jpishgar :
you think they are all doing it because they are AMD fanboys?
...yes. None of the comments so far have been critical of methodology, approach, findings, or the analysis itself. For the most part, they've been "OMG Tom's how much in the briefcase from Intel? lol" and the like. There was no briefcase. There was no pay-for-play. No check in the mailbox. That's not the way things are done here, and we're one of the few spots left that's actually been able to retain editorial independence and integrity. We do, however, get this variety of reaction on almost every review that is critical of or makes conclusions about certain pieces of hardware, usually CPUs and Graphics Cards. I mean, goodness, the author posted about the Ryzen series in a beaming review earlier in the week, and even in that piece there were some critical comments from Intel fanboys about how "AMD is stealing your money" and questioning the editorial bias of Tom's.
there are reviews all over the internet who clearly show how an i5 can easily fall behind a 6 core Ryzen when push comes to shove
On productivity, not particularly gaming. That's even called out in the review.
Not everyone is a fan boy, and it is everyone, just about as just about everyone here thinks your conclusions are very very wrong.
Explain why. We like feedback. We dig dissent. Hell, we have an entire forum category dedicated to the wars that rage between Intel and AMD in the CPU market, and fiery speculation threads with people tossing molotovs at each other hourly over the course of years. There's absolutely no problem with telling us you disagree with us - our editorial guys are usually pretty super-confident in their findings and have the data and evidence to back it up. But, we also value our reputation, and have to insist that the baseless accusations of bias and pay-for-play accompany some kind of proof. We simply don't do that. Fighting against that crap is how we got our start. That's where our roots are.
-JP
Ok, I hope I'm sufficiently an old enough hand you won't dismiss my comments as being some sort of paid AMD shill.
FIRST let me preface by stating a few things
-I own a desktop and laptop, both sport intel CPUS
-I am an overclocker/computer enthusiast
-I work in IT, and handle a wide range of technology, am intimately familiar with both AMD and Intel product lines.
The problem with your methodology is the lack of "real world" experience.
I'll explain. If you look at pure benches for a modern kaby i3, you'll see it's a respectable gaming cpu. However, anyone using a Kaby i3 on a "real" computer with common "software" installed on it, not a test bench pcs which has just the OS, drivers and testing software, you get a very different experience. I know, I play on enough modern i3s to know they're BARELY functional cpus in OFFICE environments let alone gaming environments. They're crippled by the lack of real cores critically. A fact that amazingly an end user in the real world can experience and see with ease, but somehow escapes most reviewers tests or experience.
-THIS IS THE DISCONNECT- which i've complained about with almost every single review i've seen here and in other places of both cpus and gpus since probably 2012. There is a real and appreciable difference between the testing results in your benches and the end user's experience when they have a real computer in their house, with steam installed on it, an AV installed on it, ms office installed on it, google chrome installed on it, windows 10 doing it's stupid stuff.. the list goes on and on and on for all the things THG and other reviewers DONT test for or even experience in their test rigs, throwing into question the very results they are purportedly are testing for.
-NOW- I'm not a professional bencher nor am I a game tester, I don't believe I can give you guys good advice on how to more accurately demonstrate a real world pc environment with your gaming benches without probably suggesting something which will invalidate my point. But I think my overall point is valid. There isn't a single pentium cpu I'd ever suggest to anyone in the year 2017 to run their office pc, let alone as a "gaming quality" cpu, which raises the question just what on earth do you test?!!! How are you test benches set up that you can't see the criminally poor performance of a pure dual core cpu in this day and age?
This whole probably came into clear focus with me when I moved to this i5-4690k from a piledriver fx8320. My old 8320 was a great overclocker, hit 5.2ghz, I kept it at 5.0 24/7. Had an ssd as a main hard drive. (I need to point that out btw). When I lost that system I replaced it with this DC i5-4690k, also a ssd (both samsung evos, the piledriver was a 750evo i think and this is an 850evo). Now I admit I'm a bit more sensitive to cpu performance that the normal user, and this i5 has been fine (mostly). But I could tell immediately I lost 4 cores when I "upgraded" to this DC chip. Even lightly overclocked (this was a bum chip, never could get it over 4.3, only really stable at 4.2) I've noticed the lack of logical cores (slightly). It does hang briefly playing games that my old fx8320 + r9-280x smashed without issue, this i5+gtx970 seem to stagger from time to time, to straight up hang. Same games, basically same hard drive, same vsync locked 60fps on the same 60hz 1080p monitor, yet the i5+gtx970 gives me a worse experience in some games then the inferior piledriver did?
hard to figure. But then It's not that hard to figure if you think about it a bit. We're talking about a game that smashes 3 cores. The gpus don't matter because they both were able to hit a full solid 60fps in the games in question at the same resolution. but the experience on the i5 is worse (slightly) because it can't manage all the other crap I have going on, on the other 2 screens while I play these games. those 4 cores aren't enough, even though they're undoubtedly much much faster.
Now I've not tried ryzen yet. I've actually purchased a r7-1700, but it's still in it's box because I'm waiting for the itx motherboards (I got a case for it too, but still waiting on the motherboards). SO I don't know how it behaves. But I do suspect all those extra cores will show up in a real world computer pretty obviously like my old piledriver used to. Is this ryzen a sidegrade from my i5, probably, I mean the ipc is closer to broadwell then haswell for ryzen, but the clock speeds are pretty slow, of course my cpu couldn't overclock anyway. I'm not moving to ryzen because I expect it to be faster then what I have, I'm going to ryzen because it will go faster, cooler, lower power consumption and have many many many times the cores then i'm currently using, and I suspect that will show up in a lot of little ways in my daily pc use, which isn't represented by your benching.
As a sidenote: I couldn't with a straight face ever suggest someone purchase a i3 or pentium cpu in this day and age. Furthermore with the ryzen r5-1500(x) cpus sporing 4c/8t I can't imagine any NON-K cpu should be suggested ever again in it's price range. Frankly, I don't think it's actually possible to justify a i~ anything over an r5 right now. Even the i7-7700k. the math actually looks a little worse when comparing my r7 vs the i7-7700k, while the r7 is nice it's just not the same league price/performance, but the r5-1600? as long as that's out there buying a 7700k is just a vanity purchase.
Facebook
Twitter
Instagram