Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
99% will probs go with new 9xx series, but gonna w8 on new mobo releases, ain't that long (maybe x3d be close then)
Isn't it the case the only difference with the new mobos is mandatory USB4, of which there are currently no USB4 periperals available? If so probably no reason to get one and X670 and below might become a little cheaper
I'm getting that upgrade itch but I want to only do a single mb/ram purchase for the whole DDR5 life cycle. Looking at 96GB ddr5 (2x48) but unsure if its worth splashing out more on higher speed ram. I do prefer just to set XMP (EXPO now?) rather than fiddle too much.
Worth spending more to grab some 7200+ just in case and set it to 6400 for the 1:1 ratio until the 10k series come out?
But I don't like the thought of looser timings like CL40 rather than 32. Does reducing the MHz also reduce timings on these faster ram modules automatically rather than manual fiddling?
We can only speculate but how high do you chaps think the 1:1 infinity fabric ratio will go on AM5?
Rumours currently have it at 6400MHz for the next gen.
I'm getting that upgrade itch but I want to only do a single mb/ram purchase for the whole DDR5 life cycle. Looking at 96GB ddr5 (2x48) but unsure if its worth splashing out more on higher speed ram. I do prefer just to set XMP (EXPO now?) rather than fiddle too much.
Worth spending more to grab some 7200+ just in case and set it to 6400 for the 1:1 ratio until the 10k series come out?
But I don't like the thought of looser timings like CL40 rather than 32. Does reducing the MHz also reduce timings on these faster ram modules automatically rather than manual fiddling?
New benchmark scores for the 16-core AMD Ryzen 9 9950X processor were revealed, showcasing its performance enhancements over previous models. The 16-core processor, identified as "100-000001277", utilizes two Zen 5 CCDs and one IOD, supporting 16 cores and 32 threads. This unit is backed by 80MB of total cache memory, comprising 64MB of L3 and 16MB of L2 cache, and operates within a thermal design power (TDP) of 170 watts. The Ryzen 9 9950X has a base clock speed of 4.3 GHz with the potential to boost up to 5.7 GHz. This boost clock matches that of the Ryzen 9 7950X, though the base clock is decreased by 200 MHz.
The processor's performance was assessed using GeekBench version 6.3.0, where it achieved a single-core score of 3359 and a multi-core score of 20550. The testing platform included the ASUS ROG Crosshair X670E HERO motherboard paired with 32GB of DDR5-6000 memory in a dual-channel configuration. In comparison to the Ryzen 9 7950X, the Ryzen 9 9950X exhibited a 14% increase in single-core performance and a 7% rise in multi-core performance. When compared to the Ryzen 9 9900X, which is a 12-core model, the 9950X was slightly slower in single-core tests but showed a 4% improvement in multi-core tests.
full article here https://www.guru3d.com/story/ryzen-...re-is-higher-than-ryzen-9-7950x-multicore-is/
Also, they only shown high end, expensive mobo on the new chipset so far, with bits added that haven't been available in current gen - all that means that will be expensive and above current ones in pricing, like w tier higher because you get more fluff (but not more performance).Isn't it the case the only difference with the new mobos is mandatory USB4, of which there are currently no USB4 periperals available? If so probably no reason to get one and X670 and below might become a little cheaper
I have seen no confirmation for that at all, just rumors at best.Biggest difference is that the x870e chipsets will support much faster memory at XMP/EXPO settings.
These work on current boards already just not 1:1 with CPU. That won't change on new chipset (reason stated below).How much of a performance difference it'll make remains to be seen, though I assume we'll see reviewers benchmark with 7200Mhz-8000Mhz DDR5 kits (or faster),
4 dimm slots not working today well are mostly the fault of the board (daisy chain topology of ram slots etc.) and not chipset's. Chipset has nothing to do with RAM, memory controller is inside the CPU for a long time now and is said to be identical in coming CPUs.which simply have no chance of working out of the box on EXPO/XMP on X670 boards with 4 DIMM slots.
You can achieve that already on same boards with new UEFI versions (MSI has quite a few they claim this will work on).We can only speculate but how high do you chaps think the 1:1 infinity fabric ratio will go on AM5?
Rumours currently have it at 6400MHz for the next gen.
I much prefer to set expo with a ram that's tested by MSI (as that's the vendor of the board I have now), where they also provide optimise timings in uefi tested by them with specific ram kit. Works great in my case, cut a lot of secondary timings by 50%+ and it's fully stable, on stock expo voltages. That's a bit of a matter of luck, I suspect. My CPU can for example only do -10 on all cores with undervolting and I can't be bothered to try and tweak that more - I just used the MSI preset for that, confirmed it's all 100% stable and be done with it.I'm getting that upgrade itch but I want to only do a single mb/ram purchase for the whole DDR5 life cycle. Looking at 96GB ddr5 (2x48) but unsure if its worth splashing out more on higher speed ram. I do prefer just to set XMP (EXPO now?) rather than fiddle too much.
IMHO no - there's no indication this will work 1:1 with new CPUs, 6400 seems to be the limit still. Ram controller in the CPU seems to be identical to current 7k series.Worth spending more to grab some 7200+ just in case and set it to 6400 for the 1:1 ratio until the 10k series come out?
You should be able to manually reduce timings with clock but it won't be as straight forward.But I don't like the thought of looser timings like CL40 rather than 32. Does reducing the MHz also reduce timings on these faster ram modules automatically rather than manual fiddling?
Maybe, but Geekbench's MT Score is meant be very poor with huge variations between runs and it scales really poorly too. In other words, not what we would expect from any multi threaded benchmark.this is not exactly stellar performance, those with 7000 series chips not worth upgrading tbh not for 14%-16% anyway. Multicore performance is even worse, assuming this might be driver issue?
New benchmark scores for the 16-core AMD Ryzen 9 9950X processor were revealed, showcasing its performance enhancements over previous models. The 16-core processor, identified as "100-000001277", utilizes two Zen 5 CCDs and one IOD, supporting 16 cores and 32 threads. This unit is backed by 80MB of total cache memory, comprising 64MB of L3 and 16MB of L2 cache, and operates within a thermal design power (TDP) of 170 watts. The Ryzen 9 9950X has a base clock speed of 4.3 GHz with the potential to boost up to 5.7 GHz. This boost clock matches that of the Ryzen 9 7950X, though the base clock is decreased by 200 MHz.
The processor's performance was assessed using GeekBench version 6.3.0, where it achieved a single-core score of 3359 and a multi-core score of 20550. The testing platform included the ASUS ROG Crosshair X670E HERO motherboard paired with 32GB of DDR5-6000 memory in a dual-channel configuration. In comparison to the Ryzen 9 7950X, the Ryzen 9 9950X exhibited a 14% increase in single-core performance and a 7% rise in multi-core performance. When compared to the Ryzen 9 9900X, which is a 12-core model, the 9950X was slightly slower in single-core tests but showed a 4% improvement in multi-core tests.
full article here https://www.guru3d.com/story/ryzen-...re-is-higher-than-ryzen-9-7950x-multicore-is/