I was about to ask the same thing, especially since the article doesn't go out of its way to explain the anomaly, unless it's that "new bench suite". But then the charts would not be indicative of anything.
I guess I was expecting 300MHz to make more of a difference. At least enough to keep the 3800XT consistently above the 3700X even if in the margin of error. Are those benchmarks so insensitive to frequency? I can wait for the full reviews, or the new benchmarking suite.
While this is a technically accurate explanation it shouldn't be the case on such a review. I mean a benchmarked CPU shouldn't be limited by its cooling. The point is to show what the CPU is capable of (at least with reasonable cooling options) not highlight the inadequacy of the cooling used. I hope that at least AMD's recommendation of an "AIO solution with a minimum 280mm radiator" was followed. I'm trying not to speculate since I have nothing to base this on.
For a benchmark I'd definitely oversize the cooling. People looking at higher end CPUs and who care about top performance will probably do the same.
Isn't this a classic case of not showing the proper comparisons? I get that AT got to test the 3800XT in this round, but it's unfortunate we can't compare against a 3800X.
I just feel like sometimes the choice of comparisons here doesn't make any sense. You need to look at what consumers will actively compare against. For instance does a user today buy a 3800X at a reduced price or wait for their retailer to have a 3800XT in stock? Those are questions a reader might want to have answered.
They are just getting started and I assume the comparisons you want will be coming. However they will show that the XT CPU's are essentially pointless. Obviously retailers can't sell an X and an XT for the same price, so they wont be the same price. Example: At Microcenter the 3900XT is $499.99 while the 3900X is $399.99. But who is going to pay $100 more for the XT when the only difference is 100 Mhz on the maximum boost? They don't have a 3800XT yet but 3600 XT Vs 3600X is $249.99 to $199.99. Who will pay even $50 extra for the added 100 Mhz?
So I expect the benchmarks of any XT to be close enough of a tie with the corresponding X as to make no difference. Even before this, many including me were passing on the 3800X in favor of the 3700X because the extra Mhz mostly didn't matter. The only way the XT's make sense is if the older X CPU's disappear from the market.
Given a long enough time-frame I'd have expected that to happen - but with the 4000 series approaching, it seems like they might all end up going EoL around about the same time.
"So I expect the benchmarks of any XT to be close enough of a tie with the corresponding X as to make no difference. "
That's the point of benchmarks. If we can just wave our hands and say that we don't need Anandtech. These guys are supposed to do the work that we can't afford to do--buy 2 CPUs and test them side by side.
I appreciate the articles coming out here, but I feel like logical comparisons need to be made, and if XT are supposed to replace X, then we really need XT vs X heads up comparisons along with the obvious Intel competitors. Think of what consumers might wonder when they want to buy a CPU and put in all the logical comparisons.
It's been a while since I bought a CPU but I paid attention during the Kaby Lake Skylake X launch to try to figure out which CPU to buy. The problem with AT was for the Skylake X benchmarks, they neglected to include ANY consumer level CPUs like Kaby Lake or even Skylake CPUs just to compare. Of course these may be two different market segments, but the lowest tier of the HEDT line might be competing against the high end of the mainstream consumer line especially from a price perspective. Someone might wonder "Should I splurge on a 7800X which is closer to $400 or settle for a 7700k closer to $300." This is a completely logical comparison that a shopper might consider.
The XTs replace the X's in the market place, but they're not necessarily superior. No-one is going to buy an XT to replace the corresponding X model (I hope).
3700x is binned for power/temp, not clocks. so it can effectively have a better clock average due to lower temperatures, and in some cases this will be better performance (depends on test in case).
I always felt the 3700x is the gem of this generation. too bad I could not purchase one while it was relatively cheap at the start of this year (after C19 crisis, all prices outside US sky-rocketed)
The prices in your charts appear to be pure fabrication.
I could not find an i9-10900K for sale on Amazon for less than $1,399. The Ryzen 9 3900X is available with next day shipping for $422.
The i9-10900K is available at Newegg and listed at $529 there. The Ryzen 9 3900X is $429.
How can you in good conscience provide such obviously faulty information? When I was reading the charts I was actually thinking, wow, Intel is quite competitive with Ryzen at the high end, then I decided to actually do some research and discovered that the information you are providing is completely inaccurate.
Also before you reply that the information you are providing is just the MSRP, consider that this is not relevant in any way to an actual evaluation of the product. If you are going to include MSRP as your "price" benchmark, why not just use manufacturer supplied benchmarks as well instead of running your own benchmarks? Why is manufacturer supplied data acceptable for one half of the value equation, but manufacturer supplied benchmarks not acceptable for the other half of the value equation?
Although maybe he's trying to make his point a little harshly, still, the main point remains. The products aren't truly comparable to the consumer unless they can see how much it will cost them now. When availability is limited by by the supply, then you will get pricing all over the board. But the fact remains, if you want to "buy" it now, this is the price you'll have to pay.
I support both showing the MSRP as well as the current Amazon/Newegg pricing. It helps the consumer more fully realize the value equation when considering their purchase. Newegg might be more accurate because at least they show what the price it would normally be sold when they (Newegg) are the seller.
The CPU's were only released today. Pretty much everyone who has them for sale is selling the XT's for Full MSRP. You can look up street prices of the X's yourself easily enough. But such comparisons on day one are rather pointless unless you are one of those that just has to have the latest and greatest on the first day. No one really knows where things will stabilize so there isn't much point in comparing old vs new other than that new will cost more than old for pretty much no gain in performance. That extra 100 Mhz boost isn't going to get you much.
I agree with romrunning that he made his point unnecessarily antagonistically, but I agree with it-- MSRP is _really_ not a useful measure for consumers.
How is using a price that is ALREADY outdated and wrong any better? ASP over the previous week or month before publication would be ideal, but the current selling price is good enough.
It's worth is as an artificial reference point. It places it more clearly in the line and shows how the manufacturers viewed their product against the competition. Take those numbers as having no units and only relevant relative to the generation.
One possibility is to use a graph/ data from camelcamelcamel - they track Amazon price history. In this case, the price history of the 6 or 7 CPUs discussed in this article can all be shown on a single graph and provide quite a lot of information to the reader. Camelcamelcamel only shows a single product’s price history so if Anandtech can show all the CPUs in one graph, that’s added value for Anand readers. Depending on camelcamelcamel’s API, it might be possible to show a live graph so that a reader coming back will see updated price information - this would be useful in the weeks after a new processor comes out, while the street price is still settling.
Why on earth would a static analysis reference a live price? A statement about value with a moving price is absolutely idiotic, and MSRP reflects the manufacturer's strategy in a way that's more interesting than retailers' strategy anyway. Obviously I wasn't going to buy an AMD card when those were the mining card of choice, but it was still interesting to note how close their MSRP/performance ratio was or wasn't.
Anandtech also does regular roundups that would, if not for the wildly fluctuating prices, sate your very need.
I don't come here to read about "manufacturer's strategy", and MSRP doesn't really reflect it anyway. Neither does current pricing necessarily only reflect an individual retailer's strategy.
If all (or most) retailers are providing a price lower than MSRP, you can infer that the manufacturer has lowered the wholesale price. That's entirely separate from the sort of market distortions we saw with mining.
