Unusual actions observed while upgrading.
Unusual actions observed while upgrading.
So while I was waiting on AMD to release hopefully a 7000 series 3d, I bought a 5500 because it was 120$ and in stock at work. Same day I brought it home and updated my Bios from 3.50 (Latest for my 2700x) to 4.70, which was the last stable version of bios my x470 Taichi MB has. Install went fine, but trying to boot I noticed it never fully went through all of the Dr debug codes like it should during the 3 restarts, it got hung up on the 2nd one, with the error 95. Tried to remove everything possible, cleared Cmos, same error. Tried all combos of ram I can do. Put in my old 2700x, and for some reason it boots? I was told this from Taichi's page, Do I assume that my bios update went poorly? or that I should just return the CPU? Pic of Bios and cpu Also Should I go back to the old bios 3.50 for my 2700x in the meantime?
The difference between 5500 and 2700x lies beyond basic performance gains; it represents a meaningful leap in capabilities. Are there additional improvements beyond speed that you haven’t noticed?
I was checking if using Smart Access Memory or a newer CPU would boost my 6900xt’s performance. My 2700x had some stability issues, and I’ve heard XMP settings help. The 5500 supports faster RAM too. It was available at work, so it felt like a quick upgrade. Maybe the RDNA 2 changes were more than just the usual 12-14% gains—overclocking seemed to offer extra benefits beyond what’s advertised.
It was just a coincidence I hadn't considered SAM, and I didn't mention the GPU either. Keep going, and good luck finding the answer.
Inspect the 5500 for bent pins—could be a small mistake. It’s close to a 3600, offering improved gaming performance compared to a 2700X. A 5700X would be an even stronger upgrade. Alternatively, the 5800X3D handles Zen 4 games well.
5500 sits between 3600 and 5600 in performance. Zen3 cores outperform Zen2 cores, though slower Zen2 still edges out faster Zen3 (which tops out around 4.6–4.8g). The cache hurts its edge, so it falls short against the 5600. This gap can be bridged by optimizing clock speeds—around 2500 fclk is achievable on these Cezanne monolithic ryzens. A 5500 might hit close to 5000 fclk with a 8Gbit rev e/j setup, using ballistix or 2666c19 bare PCBs, or solid DDR3L variants like those in 3600+ builds. You’d need to trade some RAM capacity and cooling for that boost. For multicore performance, it should be on par with the 2700, but it loses in singlecore. If you want more cores, consider a beta BIOS; otherwise, stick with single-core if that’s your goal. Right now, testing a beta BIOS would be wise to confirm support for the 5500.
I could possibly attempt that tomorrow night if I don’t have to return the CPU. Considering whether I’d just try to secure an AIO using the refund or purchase another 5000 series to see if the problem persists.
Aio is useless unless you require cooling. You should consider a 5600 if it’s not significantly pricier than the 5500 (over $40 more). You won’t really benefit from that massive iMC on a top board—top boards don’t reach such high clocks as a daisy chain, and Ryzen iMC performance drops with more ranks. The 5600 offers better cache, so you don’t need to push the RAM too high.
Follow the AGESA 1207 instructions to install the 5500. The 95 Dr Debug code is available—have you already tried it?