There's a reason why this CPU is being offered for less than even the R9-7900 and that reason is the fact that the R9-7900X3D is easily the worst Zen4 CPU ever produced.
Sounds like apocalyptic and deeply cynical take on the CPU. Let's examine it.
I said it when it was released and I still say it now. The Ryzen 9 X3D CPUs were a huge mistake on AMD's part and the cynicism of its creation really irked me. The R9-7900X3D is especially bad as there's literally nothing that stands out about this CPU to make it compelling to anyone.
- 'to anyone'? Crude generalization that makes this negative narrative suspicious out of the gate
- it sells between 40-100 units on a weekly basis on Mindfactory alone, so there are thousands, if not tens of thousands of folks globally who already enjoy it and find it interesting for their needs. Are you able to recognize that others actually buy this CPU, even if you do not like it?
- therefore, what you wrote above is blatantly wrong
The R9-7950X3D is at least compelling as AMD's current flagship CPU (overall, it still sucks though).
The R9-7950X is compelling as the most potent productivity CPU that AMD currently produces.
The R9-7900 offers the best overall value in the Ryzen 9 lineup along with the lowest power draw (TDP of only 65W) and comes with the Wraith Prism included, a CPU cooler that can easily handle it.
- none of the above is an argument against 7900X3D; each chip can find buyers with specific needs/workloads; it's called diversity
- you know, people buy different types of croissants in a bakery, each one to their taste and liking. Who is to say whether chocolate, almond or butter croissant is the best? Butter croissant definitively sells the most globally, but almond croissant has its crowd too. It's not as loud and popular as the butter croissant crowd is though.
The R9-7900X3D can't game as well as the R7-7800X3D, it's not as productive as the R9-7900 or R9-7900X because of the clock and power limitations imposed upon it by the 3D V-Cache, something that has no place on a CPU that is most suited to productivity.
- gaming experience also depends on preferred titles, GPU and display resolutions
- in 4K gaming, there are 15 top CPUs, including 7900X3D, that are only 6-7% away from each other (TPU charts), on average, so it's largely negligible
- 7900X3D was neither designed to be the best in gaming nor in productivity. Why would anyone ever expect that from this CPU?
- 7900X is faster in productivity, of course, but it guzzles way more power at it; e.g. in Handbrake x265 it is just a tad faster but way less efficient
- you present a binary take both on gaming and productivity, all-or-nothing approach, which is not how real world works
The 3D V-Cache is also only on one of the R9-7900X3D's two CCXs which means that it's essentially a hexacore gaming CPU that can experience scheduling issues. Now, at $329, at least its price is compelling, because nothing else about it is.
- price is compelling, I agree; actually, very compelling for the product, the best I have seen so far
- "nothing else" is not correct, as explained above; in gaming it's very close to top CPUs, especially in higher resolutions, and in productivity it is miles faster than 7800X3D, with 4 more core, and much cheaper than higher SKUs, a good choice for someone who does productivity, such as media encoding, without needing the top SKU
- I agree that there are a few scheduling issues in some games; those have been identified and measured, and it's up to buyers to look at it, whether this affects them. For example, I do not play any of those games, so the issue does not concern me at all.
Whoever made the decision to produce this CPU and price it the same as the R9-7950X instead of creating twice as many R5-3600X3D's should have had their employment at AMD terminated because it was literally the worst thing that they could do. High prices on AM5 parts literally crippled its adoption rate and this was only compounded by the fact that, until the R7-7800X3D was introduced, there weren't any really compelling CPUs for AM4 owners to upgrade to, because the R7-5800X3D is so damn good.
- this R9 SKU had lower volume from the outset due to its nature, being produced mostly from chiplets with one or two underperforming cores
- in many places, this chip arrived several weeks later in lower volumes; I looked at this across Europe at that time
- you forget that AMD decided to sell both platforms and they were aware that AM5 would only gradually pick up in adoption, which is exactly what happened over time. This particular chip has nothing to do with wider macroeconomic situation in the post-pandemic world where general PC sales decreased across the board around the time the new platform was released
- Zen4 X3D CPUs are not meant for vast majority of AM4 users who already have 5800X3D. Those users are good until Zen5 or later.