You're making this argument as a defence of bad information. A person doesn't have to provide a workable alternative to observe that a methodology is bad, though this is a common misconception.
what I would like to see is instead of the pointless TDP figure as we know both sides manipulate these in some way can we have the actual TDP seen at the time of the test most testers have a basic test rig spec which can me referenced in the results as well, and yes real world prices please these are easy to find by linking the prices with the article use Amazon etc as a benchmark
That would be neat, too. An average power draw for the test as performed rather than the useless manufacturer figure - that way a buyer knows the actual thermal load they'll need to dissipate.
If I'm not mistaken, the prices listed are the manufacturer's suggested price. Since price-fixing is illegal, and CPU manufacturers haven't figured out how to make it happen without getting caught, real-world prices basically never align with suggested retail.
You might misunderstand what price fixing is. It's when competing vendors coordinate their prices together. That's usually illegal. Companies are free to dictate the prices that retailers charge for their products. That's not price fixing.
Price fixing is illegal; resale price maintenance isn't. Minimum advertised price policy and unilateral minimum price policies are legal and have been used or are in use by both AMD and Intel.
Amazon and Newegg prices can change day to day. Clearly Amazon is out of 10900K stock and some marketplace seller is trying to rip someone off. MSRP is the only price that will have any consistency. The writers can't be expected to come back and update all of their charts whenever Newegg has a sale.
It would also be difficult to account for all the different retailers. As you point out, they often have different prices for the same product. Is the street price what Newegg is charging, or is it what Microcenter is charging? I don't live anywhere near a Microcenter, but that price is available to some people.
To me, if you can get something for less than MSRP, great, you're getting a deal. If it's more, don't buy it. MSRP is useful as a guide for how much something should cost and it's up to you to determine if a retailer is currently offering a good price.
I'm okay with including one or two retailers just to provide context for price comparison. Amazon/Newegg are shipping everywhere, as opposed to some of Microcenter's "in store only" deals. So use Amazon/Newegg pricing for context, and the consumer can be aware that they can always price-shop on other retail sites.
The prices can be linked dynamically through back-end APIs so that it would show that day's pricing. Then the consumer would know how much roughly it would cost them today, and if they look at the article again next week, it could show the price then.
The key point for me, as a consumer, is to get the price/value comparison today if I'm going to place an order today. So if I find the 3800X is selling at $320, the 3800XT is $400, and the 10700K is at $410, then helps me determine what my best value option. Show the MSRP only doesn't give me the full picture, especially if the 3800X is listed at the original MSRP of $399.99.
I agree msrp is most usefull! For example in europe the price difference between non x, x, and Xt versiona Are much smaller. All of those Are really close to msrp prices! I don`t mind if there is link to current street prices, but it seems that the sale of non x and x models is ower so Sooner or later the price of non x, and x models get very newar the price of Xt models Also in USA. (Until there is the next sale that can affect the Xt models too!) But in anyway msrp is the best option, it tells what price you Are expected to pay, unless there is temporary sale somewhere that you should use to your adwantage.
There's a difference between "I might be able to get this for this price" and "nobody is selling this product at MSRP", though. Of course we wouldn't be expecting Anandtech to reflect every fluctuation in price in a JPEG chart - but on the flip side of that, by using MSRP to compare a brand-new product to a product that has settled into the marketplace, they give the newcomer an unfair advantage as products are more likely to drop *below* MSRP over time than the inverse.
I agree but in a way it doesn't matter because if someone decides on the 10900K based on its performance/price and then tries to buy it they'll see that they're screwed.
This is a valid point. It's admittedly difficult for Anandtech to hit a moving target re: price, but the way pricing information is presented here is deeply misleading.
It's a shame that AMD didn't boost the TDP a little to match the Intel competitors (on paper, in reality it would still be way below) and then boost the base clocks a little as well.
In reality this seems like an easy way to get some media column inches mid-year, and AMD wins whether it's 'XT is good' or 'XT is good but X is better value'.
I would prefer we use our technological advancement to reduce TDP over time. We sort of need to do something about our civilization's energy and resource consumption before we finish trashing the only place we have to live. Cutting corners in every possible way, including in our technological devices, would help make a small difference. Then again reproducing less would be a heck of a lot better, but we should probably be addressing that issue with as many remedies as possible.
As a matter of fact, I do use an Atom notebook that I bought second hand. I have not purchased new computer hardware for over a decade and I really like how little waste heat the lowest end of the laptop compute spectrum generates.
For the amount of work it accomplishes, yes it does create quite a bit of heat and is rather inefficient. It was far better than contemporary processors in terms of heat production and power input though. It also helped get a fair number of inexpensive computers into the hands of people that used them to fit well in terms of consumption when compared against the tasks they performed. Back then you could mostly idle a desktop and easily consume 60W of power or more (assuming a fairly low end desktop configuration with iGPU only) or put an Atom n270 under moderate to high load for less than 10W of power. It seems like that was a good trade-off if your workload was not overly demanding.
It would be faster and therefore probably more power-efficient to use turbo, at least to some extent, so it can do the work and then immediately idle the chips. Maybe not the very-highest turbo bins.
Have you been asleep at the wheel? Modern computers use a fraction the energy they did even 5-6 years ago, let alone 10. But liek it or not, we are not going back to the DOS days, there will always be a demand for more computing power.
I'd bet if someone went through your house, 95% of it would be junk from China. All shipped on shipping ships buring filthy bunker oil, products manufactured in filthy factories that have 0 enviromental controls, assembled by slave labor, using materials mined by slaves in Africa polluting even more due to lack of enviromental controls.
I'd also bet you're not using a 15 year old second hand computer to post here, and have a cell phone made with those same rare earth minerals. You likely drive a car, and clearly use the internet.
Practice what you preach.
Also, if you wanna go on about "muh reproduction" you are free to go to africa, the biggest population growth market on the planet, and tell them to stop reproducing. See how far that gets you.
Acer Aspire ES1-111M-C7DE bought from ebay for less than $100 USD running Linux Mint. My phone is LG Rebel 3 - a $10 refurb from my carrier (Tracfone) so it isn't new either. My car was given to me in 2005 by my mother who was sick of watching me drive a car that was built in 1990 and I still drive it today, though I don't travel much so it has less than 90k miles on it. My home is a 109 year old farmhouse that I rent, my furniture is mostly second hand from relatives or my ex's in-laws, in good shape, but not new to me. I consume bare necessities, subscribe to no recurring entertainment services besides a 10mbit internet line (no TV, no satellite, no streaming media, etc.). Mostly, I admit I do that sort of thing to pack away income for the future in the perhaps deluded hopes that I can live exclusively off interest it generates and continue folding excess back into the principal so there are no future concerns about retirement so mainly greed is my driver, but if it helps out the planet along the way then why not?
Give it a try someday. It's good to keep the expenses low, have a healthy emergency fund of cash, and I don't really feel like I'm missing out on anything by not burning more in material goods at the expense of being in any sort of debt. Or, I dunno, sit there being upset about it and ranting anti-Asian and anti-African rhetoric - you're welcome to do that too.
