This is no different than GPU manufacturer's launches. AMD's RX 7900 XT & XTX was released in December and the RX 7700 and RX 7600 series in January. Nvidia does the same with its 3000 and 4000 series, though the 4000 series gets a little muddled with the whole 4080 err 4070ti naming debacle.
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
Ryzen 7xx0 3D V Cache or No?
- Thread starter Niku-Sama
- Start date
Maybe I missed something, but RX 7600/7700 wasn't released yet.
It's always like a high version is released first, usually 2 models like RTX4090 and RTX4080 (earlier 3090/3080). Then is some break like 2-3 months, and a lower model, so this time RTX4070Ti (supposed to be more, but Nvidia failed specs, so was only one). The lowest is usually sometime later, like a half year waiting for RTXxx50/60.
With CPUs is about the same, so we had Ryzen 7950X/7900X/7700X/7600X in the first round ... next the same but without X ... and now was expected the whole X3D, there are only 3 CPUs, so I'm not sure why only 7800X3D is delayed by 2 months. It's not like they make profits on that move.
OP
- Joined
- Jan 13, 2005
- Thread Starter
- #24
honestly it doesn't hurt my feelings at all, that way they can get the bugs out of the mobo's that we know will happen and some numbers to see if it is really worth the money and time. i think it'll be good but you dont know truly until they are out.
and i am going to be honest here, i figured these were going to start at $599 and go up from there with the way everything has been lately. Surprised to see 7800x3d starting at $450
maybe some one finally came to their senses since the writings on the wall for next fiscal year already...
and i am going to be honest here, i figured these were going to start at $599 and go up from there with the way everything has been lately. Surprised to see 7800x3d starting at $450
maybe some one finally came to their senses since the writings on the wall for next fiscal year already...
@Niku-Sama let's not get the cart before the horse. We're still 3 weeks from the launch date and we have seen other products announce an MSRP of $XXXX but then when it comes time to buy its nowhere to be found for that price. I admit this was primarily with GPUs but if memory serves me correctly it happened with early DDR5 too.
I agree these are decent MSRP prices, let's just remember what the "S" stands for. I'm praying that you are right but my intuition tells me otherwise.
I agree these are decent MSRP prices, let's just remember what the "S" stands for. I'm praying that you are right but my intuition tells me otherwise.
While I think these will sell well, I'm not convinced they are going to be scalped and rare expensive birds.
What happened with ddr5 and that timing (supply chain disruptions, plant disruptions, brand spanking new item to the market) is different than these chips.
What happened with ddr5 and that timing (supply chain disruptions, plant disruptions, brand spanking new item to the market) is different than these chips.
It means something for gaming, but slower in some things, so there that. That said, the performance 'penalty' wasn't much.
...saying people dont know the benefits lends credence to the idea these won't be unicorns and scalped to high heaven.
Considering MSRP, in my local stores, inc. tax 7800X3D will cost me about ~$600. I couldn't wait until April, so I bought Ryzen 7600 (non-X). It cost me ~$250, and maybe it will be enough, and I won't buy 7800X3D at all. In the worst case, I will sell it in a couple of months. I have 7950X for tests anyway, so I can live without a 3D cache chip, even though I wish to play it a bit and test some scenarios.
Last weekend, I checked that my Ryzen 7600 runs at 5.4GHz 1.23V on all cores (AVX tests) and is stable in Cine R23 in the loop, without throttling, on Noctua NH-L9a-AM5 (so the smallest cooler available). The CPU goes up to ~99°C, but it has 95W package wattage after OC. I just doubt I will need anything better for this build, so maybe AMD's move just saved me ~$350.
Edit:
A bit off-topic with the Ryzen 7600 results, but I thought I would leave it here.
Last weekend, I checked that my Ryzen 7600 runs at 5.4GHz 1.23V on all cores (AVX tests) and is stable in Cine R23 in the loop, without throttling, on Noctua NH-L9a-AM5 (so the smallest cooler available). The CPU goes up to ~99°C, but it has 95W package wattage after OC. I just doubt I will need anything better for this build, so maybe AMD's move just saved me ~$350.
Edit:
A bit off-topic with the Ryzen 7600 results, but I thought I would leave it here.
Last edited:
OP
- Joined
- Jan 13, 2005
- Thread Starter
- #32
yeesh PBO only OC. thats kinda crappy.
i hope some mobo manufacturer will "accidentally" make a board that has unlocked multi's for them. its happened before.
its sounding like more and more the temp is causing instability on the vcache die hence the limited speeds and multi's. and more and more it sounds like a 360mm rad would fix that problem.
i'm curious how limited the memory controller is going to be on these too
i hope some mobo manufacturer will "accidentally" make a board that has unlocked multi's for them. its happened before.
its sounding like more and more the temp is causing instability on the vcache die hence the limited speeds and multi's. and more and more it sounds like a 360mm rad would fix that problem.
i'm curious how limited the memory controller is going to be on these too
If you know how to set PBO, then you can get the next 200-400MHz. Manually, you won't get much more anyway. Most Ryzen 7000 CPUs can't make more than 5.2-5.4GHz set manually for all cores. X3D is expected to run at 5.0GHz, so it will probably run at up to 5.4GHz with optimal PBO settings and low enough temps. Regular cores will boost up to ~5.8GHz.