When you're a chipmaker that has adopted the practice of long-lived platforms (like AM4 is and AM5 is supposed to be), what you want is to get as many consumers on board as you can because they become a captive audience. I would estimate that over 90% of AM4 owners had the sense to stay with AM4 the whole way because being able to drop a CPU like the R7-5700X3D or R7-5800X3D into a motherboard from 2017 and get fantastic gaming performance without needing a new board or RAM is a huge boon for most consumers. This strategy not only brought AMD back from the brink of insolvency, it turned them into the market leader in less than five years, a turnaround that has never before been seen in the industry. If you had told me just ten years ago that AMD would be the CPU market leader after Intel had held a stranglehold on the CPU market for over 30 years, I would've asked you what drugs you were on (and where could I get some).
- completely agree with this. Nothing to add.
All AMD had to do to make AM5 an unmitigated success was to limit the 3D V-Cache to Ryzen 5 and Ryzen 7 models. Their pricing strategy was that an X3D CPU would be the same price as the X CPU from the tier above like this (with the exception of the R9-7950X3D of course because it was already the top-tier):
$R9-7900X3D = $R9-7950X
$R7-7800X3D = $R9-7900X
- I agree that they could have considered releasing 7600X3D together with the other three SKUs.
- but, would it really make a difference? We might never know.
- 7600 and 7900 classes of CPUs target different audiences anyway, so the existence of 7900X3D did not prevent AMD from targetting more mainstream gaming market with lower SKU that they did not release.
- you cannot blame the existence of 7900X3D by saying that they did not release 7600X3D. Two different arguments there
- Hardware Unboxed measured '7600X3D' performance on disabled chiplet of 7900X3D in recent video and it's only ~12% faster than 5800X3D, which does not warrant a complelling upgrade from AM4 to AM5 for R5 users when 5800X3D already exists.
So, all AMD had to do was this: $R5-7600X3D = $R7-7700X and their adoption rate would have increased by a minimum of 50% with a doubling of the adoption rate not out of the realm of possiblity. Even if they didn't make huge profits initially, they would know that the users who adopted AM5 would be buying only AMD CPUs for the life of the platform. It's like Sony's brilliant tactic of making no profit (or even taking a loss) for every Playstation sold because they knew that someone who bought a Playstation would have to buy games (which is where they raked in massive profits). One cannot deny the success of Sony's method when one considers that the Playstation is the undisputed console king, holding a 57.5% share of the US console market.
- I would not bet my horses on doubling of the adoption rate if ~12% faster 7600X3D existed, considering the cost difference between entire AM4 and AM5 platform and popularity of 5800X3D
- any transition to a new platform where the older one is still so successful would always be a gradual, long process, which it is
- in this case, initial AM5 offerings were the victim of on-going AM4 success. We can't have it all - both longevity and quick adoption of new platform. Life does not work like that.
- 7900X3D is a tiny little piece of much wider picture. If they had not released it, it would not have changed much the global picture about the competition between the two platforms
- it's a good problem to have after all
An R5-7600X3D for the same price as the R7-7700X would've been a guaranteed home run for AMD. Instead they decided to bunt with the Ryzen 9 X3D CPUs (especially the R9-7900X3D) and this is the result, their attempted cash-grab is sitting on the shelves gathering dust just like the RX 7900 XT, RX 7700 XT, RTX 4060 16GB and RTX 4080 did.
- as said above, this was not set in stone as minor gains of 7600X3D over 5800X3D would not necessarily motivate people enough to switch to more expensive platform and invest into all new components. ~12% uplift does not sound that inviting.
- I'd argue quite opposite. R5 buyers are the group that is most price sensitive and minor performance gains in gaming of ~12% on '7600X3D' over 5800X3D would not have looked too attractive to motivate them to spend more on a new platform.
- that's probably the main reason why they did not bother with 7600X3D so far, but went straight to 7800X3D that does look significantly better
- again, 7900X3D has nothing to do with this battle of perceived improvements in lower segment
- so, I consider this 'crusade' against 7900X3D an unhelpful rant, scape-goating a CPU that has not done anything wrong to anyone
- like 5900X, 7900X3D does not need to be the best in either gaming or productivity, and it's not its job to be that. It's just another SKU as option for more niche group of thousands of buyers who already have it around the world and simply enjoy it.
- it's not even a victim of success of other AMD's SKUs; a good problem for AMD to have, as they can always adjust production volume depending on what sells more and what sells less
- I'd buy it now, if I needed a new system, but my upgrade is either Zen5 or Zen6, or back to Intel if new i7 turns out to be decently performant and efficient.