1) Pointing to the shittiness of our current iteration of global capitalism *is not a rebuttal to anyone advocating for making things better*; it's the laziest possible defence of the status-quo. The fact is that it's nearly impossible to buy things that aren't made using borderline-slave labour in horrible factories and shipped using dirty fuel. Is that the fault of the consumer? No, not even close. It's the fault of the various people in charge (i.e. business owners, shareholders, certain parts of the government). They're the ones who collectively decided that they couldn't make enough profits from paying local workers a living wage, and decided to move their factories abroad in a global race to the bottom. We who merely live in societies that work on that arrangement have no real choice but to participate in it, ergo, it is not hypocritical to be critical of said society whilst existing within it unless you're a billionaire, business owner, etc. Individuals following TheinsanegamerN's vapid prescription wouldn't make a fart in a volcano's worth of difference on a societal scale, because it's not a solution: it's a hack that allows people who don't want anything to change to feel morally superior to people who've noticed the problem. I point you to the cartoons with the guy in the well; TheinsanegamerN is the smug idiot in the well.
2) Population growth is a function of poverty and education. If you want to curb population growth in Africa, the best place to have started was about 500 years ago when the west started colonising the place and using the locals as slave labour for resource extraction. The next best place to start is now, by funding infrastructure, healthcare and education. If your first response to someone talking about population growth is "but muh Africa", you're demonstrating three things: your ignorance of the causes, your ignorance of the solutions, and your preference for racist rhetoric over actual scientific data.
In conclusion: I don't usually agree with PeachNCream, but your reply here is garbage.
@Spunjji Very nice analogy but I wouldn’t use intellect to reply to idiots like TheinsanegamerN because he’s just a stupid racist westerner, probably an American who’s well-fed with BS by the MSM. It baffles me why people in the west always stick their noses in countries half way around the earth and act like they don’t have any problem at home.
It *is* legit, in its own way. Most of the time when I respond to specific points of someone's fallacious argument, they either set off a gish-gallop or engage in ever-tightening circles of sophistry and empty rhetoric. There's nothing productive about that.
Sometimes it's okay to just say "you're too full of shit to see how full of shit you are". It may not be persuasive on its own, but if it's accurate in context, then it's legit.
FYI: I'm British. I grew up "well-fed with BS by the MSM" and my nation pretty much wrote the book on the issues America has now. It hasn't rendered me incapable of understanding them, but it did place a lot of barriers in front of me *seeing* them.
I'm not especially fussed about whether I convince *him* - that can be a fool's errand - but I leave comments like that because I feel it's important to push back against insidious racist narratives in a way that makes it clear I'm not just objecting to the tone, but the substance. Responding with a generalised critique, even if it's often an accurate one, doesn't really serve that purpose.
You do you, of course - but I'mma keep doing me. :)
The coming crisis is not fixed by lowering TDP. It could only be fixed by most, if not everyone drastically changing their lifestyle. But that's not going to happen. We happily go out on the streets, demonstrate and protest for a few hours, asking the politicans to invent some magic wand so we don't have to change ourselves. And then we go back into the routine of our modern lifestyle... ¯\_(ツ)_/¯
We're not the ones with the power to change it. Even all of the people protesting dramatically changing our own lifestyles does nothing more than collapse the economy, because the people participating in it are not the ones with the power to effect the necessary changes.
Of course, some of us *have made the changes anyway*.
As others mentioned, the manufacturing cost is going to blow any kind of TDP saving out of the water.
One of the most evironmentally friendly thighs you can do is keep old tech/machinery chugging as long as you can, so you don't have to buy new widgets. Hence the whole right to repair movement.
Lowering the thermals is also nice for those of us who live in areas that require air conditioning in the summer. My office gets very toasty if I am not careful about what parts go into my PC, which is why I tend to stick with ≤65W APUs and ≤350W power supplies. My next build is going to be a mini-ITX box that will hopefully be under 120W total.
Attempting to limit energy(resource) USE sociologically generally leads to stunted economic growth and increased mortality. In contrast, enhancing energy(resource) PRODUCTION generally leads to increased economic growth and extended life spans.
In short, don't use less; make more.
Just a philosophical bone throw. I see too many people thinking about resources as things that can only be destroyed, when that isn't the case.
More production requires more input. Inputs are finite. Making more things in the name of economic growth seems misguided at best. What might be smarter is finding a way to gracefully decline toward a sustainable steady state. While that might mean economic decline for a while and long-term stagnation our species will at least survive well enough to be able to debate the merits of whether or not more is better with one another. I like that option better than I like the idea of generations of people that come after us struggling to exist in the wasteland of our prior excesses with no means of debate at their disposal besides tossing rocks at one another while they starve to death.
You realise the outputs of these systems feed back into the inputs, right?
Eventually you hit a point where your paradigm no longer holds true. It's an observation of a system operating within certain parameters, not a universal law.
Increasing tdp does not help 7nm product much... one reason why Intel highend still is 14nm... it scales with super high tdp better than 10nm or 7nm...
When I saw the title I wanted to jump to the comments and say thumbs up for not bothering to test a 100mhz bump in frequency, but I see you did. Anyway, these are basically a rip off. 50-100$ more than X parts for 1-2% better performance.
To sucker people out of more $$$ for buying a year+ old CPU arch with a slightly better bin. And you KNOW there are people who will buy one of these then turn around and buy a 4000 series 5 months later.
I don't "know" that at all. In fact, I'd propose that you're talking out of your arse.
The point of the exercise is to "refresh" their product range to compete with Intel's own refreshed products for this year. It's that simple. It's been happening in the IT industry for decades; if you find it novel or surprising now then you're an ignoramus.
So they can be in the headlines again for releasing "new" products. It's important to keep the product line "fresh", especially when a major competitor is also releasing "new" products.
these guys now have a more mature silicon that clocks a lil bit higher than a year ago and now they're like shit we dont want to give u guys this perfomance for free that aint fair right? so here u go....the "new" XT CPUs lmao. Thats a page out of the Intel book right here.
That AMD does not recommend its own coolers with the XT for the same TDP makes me wonder perhaps its vaunted stock coolers are as good as people said they are. Kinda bummer that switching over to team Red is not as good a deal I thought I was getting now that I still have to get new CPU cooler.
Defaul cooler is ok. But of you want to get better boost clocks you need aftermarket cooler! I think that it is sensible to leave the cooler away from these halo products. If you Are willing to pay much more for 1 to 2% more speed you definitely Are going to buy aftermarket cooler in anyway to get that last 0.5% (and less noise!) I did buy 65w 3700x and do buy the most expensive notchtua cooler to it, even the wraith cooler is reasonable. Not because of better cooling. I wanted the Computer be as silent as possible! So now I have completely useless box cooler and really silent and really low temperature amd cpu in my rig.
They're precisely as good as they always were. You always wanted to replace them if you wanted to maintain full boost clocks for long periods of time and/or enjoy lower noise.
If you don't, then you're better off saving the money and buying a lower bin chip. You literally will not notice the difference, but the solid stock cooler will save you some cash.
My advice is, don't waste money on XT. They are like few percent better, half or less in gaming, plus they reset MSRP, while old parts can be gotten with discount and box cooler. Like 3600 can be had for 170USD, while 3600XT is 250USD. That is far more than just 2-4% you are getting in the best case. And 3700x can be had for 270USD, while 3800XT is 400USD, for again only 2-4% gains at best. And if you want yo say it might matter for productivity, 3900X is 422USD on Amazon right now and it will absolutely demolish 3800XT. Oh, and if you want yo say you want max FPS in gaming and you are willing to pay more, Intel is still undisputed king there, even if lead is small, XT chips are still far closer to rest of Ryzen lineup, than Intel. And that is coming from someone that will happily recommend Ryzen.