There was no way to unlock OC on 5800X3D, so I assume it's just locked on the CPU level.
There was no way to unlock OC on 5800X3D, so I assume it's just locked on the CPU level.
Without the manual settings or bclk you may be stuck at +200MHz like on most other chips. On Ryzen 7600, I have up to 5.15GHz at auto, and PBO lets it go up to 5.35GHz, even though the CPU can make more. It's regardless of temps or anything else. I can set only +200MHz max boost, and all other options are useless (curve and other things are not helping in anything). When I run Ryzen Master auto OC then it reaches 5.35GHz on all cores, not like with higher CPUs that it sets higher or lower frequencies depending on the core capability.
Considering the above, you can count that cores with V cache will go +200MHz, and all other cores will boost much higher, up to about 5.9GHz (depending on the CPU).
I don't think you will need more than 5.2GHz for anything you do.
Edit:
I'm not sure if X3D chips will have it unlocked, but you may think about a motherboard with asynchronous bclk (one for CPU and one for everything else). Higher ASUS mobos have it, like the Crosshair series. Then you can set much higher bclk and you are not limited with ~103MHz.
Considering the above, you can count that cores with V cache will go +200MHz, and all other cores will boost much higher, up to about 5.9GHz (depending on the CPU).
I don't think you will need more than 5.2GHz for anything you do.
Edit:
I'm not sure if X3D chips will have it unlocked, but you may think about a motherboard with asynchronous bclk (one for CPU and one for everything else). Higher ASUS mobos have it, like the Crosshair series. Then you can set much higher bclk and you are not limited with ~103MHz.
Last edited:
I think these will be like the current X3D line, just as locked down. But of course you get the higher core counts, clocks and IPC to go with it.
Current X3D does have PBO and CO, just like they said the new X3D will have. But they didn't say anything about boost override.. or did they and I missed it?
Current X3D does have PBO and CO, just like they said the new X3D will have. But they didn't say anything about boost override.. or did they and I missed it?
- Joined
- Aug 14, 2014
Gamers Nexus has entered the thread 
"This video benchmarks and tests various points of failure for the AMD X3D CPUs in particular, but also just generally AM5 / Ryzen 7000 (Zen 4) CPU failures. The most eye-catching failure mode is catastrophic: Users have posted photos of exploded CPUs with blown-out motherboard sockets and melted plastic. We bought those parts from the users to perform a failure analysis, then set forth testing our own AMD Ryzen 7 7800X3D CPUs to see if we could make them explode. Spoiler alert: It worked. We were able to melt new AMD X3D CPUs in the socket of an ASUS X670E Crosshair Hero motherboard, but we also found severe BIOS bugs in Gigabyte's BIOS. SOC voltage isn't the only concern, here. It's also poor OCP implementation on at least the ASUS board we used, plus a combination of the erosion of thermal and over-current protections. As stated in the video, no test of this type is fully conclusive: These are complex issues with functionally infinite variables between users (and possible error), manufacturers, BIOS versions, and AMD itself. As such, we can only speak to our findings in some instances -- but we have some that do point to specific issues.
We killed 2x 7800X3Ds and 1x ASUS motherboard, but we also had 2 failed CPUs and 1 failed board from viewers."
"This video benchmarks and tests various points of failure for the AMD X3D CPUs in particular, but also just generally AM5 / Ryzen 7000 (Zen 4) CPU failures. The most eye-catching failure mode is catastrophic: Users have posted photos of exploded CPUs with blown-out motherboard sockets and melted plastic. We bought those parts from the users to perform a failure analysis, then set forth testing our own AMD Ryzen 7 7800X3D CPUs to see if we could make them explode. Spoiler alert: It worked. We were able to melt new AMD X3D CPUs in the socket of an ASUS X670E Crosshair Hero motherboard, but we also found severe BIOS bugs in Gigabyte's BIOS. SOC voltage isn't the only concern, here. It's also poor OCP implementation on at least the ASUS board we used, plus a combination of the erosion of thermal and over-current protections. As stated in the video, no test of this type is fully conclusive: These are complex issues with functionally infinite variables between users (and possible error), manufacturers, BIOS versions, and AMD itself. As such, we can only speak to our findings in some instances -- but we have some that do point to specific issues.
We killed 2x 7800X3Ds and 1x ASUS motherboard, but we also had 2 failed CPUs and 1 failed board from viewers."
In short, they did that on purpose to make noise on the web as most brands released BIOS (even as slow as Biostar) to limit voltages a couple of days ago, and they had to be aware of that, considering they were showing the AMD statement and info about BIOS updates.
As far as seeing that is always somehow funny (when $1k+ components are not yours), then I feel they try hard to impress teenagers.
As far as seeing that is always somehow funny (when $1k+ components are not yours), then I feel they try hard to impress teenagers.
