to nit pick. I am sure it will have something to do with testing two different graphics cards, on two different platforms, with twenty times, with three runs each, can you imagine how much longer that would take with thirty? What would it add, probably 1-05.-1% difference if that, since some of the games would have favoured the Ryzen, I am sure much to your disappointment.
The guy linked to his 3 month old review where he picked a set of games, and justified his game set there and showed he wasn't biased by including 'bad games for each platform'.
Then he re-reviews it and removes the worst game on the Ryzen test bed, Gears of War 4.
He's chosen a bunch of GPU bound games that show the Ryzen in a favorable light by averaging the performance of '20 games' but only really comparing 6 games that do anything with the CPU.
Who really thinks adding DOOM's results of 200/200 for literally every processor is beneficial?
I don't see 'TechSpot' as a beacon of testing integrity, this guy links to 'AdoredTV' as some trusted source. AdoredTV might as well just be a giant paid advertisement for all things AMD.
This list of games just looks cherry picked to allow the conclusion to be 'AMD has closed the gap, #finewine' when we're not focusing in on the fact that many titles are still 20% slower.
Because how on earth does this only suggest that Ryzen is '5% slower'?
Like really, we've got some pretty old games on this list, yet GOW4 is off? I think this graph would just look bad if you plopped in GOW4 being 30% slower. Where is FO4? GTA5?
You can easily massage results, and trying to suggest these games 'take a long time to test' is just outright untrue.
This guy is really good, from GamerNexus.
Anyway, techpowerup does a great job, and they don't adjust their testbed without strong thought. The idea that the set of games is randomly chosen for each review should absolutely flag techspot up in your mind as something weird.