As for who might want it, it is people who like overclocking, tinkering and benchmarking. Bit higher silicon quality will help there and open up more possibilities. But for vast majority, you don't game benchmarks and tier higher will always beat XT. So unless prices get sharp drop, my recommendation is to not bother and either get old or wait for Zen3.
Full agreement. These are for people who want the absolute best AMD chip they can get *right now* at any cost - I suspect most of them will go to systems integrators.
Just did a quick scan of the charts... It looks as if in the Multi-threaded charts the highest core count wins by a margin related to cores, not speed, where in the single core charts the winner is always determined by the speed of the cores.. Looks like there is room for some more granularity in the MT charts....
True, but I feel the limited data provided is reasonable given the time constraints in play. He doesn't "have the numbers" because the tests have changed, so if being thorough is the goal, chucking in old numbers isn't going to get us there!
No, it's to demonstrate that this is an example of the inefficiencies of capitalism.
Only in capitalism are products made that have no reason to exist in the market other than to trick people into spending more of their money than they should, via smoke and mirrors tactics.
In a reasonable system, AMD would simply let some people be lucky by getting a little better turbo performance out of the existing line. But, we can't have that. We have to have as many redundancies as possible. Tack on a few extra letters and send out the samples!
I don't necessarily disagree, but I also don't think "demonstrating the inefficiencies of capitalism" is necessarily Anandtech's domain. 😅
AMD aren't operating within a reasonable system. They're operating within the system we have, where having new names and bigger numbers is a requirement to compete *even when you have an objectively superior product*, and even then, some people will still buy your competitor's products for the most specious reasons imaginable.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
111 Comments
Back to Article
drexnx - Tuesday, July 7, 2020 - link
weird to see in a couple cases the 3700X beating the 3800XT when they're the same part binned differently.close - Tuesday, July 7, 2020 - link
I was about to ask the same thing, especially since the article doesn't go out of its way to explain the anomaly, unless it's that "new bench suite". But then the charts would not be indicative of anything.Ian Cutress - Tuesday, July 7, 2020 - link
It's within 3%, which could just be run-to-run variation.close - Tuesday, July 7, 2020 - link
I guess I was expecting 300MHz to make more of a difference. At least enough to keep the 3800XT consistently above the 3700X even if in the margin of error. Are those benchmarks so insensitive to frequency? I can wait for the full reviews, or the new benchmarking suite.haukionkannel - Wednesday, July 8, 2020 - link
The heat, the heat, the heat... 3700x runs a Little bit cooler and can maintain higher clock a Little bit longer.close - Wednesday, July 8, 2020 - link
While this is a technically accurate explanation it shouldn't be the case on such a review. I mean a benchmarked CPU shouldn't be limited by its cooling. The point is to show what the CPU is capable of (at least with reasonable cooling options) not highlight the inadequacy of the cooling used. I hope that at least AMD's recommendation of an "AIO solution with a minimum 280mm radiator" was followed. I'm trying not to speculate since I have nothing to base this on.For a benchmark I'd definitely oversize the cooling. People looking at higher end CPUs and who care about top performance will probably do the same.
Meteor2 - Thursday, July 9, 2020 - link
Then do more runs? Otherwise the test has no value.OddFriendship8989 - Tuesday, July 7, 2020 - link
Isn't this a classic case of not showing the proper comparisons? I get that AT got to test the 3800XT in this round, but it's unfortunate we can't compare against a 3800X.I just feel like sometimes the choice of comparisons here doesn't make any sense. You need to look at what consumers will actively compare against. For instance does a user today buy a 3800X at a reduced price or wait for their retailer to have a 3800XT in stock? Those are questions a reader might want to have answered.
Ratman6161 - Tuesday, July 7, 2020 - link
They are just getting started and I assume the comparisons you want will be coming. However they will show that the XT CPU's are essentially pointless. Obviously retailers can't sell an X and an XT for the same price, so they wont be the same price. Example: At Microcenter the 3900XT is $499.99 while the 3900X is $399.99. But who is going to pay $100 more for the XT when the only difference is 100 Mhz on the maximum boost? They don't have a 3800XT yet but 3600 XT Vs 3600X is $249.99 to $199.99. Who will pay even $50 extra for the added 100 Mhz?So I expect the benchmarks of any XT to be close enough of a tie with the corresponding X as to make no difference. Even before this, many including me were passing on the 3800X in favor of the 3700X because the extra Mhz mostly didn't matter. The only way the XT's make sense is if the older X CPU's disappear from the market.
velanapontinha - Wednesday, July 8, 2020 - link
"The only way the XT's make sense is if the older X CPU's disappear from the market." - exactly!Spunjji - Wednesday, July 8, 2020 - link
Given a long enough time-frame I'd have expected that to happen - but with the 4000 series approaching, it seems like they might all end up going EoL around about the same time.OddFriendship8989 - Wednesday, July 8, 2020 - link
"So I expect the benchmarks of any XT to be close enough of a tie with the corresponding X as to make no difference. "That's the point of benchmarks. If we can just wave our hands and say that we don't need Anandtech. These guys are supposed to do the work that we can't afford to do--buy 2 CPUs and test them side by side.
I appreciate the articles coming out here, but I feel like logical comparisons need to be made, and if XT are supposed to replace X, then we really need XT vs X heads up comparisons along with the obvious Intel competitors. Think of what consumers might wonder when they want to buy a CPU and put in all the logical comparisons.
It's been a while since I bought a CPU but I paid attention during the Kaby Lake Skylake X launch to try to figure out which CPU to buy. The problem with AT was for the Skylake X benchmarks, they neglected to include ANY consumer level CPUs like Kaby Lake or even Skylake CPUs just to compare. Of course these may be two different market segments, but the lowest tier of the HEDT line might be competing against the high end of the mainstream consumer line especially from a price perspective. Someone might wonder "Should I splurge on a 7800X which is closer to $400 or settle for a 7700k closer to $300." This is a completely logical comparison that a shopper might consider.
Meteor2 - Thursday, July 9, 2020 - link
The XTs replace the X's in the market place, but they're not necessarily superior. No-one is going to buy an XT to replace the corresponding X model (I hope).marc1000 - Tuesday, July 7, 2020 - link
3700x is binned for power/temp, not clocks. so it can effectively have a better clock average due to lower temperatures, and in some cases this will be better performance (depends on test in case).I always felt the 3700x is the gem of this generation. too bad I could not purchase one while it was relatively cheap at the start of this year (after C19 crisis, all prices outside US sky-rocketed)
bji - Tuesday, July 7, 2020 - link
The prices in your charts appear to be pure fabrication.I could not find an i9-10900K for sale on Amazon for less than $1,399. The Ryzen 9 3900X is available with next day shipping for $422.
The i9-10900K is available at Newegg and listed at $529 there. The Ryzen 9 3900X is $429.
How can you in good conscience provide such obviously faulty information? When I was reading the charts I was actually thinking, wow, Intel is quite competitive with Ryzen at the high end, then I decided to actually do some research and discovered that the information you are providing is completely inaccurate.
bji - Tuesday, July 7, 2020 - link
Also before you reply that the information you are providing is just the MSRP, consider that this is not relevant in any way to an actual evaluation of the product. If you are going to include MSRP as your "price" benchmark, why not just use manufacturer supplied benchmarks as well instead of running your own benchmarks? Why is manufacturer supplied data acceptable for one half of the value equation, but manufacturer supplied benchmarks not acceptable for the other half of the value equation?hnlog - Tuesday, July 7, 2020 - link
Stop trollingromrunning - Tuesday, July 7, 2020 - link
Although maybe he's trying to make his point a little harshly, still, the main point remains. The products aren't truly comparable to the consumer unless they can see how much it will cost them now. When availability is limited by by the supply, then you will get pricing all over the board. But the fact remains, if you want to "buy" it now, this is the price you'll have to pay.I support both showing the MSRP as well as the current Amazon/Newegg pricing. It helps the consumer more fully realize the value equation when considering their purchase. Newegg might be more accurate because at least they show what the price it would normally be sold when they (Newegg) are the seller.
Ratman6161 - Tuesday, July 7, 2020 - link
The CPU's were only released today. Pretty much everyone who has them for sale is selling the XT's for Full MSRP. You can look up street prices of the X's yourself easily enough. But such comparisons on day one are rather pointless unless you are one of those that just has to have the latest and greatest on the first day. No one really knows where things will stabilize so there isn't much point in comparing old vs new other than that new will cost more than old for pretty much no gain in performance. That extra 100 Mhz boost isn't going to get you much.Spunjji - Wednesday, July 8, 2020 - link
His complaint wasn't about the XT pricing, it was about the X and the i9.schizoide - Tuesday, July 7, 2020 - link
I agree with romrunning that he made his point unnecessarily antagonistically, but I agree with it-- MSRP is _really_ not a useful measure for consumers.SirMaster - Tuesday, July 7, 2020 - link
But how is using a price that will likely be outdated and wrong by the time someone reads the article any more useful?WaltC - Tuesday, July 7, 2020 - link
3900X has been on sale at AMZN for months @ ~$420. Bought min in March.https://www.amazon.com/AMD-Ryzen-3900X-24-Thread-P...
qwertymac93 - Tuesday, July 7, 2020 - link
How is using a price that is ALREADY outdated and wrong any better? ASP over the previous week or month before publication would be ideal, but the current selling price is good enough.MSRP isn't worth the paper it's written on.
close - Wednesday, July 8, 2020 - link
It's worth is as an artificial reference point. It places it more clearly in the line and shows how the manufacturers viewed their product against the competition. Take those numbers as having no units and only relevant relative to the generation.bji - Thursday, July 9, 2020 - link
I apologize for being "unnecessarily antagonistic". I was just trying to make my point very clearly, not intending to be antagonizing.velanapontinha - Wednesday, July 8, 2020 - link
He makes a very good pointextide - Tuesday, July 7, 2020 - link
So what price should they use? Even you couldn't pick out a single price for the two parts you mentioned.Tomatotech - Wednesday, July 8, 2020 - link
One possibility is to use a graph/ data from camelcamelcamel - they track Amazon price history. In this case, the price history of the 6 or 7 CPUs discussed in this article can all be shown on a single graph and provide quite a lot of information to the reader. Camelcamelcamel only shows a single product’s price history so if Anandtech can show all the CPUs in one graph, that’s added value for Anand readers. Depending on camelcamelcamel’s API, it might be possible to show a live graph so that a reader coming back will see updated price information - this would be useful in the weeks after a new processor comes out, while the street price is still settling.Spunjji - Wednesday, July 8, 2020 - link
That would be ideal, really.lmcd - Thursday, July 9, 2020 - link
Why on earth would a static analysis reference a live price? A statement about value with a moving price is absolutely idiotic, and MSRP reflects the manufacturer's strategy in a way that's more interesting than retailers' strategy anyway. Obviously I wasn't going to buy an AMD card when those were the mining card of choice, but it was still interesting to note how close their MSRP/performance ratio was or wasn't.Anandtech also does regular roundups that would, if not for the wildly fluctuating prices, sate your very need.
Spunjji - Monday, July 13, 2020 - link
I don't come here to read about "manufacturer's strategy", and MSRP doesn't really reflect it anyway. Neither does current pricing necessarily only reflect an individual retailer's strategy.If all (or most) retailers are providing a price lower than MSRP, you can infer that the manufacturer has lowered the wholesale price. That's entirely separate from the sort of market distortions we saw with mining.
Zizy - Wednesday, July 8, 2020 - link
The cheapest one you can realistically buy the part for, what else?Spunjji - Wednesday, July 8, 2020 - link
You're making this argument as a defence of bad information. A person doesn't have to provide a workable alternative to observe that a methodology is bad, though this is a common misconception.alufan - Tuesday, July 7, 2020 - link
what I would like to see is instead of the pointless TDP figure as we know both sides manipulate these in some way can we have the actual TDP seen at the time of the test most testers have a basic test rig spec which can me referenced in the results as well, and yes real world prices please these are easy to find by linking the prices with the article use Amazon etc as a benchmarkSpunjji - Wednesday, July 8, 2020 - link
That would be neat, too. An average power draw for the test as performed rather than the useless manufacturer figure - that way a buyer knows the actual thermal load they'll need to dissipate.Lord of the Bored - Tuesday, July 7, 2020 - link
If I'm not mistaken, the prices listed are the manufacturer's suggested price. Since price-fixing is illegal, and CPU manufacturers haven't figured out how to make it happen without getting caught, real-world prices basically never align with suggested retail.JoeDuarte - Tuesday, July 7, 2020 - link
You might misunderstand what price fixing is. It's when competing vendors coordinate their prices together. That's usually illegal. Companies are free to dictate the prices that retailers charge for their products. That's not price fixing.qwertymac93 - Tuesday, July 7, 2020 - link
Price fixing is illegal; resale price maintenance isn't. Minimum advertised price policy and unilateral minimum price policies are legal and have been used or are in use by both AMD and Intel.cfenton - Tuesday, July 7, 2020 - link
Amazon and Newegg prices can change day to day. Clearly Amazon is out of 10900K stock and some marketplace seller is trying to rip someone off. MSRP is the only price that will have any consistency. The writers can't be expected to come back and update all of their charts whenever Newegg has a sale.It would also be difficult to account for all the different retailers. As you point out, they often have different prices for the same product. Is the street price what Newegg is charging, or is it what Microcenter is charging? I don't live anywhere near a Microcenter, but that price is available to some people.
To me, if you can get something for less than MSRP, great, you're getting a deal. If it's more, don't buy it. MSRP is useful as a guide for how much something should cost and it's up to you to determine if a retailer is currently offering a good price.
romrunning - Tuesday, July 7, 2020 - link
I'm okay with including one or two retailers just to provide context for price comparison. Amazon/Newegg are shipping everywhere, as opposed to some of Microcenter's "in store only" deals. So use Amazon/Newegg pricing for context, and the consumer can be aware that they can always price-shop on other retail sites.The prices can be linked dynamically through back-end APIs so that it would show that day's pricing. Then the consumer would know how much roughly it would cost them today, and if they look at the article again next week, it could show the price then.
The key point for me, as a consumer, is to get the price/value comparison today if I'm going to place an order today. So if I find the 3800X is selling at $320, the 3800XT is $400, and the 10700K is at $410, then helps me determine what my best value option. Show the MSRP only doesn't give me the full picture, especially if the 3800X is listed at the original MSRP of $399.99.
haukionkannel - Wednesday, July 8, 2020 - link
I agree msrp is most usefull!For example in europe the price difference between non x, x, and Xt versiona Are much smaller. All of those Are really close to msrp prices!
I don`t mind if there is link to current street prices, but it seems that the sale of non x and x models is ower so Sooner or later the price of non x, and x models get very newar the price of Xt models Also in USA. (Until there is the next sale that can affect the Xt models too!)
But in anyway msrp is the best option, it tells what price you Are expected to pay, unless there is temporary sale somewhere that you should use to your adwantage.
Spunjji - Wednesday, July 8, 2020 - link
There's a difference between "I might be able to get this for this price" and "nobody is selling this product at MSRP", though. Of course we wouldn't be expecting Anandtech to reflect every fluctuation in price in a JPEG chart - but on the flip side of that, by using MSRP to compare a brand-new product to a product that has settled into the marketplace, they give the newcomer an unfair advantage as products are more likely to drop *below* MSRP over time than the inverse.shabby - Tuesday, July 7, 2020 - link
I agree but in a way it doesn't matter because if someone decides on the 10900K based on its performance/price and then tries to buy it they'll see that they're screwed.Spunjji - Wednesday, July 8, 2020 - link
This is a valid point. It's admittedly difficult for Anandtech to hit a moving target re: price, but the way pricing information is presented here is deeply misleading.Valantar - Tuesday, July 7, 2020 - link
No news on Renoir APUs launching today? Was really hoping for those to be announced at the same time as the XT chips ...yeeeeman - Tuesday, July 7, 2020 - link
july 21stValantar - Tuesday, July 7, 2020 - link
But that's soooooo faaaaaaar into the future :'( Two whole weeks more? Man, I was hoping to get my new HTPC built this summer ...psychobriggsy - Tuesday, July 7, 2020 - link
It's a shame that AMD didn't boost the TDP a little to match the Intel competitors (on paper, in reality it would still be way below) and then boost the base clocks a little as well.In reality this seems like an easy way to get some media column inches mid-year, and AMD wins whether it's 'XT is good' or 'XT is good but X is better value'.
yeeeeman - Tuesday, July 7, 2020 - link
they are already turbo-ing very high all core, so AMD can't do too much to improve these.PeachNCream - Tuesday, July 7, 2020 - link
I would prefer we use our technological advancement to reduce TDP over time. We sort of need to do something about our civilization's energy and resource consumption before we finish trashing the only place we have to live. Cutting corners in every possible way, including in our technological devices, would help make a small difference. Then again reproducing less would be a heck of a lot better, but we should probably be addressing that issue with as many remedies as possible.Achaios - Tuesday, July 7, 2020 - link
Be the change you wanna see in the world.Buy a CPU and operate it always with its turbo disabled. A laptop with an U type CPU operated this way should consume abt 8 Watts per hour.
Next stage is to buy a bicycle to provide energy for the laptop in your trailer, like those hippy guys did in the X-files.
TheinsanegamerN - Tuesday, July 7, 2020 - link
This is terrible advice, the energy used to make that laptop will never be recouped.He should be using a atom powered netbook bought secondhand. Remember, reduce, reuse, recycle!
PeachNCream - Tuesday, July 7, 2020 - link
As a matter of fact, I do use an Atom notebook that I bought second hand. I have not purchased new computer hardware for over a decade and I really like how little waste heat the lowest end of the laptop compute spectrum generates.ads295 - Tuesday, July 7, 2020 - link
Hahaha, my Atom N270 netbook produces so much waste heat it's crazy.PeachNCream - Wednesday, July 8, 2020 - link
For the amount of work it accomplishes, yes it does create quite a bit of heat and is rather inefficient. It was far better than contemporary processors in terms of heat production and power input though. It also helped get a fair number of inexpensive computers into the hands of people that used them to fit well in terms of consumption when compared against the tasks they performed. Back then you could mostly idle a desktop and easily consume 60W of power or more (assuming a fairly low end desktop configuration with iGPU only) or put an Atom n270 under moderate to high load for less than 10W of power. It seems like that was a good trade-off if your workload was not overly demanding.GreenReaper - Tuesday, July 7, 2020 - link
It would be faster and therefore probably more power-efficient to use turbo, at least to some extent, so it can do the work and then immediately idle the chips. Maybe not the very-highest turbo bins.Foeketijn - Wednesday, July 8, 2020 - link
8 Watts per hour. That hurts my brain.Use a 400 series chipset instead of the 570 is even more better.
TheinsanegamerN - Tuesday, July 7, 2020 - link
Have you been asleep at the wheel? Modern computers use a fraction the energy they did even 5-6 years ago, let alone 10. But liek it or not, we are not going back to the DOS days, there will always be a demand for more computing power.I'd bet if someone went through your house, 95% of it would be junk from China. All shipped on shipping ships buring filthy bunker oil, products manufactured in filthy factories that have 0 enviromental controls, assembled by slave labor, using materials mined by slaves in Africa polluting even more due to lack of enviromental controls.
I'd also bet you're not using a 15 year old second hand computer to post here, and have a cell phone made with those same rare earth minerals. You likely drive a car, and clearly use the internet.
Practice what you preach.
Also, if you wanna go on about "muh reproduction" you are free to go to africa, the biggest population growth market on the planet, and tell them to stop reproducing. See how far that gets you.
PeachNCream - Tuesday, July 7, 2020 - link
Acer Aspire ES1-111M-C7DE bought from ebay for less than $100 USD running Linux Mint. My phone is LG Rebel 3 - a $10 refurb from my carrier (Tracfone) so it isn't new either. My car was given to me in 2005 by my mother who was sick of watching me drive a car that was built in 1990 and I still drive it today, though I don't travel much so it has less than 90k miles on it. My home is a 109 year old farmhouse that I rent, my furniture is mostly second hand from relatives or my ex's in-laws, in good shape, but not new to me. I consume bare necessities, subscribe to no recurring entertainment services besides a 10mbit internet line (no TV, no satellite, no streaming media, etc.). Mostly, I admit I do that sort of thing to pack away income for the future in the perhaps deluded hopes that I can live exclusively off interest it generates and continue folding excess back into the principal so there are no future concerns about retirement so mainly greed is my driver, but if it helps out the planet along the way then why not?Give it a try someday. It's good to keep the expenses low, have a healthy emergency fund of cash, and I don't really feel like I'm missing out on anything by not burning more in material goods at the expense of being in any sort of debt. Or, I dunno, sit there being upset about it and ranting anti-Asian and anti-African rhetoric - you're welcome to do that too.
Ranger1065 - Tuesday, July 7, 2020 - link
It's tricky to fault the logic behind consuming and reproducing less and worrisome to contemplate how many fail to appreciate it.Foeketijn - Wednesday, July 8, 2020 - link
+1Spunjji - Wednesday, July 8, 2020 - link
Not just fail to appreciate it - get steaming mad, throw their toys out of the pram and call *everyone else* a hypocrite.Spunjji - Wednesday, July 8, 2020 - link
1) Pointing to the shittiness of our current iteration of global capitalism *is not a rebuttal to anyone advocating for making things better*; it's the laziest possible defence of the status-quo.The fact is that it's nearly impossible to buy things that aren't made using borderline-slave labour in horrible factories and shipped using dirty fuel. Is that the fault of the consumer? No, not even close. It's the fault of the various people in charge (i.e. business owners, shareholders, certain parts of the government). They're the ones who collectively decided that they couldn't make enough profits from paying local workers a living wage, and decided to move their factories abroad in a global race to the bottom. We who merely live in societies that work on that arrangement have no real choice but to participate in it, ergo, it is not hypocritical to be critical of said society whilst existing within it unless you're a billionaire, business owner, etc. Individuals following TheinsanegamerN's vapid prescription wouldn't make a fart in a volcano's worth of difference on a societal scale, because it's not a solution: it's a hack that allows people who don't want anything to change to feel morally superior to people who've noticed the problem. I point you to the cartoons with the guy in the well; TheinsanegamerN is the smug idiot in the well.
2) Population growth is a function of poverty and education. If you want to curb population growth in Africa, the best place to have started was about 500 years ago when the west started colonising the place and using the locals as slave labour for resource extraction. The next best place to start is now, by funding infrastructure, healthcare and education. If your first response to someone talking about population growth is "but muh Africa", you're demonstrating three things: your ignorance of the causes, your ignorance of the solutions, and your preference for racist rhetoric over actual scientific data.
In conclusion: I don't usually agree with PeachNCream, but your reply here is garbage.
sonny73n - Thursday, July 9, 2020 - link
@SpunjjiVery nice analogy but I wouldn’t use intellect to reply to idiots like TheinsanegamerN because he’s just a stupid racist westerner, probably an American who’s well-fed with BS by the MSM.
It baffles me why people in the west always stick their noses in countries half way around the earth and act like they don’t have any problem at home.
SkyBill40 - Friday, July 10, 2020 - link
So rather than actually do something productive like addressing any specific points, you'd just rather ad hominem instead and be that guy?Got it. That totally seems legit.
Spunjji - Friday, July 10, 2020 - link
It *is* legit, in its own way. Most of the time when I respond to specific points of someone's fallacious argument, they either set off a gish-gallop or engage in ever-tightening circles of sophistry and empty rhetoric. There's nothing productive about that.Sometimes it's okay to just say "you're too full of shit to see how full of shit you are". It may not be persuasive on its own, but if it's accurate in context, then it's legit.
Spunjji - Friday, July 10, 2020 - link
FYI: I'm British. I grew up "well-fed with BS by the MSM" and my nation pretty much wrote the book on the issues America has now. It hasn't rendered me incapable of understanding them, but it did place a lot of barriers in front of me *seeing* them.I'm not especially fussed about whether I convince *him* - that can be a fool's errand - but I leave comments like that because I feel it's important to push back against insidious racist narratives in a way that makes it clear I'm not just objecting to the tone, but the substance. Responding with a generalised critique, even if it's often an accurate one, doesn't really serve that purpose.
You do you, of course - but I'mma keep doing me. :)
SkyBill40 - Saturday, July 11, 2020 - link
Fair enough.MrVibrato - Tuesday, July 7, 2020 - link
The coming crisis is not fixed by lowering TDP. It could only be fixed by most, if not everyone drastically changing their lifestyle. But that's not going to happen. We happily go out on the streets, demonstrate and protest for a few hours, asking the politicans to invent some magic wand so we don't have to change ourselves. And then we go back into the routine of our modern lifestyle... ¯\_(ツ)_/¯Spunjji - Wednesday, July 8, 2020 - link
We're not the ones with the power to change it. Even all of the people protesting dramatically changing our own lifestyles does nothing more than collapse the economy, because the people participating in it are not the ones with the power to effect the necessary changes.Of course, some of us *have made the changes anyway*.
Deicidium369 - Tuesday, July 7, 2020 - link
Your people are doomed. I think we modified the wrong species.PeachNCream - Wednesday, July 8, 2020 - link
I see you have your fishing rod out. It doesn't look like you put the right bait on the hook.Spunjji - Wednesday, July 8, 2020 - link
Whose people? What species?I can't tell if this is a joke or really, really gross.
brucethemoose - Tuesday, July 7, 2020 - link
As others mentioned, the manufacturing cost is going to blow any kind of TDP saving out of the water.One of the most evironmentally friendly thighs you can do is keep old tech/machinery chugging as long as you can, so you don't have to buy new widgets. Hence the whole right to repair movement.
Lucky Stripes 99 - Tuesday, July 7, 2020 - link
Lowering the thermals is also nice for those of us who live in areas that require air conditioning in the summer. My office gets very toasty if I am not careful about what parts go into my PC, which is why I tend to stick with ≤65W APUs and ≤350W power supplies. My next build is going to be a mini-ITX box that will hopefully be under 120W total.qwertymac93 - Tuesday, July 7, 2020 - link
Attempting to limit energy(resource) USE sociologically generally leads to stunted economic growth and increased mortality. In contrast, enhancing energy(resource) PRODUCTION generally leads to increased economic growth and extended life spans.In short, don't use less; make more.
Just a philosophical bone throw. I see too many people thinking about resources as things that can only be destroyed, when that isn't the case.
PeachNCream - Wednesday, July 8, 2020 - link
More production requires more input. Inputs are finite. Making more things in the name of economic growth seems misguided at best. What might be smarter is finding a way to gracefully decline toward a sustainable steady state. While that might mean economic decline for a while and long-term stagnation our species will at least survive well enough to be able to debate the merits of whether or not more is better with one another. I like that option better than I like the idea of generations of people that come after us struggling to exist in the wasteland of our prior excesses with no means of debate at their disposal besides tossing rocks at one another while they starve to death.Spunjji - Wednesday, July 8, 2020 - link
You realise the outputs of these systems feed back into the inputs, right?Eventually you hit a point where your paradigm no longer holds true. It's an observation of a system operating within certain parameters, not a universal law.
haukionkannel - Wednesday, July 8, 2020 - link
Increasing tdp does not help 7nm product much... one reason why Intel highend still is 14nm... it scales with super high tdp better than 10nm or 7nm...yeeeeman - Tuesday, July 7, 2020 - link
When I saw the title I wanted to jump to the comments and say thumbs up for not bothering to test a 100mhz bump in frequency, but I see you did. Anyway, these are basically a rip off. 50-100$ more than X parts for 1-2% better performance.TheinsanegamerN - Tuesday, July 7, 2020 - link
Absolutely. Anyone paying more money for one of these is just as dumb as someone buying a 3800x when the 3700x exists and hits the same boost clocks.patel21 - Tuesday, July 7, 2020 - link
Why would AMD even bother releasing these as new models.TheinsanegamerN - Tuesday, July 7, 2020 - link
To sucker people out of more $$$ for buying a year+ old CPU arch with a slightly better bin. And you KNOW there are people who will buy one of these then turn around and buy a 4000 series 5 months later.Spunjji - Wednesday, July 8, 2020 - link
I don't "know" that at all. In fact, I'd propose that you're talking out of your arse.The point of the exercise is to "refresh" their product range to compete with Intel's own refreshed products for this year. It's that simple. It's been happening in the IT industry for decades; if you find it novel or surprising now then you're an ignoramus.
meacupla - Tuesday, July 7, 2020 - link
Because if they sell it under the same part number, the would have to be sold at the same price when they eventually go on discount.Because you don't have to play the CPU manufactured date/fab lottery that has happened in the past with Intel and AMD.
However, on the flip side, this doesn't help AMD's lack of memory BIOS problems.
qwertymac93 - Tuesday, July 7, 2020 - link
So they can be in the headlines again for releasing "new" products. It's important to keep the product line "fresh", especially when a major competitor is also releasing "new" products.Hxx - Tuesday, July 7, 2020 - link
these guys now have a more mature silicon that clocks a lil bit higher than a year ago and now they're like shit we dont want to give u guys this perfomance for free that aint fair right? so here u go....the "new" XT CPUs lmao. Thats a page out of the Intel book right here.haukionkannel - Wednesday, July 8, 2020 - link
Nah... amd deparment is full of amatours!Intel would have made new socket for this so that you Also would have to buy new motherboard!
Spunjji - Wednesday, July 8, 2020 - link
In MSRP terms they are giving you the new performance for free.Some people don't just look a gift horse in the mouth, they give it a full-on colon inspection.
wr3zzz - Tuesday, July 7, 2020 - link
That AMD does not recommend its own coolers with the XT for the same TDP makes me wonder perhaps its vaunted stock coolers are as good as people said they are. Kinda bummer that switching over to team Red is not as good a deal I thought I was getting now that I still have to get new CPU cooler.haukionkannel - Wednesday, July 8, 2020 - link
Defaul cooler is ok. But of you want to get better boost clocks you need aftermarket cooler!I think that it is sensible to leave the cooler away from these halo products. If you Are willing to pay much more for 1 to 2% more speed you definitely Are going to buy aftermarket cooler in anyway to get that last 0.5% (and less noise!)
I did buy 65w 3700x and do buy the most expensive notchtua cooler to it, even the wraith cooler is reasonable. Not because of better cooling. I wanted the Computer be as silent as possible!
So now I have completely useless box cooler and really silent and really low temperature amd cpu in my rig.
Spunjji - Wednesday, July 8, 2020 - link
They're precisely as good as they always were. You always wanted to replace them if you wanted to maintain full boost clocks for long periods of time and/or enjoy lower noise.If you don't, then you're better off saving the money and buying a lower bin chip. You literally will not notice the difference, but the solid stock cooler will save you some cash.
Astroturf whining is tedious.
explodingbullet - Tuesday, July 7, 2020 - link
SEP? wtf is an etailer? You mean MSRP?Slash3 - Tuesday, July 7, 2020 - link
Surfin' those nets on that information superhighway with your double barrel 128Kbit ISDN link, yeeee hawwwww NETSCAAAAAAAPPEEESpunjji - Wednesday, July 8, 2020 - link
cYbErSpAcEsuperflex - Tuesday, July 7, 2020 - link
Sucks that you have to buy that crap on your McDonalds wage.I'll run SuperPi 100 times tonight for you while idling my Audi V8
tygrus - Tuesday, July 7, 2020 - link
It's a lot of fuss over 1 to 2%.Yawn ... wake me up when Zen3 arrives.
School report card = C, could do much better if she applied herself.
SethNW - Tuesday, July 7, 2020 - link
My advice is, don't waste money on XT. They are like few percent better, half or less in gaming, plus they reset MSRP, while old parts can be gotten with discount and box cooler. Like 3600 can be had for 170USD, while 3600XT is 250USD. That is far more than just 2-4% you are getting in the best case. And 3700x can be had for 270USD, while 3800XT is 400USD, for again only 2-4% gains at best. And if you want yo say it might matter for productivity, 3900X is 422USD on Amazon right now and it will absolutely demolish 3800XT. Oh, and if you want yo say you want max FPS in gaming and you are willing to pay more, Intel is still undisputed king there, even if lead is small, XT chips are still far closer to rest of Ryzen lineup, than Intel. And that is coming from someone that will happily recommend Ryzen.As for who might want it, it is people who like overclocking, tinkering and benchmarking. Bit higher silicon quality will help there and open up more possibilities. But for vast majority, you don't game benchmarks and tier higher will always beat XT. So unless prices get sharp drop, my recommendation is to not bother and either get old or wait for Zen3.
Spunjji - Wednesday, July 8, 2020 - link
Full agreement. These are for people who want the absolute best AMD chip they can get *right now* at any cost - I suspect most of them will go to systems integrators.croc - Tuesday, July 7, 2020 - link
Just did a quick scan of the charts... It looks as if in the Multi-threaded charts the highest core count wins by a margin related to cores, not speed, where in the single core charts the winner is always determined by the speed of the cores.. Looks like there is room for some more granularity in the MT charts....Oxford Guy - Wednesday, July 8, 2020 - link
I would not by until there are two more Xs.hoohoo - Wednesday, July 8, 2020 - link
HEY DR. IAN CUTRESS!!!!!I KNOW YOU HAVE NUMBERS FOR THE 3600X AND 3800X. HOW ABOUT PUTTING THEM ON THE CHARTS, SO WE CAN SEE THE DIFFERENCE WITH THE XT VERSIONS.
SORRY SHOUTING BUT THIS IS JUST BLOODY RIDICULOUS.
Spunjji - Thursday, July 9, 2020 - link
It's going to be within the margins of test variance.KAlmquist - Friday, July 10, 2020 - link
You are probably correct, but the whole point of running benchmarks is to be sure that that is the case.Spunjji - Friday, July 10, 2020 - link
True, but I feel the limited data provided is reasonable given the time constraints in play. He doesn't "have the numbers" because the tests have changed, so if being thorough is the goal, chucking in old numbers isn't going to get us there!Oxford Guy - Friday, July 10, 2020 - link
"He doesn't 'have the numbers' because the tests have changed"Which makes sense since this is a completely different set of CPUs.
Oxford Guy - Friday, July 10, 2020 - link
No, it's to demonstrate that this is an example of the inefficiencies of capitalism.Only in capitalism are products made that have no reason to exist in the market other than to trick people into spending more of their money than they should, via smoke and mirrors tactics.
In a reasonable system, AMD would simply let some people be lucky by getting a little better turbo performance out of the existing line. But, we can't have that. We have to have as many redundancies as possible. Tack on a few extra letters and send out the samples!
Spunjji - Monday, July 13, 2020 - link
I don't necessarily disagree, but I also don't think "demonstrating the inefficiencies of capitalism" is necessarily Anandtech's domain. 😅AMD aren't operating within a reasonable system. They're operating within the system we have, where having new names and bigger numbers is a requirement to compete *even when you have an objectively superior product*, and even then, some people will still buy your competitor's products for the most specious reasons imaginable.
Oxford Guy - Friday, July 10, 2020 - link
IT WOULD BE INCONVENIENT FOR AMD'S MARKETING NONSENSE.Tilmitt - Thursday, October 8, 2020 - link
"We’re planning on doing a full article with our updated benchmark suite and new tests after we’ve done more regression testing."When is this going to be done?