Pasi123

The B550 Gaming Plus should be able to handle any AM4 CPU with the latest BIOS.
 
As for which CPU to get, it depends on your budget and what you're doing with the computer.
 
If your focus is gaming, then the CPU to get is either the 5800X3D or 5700X3D. They have the best gaming performance on the platform by a good margin - they're basically as fast for gaming as the 7700X.
 
If your focus is productivity, you'd want a Ryzen 9 5950X or 5900X. They have the most raw compute of any CPUs on AM4 - much more than your 3600. While not as fast for gaming, in most situations, they're perfectly sufficient for an RTX 4070.
 
Finally, if your budget is tight. You can consider the 5700X. It should be cheaper than the rest of these options while giving the same gaming performance as the Ryzen 9 parts and about 50% more productivity performance than you have now.

Phenom II X6 1100T @ 4GHz+ and GeForce GTX 580's in SLI. That's what was available in December 2010 when the 1100T was released.

Depends on if the case comes with a RGB controller or not. If not then you could buy a controller that connects to an internal USB 2.0 header.
The motherboard itself predates ARGB by multiple years so it doesn't have an ARGB header.

Depends on if the case comes with a RGB controller or not. If not then you could buy a controller that connects to an internal USB 2.0 header.
The motherboard itself predates ARGB by multiple years so it doesn't have an ARGB header.

Depends on if the case comes with a RGB controller or not. If not then you could buy a controller that connects to an internal USB 2.0 header.
The motherboard itself predates ARGB by multiple years so it doesn't have an ARGB header.

Consider this pre-testing to a wider IPC test. I ran Cinebench R15 in various configurations on 3 CPUs representing 3 different generations.
12100F - Alder Lake, Intel 7 (formerly 10 Enhanced SuperFin) 11700k - Rocket Lake, 14nm 7980XE - Skylake-X, 14nm Rocket Lake was the first desktop generation past Skylake architecture even though it was still made on 14nm process, and Alder Lake finally took desktop beyond 14nm nodes.
 
Why use Cinebench R15? Mainly because it is well known, and as older software it does not make use of AVX instructions so takes away an element of complication. It is known to not be much affected by memory performance which is good if I'm focusing on the CPU cores itself. I note the score, reported CPU power consumption, and where possible the CPU clock. In some scenarios with a power limit, the clock varied too much to get a good value. Skylake-X here should be near enough a substitute for regular Skylake and derivatives (Kaby Lake, Coffee Lake, Comet Lake).
 
Note this is "quick testing" and I didn't try to eliminate all variables and check of repeatability beyond doing a minimum of 2 runs each.
 
Relative performance (IPC best case):
Skylake-X: 52.2 points/core/GHz
Rocket Lake: 63.3 points/core/GHz, +21% relative to SKX
Alder Lake: 72.7 points/core/GHz, +39% relative to SKX, +15% relative to RKL
 
Relative efficiency:
This is where it gets really complicated. It depends on where on the efficiency curve you compare, and these are wildly differing CPU configurations. I used a metric of points/W for comparison.
 
7980XE unlimited (188W): 16.6 points/W
7980XE turbo off (117W): 20.9 points/W
11700k unlimited (183W): 12.7 points/W
11700k turbo off (88W): 20.4 points/W
11700k 45W limit: 22.5 points/W
11700k 25W limit: 23.1 points/W
12100F unlimited (56W): 20.3 points/W
12100F 45W limit: 23.9 points/W
12100F 25W limit: 30.2 points/W
 
I decided against testing the 7980XE at lower powers since it was reporting around 45W usage at idle! 
 
We still have different core configurations. I tried adjusting the 11700k further, running it with only 4 cores enabled (4c8t), and also running it with all cores and HT off (8c8t).
 
11700k 4 cores 8 threads
125W limit: 10.1 points/W
45W limit: 18.3 points/W (Alder Lake 30% more efficient)
25W limit: 18.6 points/W (Alder Lake 62% more efficient)
As generally expected, this is much worse than running with 8 cores. While each core has more power available, it runs in a less efficient area. Alder Lake does seem to scale better at lower powers. Note although is is 4 core vs 4 core, by disabling 4 cores of Rocket Lake it may not scale exactly. Intel did not make a 4 core Rocket Lake.
 
11700k 8 cores 8 threads (HT off)
Unlimited (159W): 11.1 points/W
45W limit: 18.5 points/W
Cinebench R15 is on the higher end of HT scaling workloads. For a 45W power limit, it is doing 18% less work than with HT on. While it was not a part of my testing this time, previously I've seen typically around 30% more throughput at the same clock with HT, implying the difference here is due to the extra power consumed by HT. Arrow Lake is rumoured to not have HT so it will be interesting to see how that goes overall.
 
Overall we do see a bit more efficiency from the newer process, especially at lower powers. There is also a clear improvement in IPC between the generations. Of course, this is very limited in only looking at Cinebench R15, and I hope to expand this to wider workloads over time.

Honestly I can't foresee *anyone* needing PCIe5 SSD speeds in the next 5+ years. The spec seems to exist primarily for hyperscale data centers

That board doesn't have a M.2 slot but it does support NVMe boot so that is something that could be easily added with a PCIe card. Though like you said if it's running fine there is no need to upgrade. The difference between NVMe and SATA SSDs isn't that big in normal use.
 
Another thing mentioned in this BIOS update is support for a USB 3.1 (10Gbps?) add-on card, but for upgrading to that to make sense you'd need to have some USB devices that benefit from speeds faster than USB 3.0 5Gbps. Same thing for network cards, you'd need a network faster than 1Gbps for any 2.5Gbps or faster NIC to make sense.

you tell us......

name the thing you want updated.
one the harder things on the forum isn't fixing or upgrading computers...... it's trying to explain to the owner that all that matters is the machine does the tasks that it needs to do.

getting into the mentality that you HAVE to upgrade or stay current regardless of how well your current machine works is mostly just consumerist propaganda that companies have been pushing since the 1950s. The fear of being "not current" in tech.

Just use your computer. Update an aspect of it when theres a problem or a need.

Figure I'd bump the old thread with my recent X58 stuffs since LTT posted a video with the SR2, shame they didn't try setting affinity the same way all the new hybrid platforms have to.

 
I pulled the EVGA E758 out of the spare tower corner it's been living in since the board is all tweaked from having one of those Arctic plastic mount coolers on it for probably 15 years and I just can't get it to read the third memory channel. I can see the dimms there and connected but the system just doesn't use it. I picked up a new cooler that advertised 1366/775 support, both of which I have a few boards for. Of course, this was just an incorrect amazon listing. Whatever, it was $13 and my 212 LED that's going into the trash the next time I touch it because it's drawn more blood than any other PC part I've ever owned had a MIR to be $15. Whenever Noctua releases their new D15 cooler I'll give it another shot with the NH-D14 before it goes into the home server to see if a proper backplate resolves the issue.
 
Had to destructively remove the Arctic cooler because one of the pull clips snapped. New cooler obviously then didn't end up fitting so I tossed my stock i7 930 cooler on it and I really don't remember it running this cool with the i7, maybe I just have really bad memories of my Pentium D and I stuck a NH-D14 on it day 1. Temps look really decent on the X5670. It's still missing the third channel though, which is an issue my Gigabyte board had for a while but a reseat fixed. I might have to try it with an i7 again but with how bowed the board is around the socket area at the top edge I don't have high hopes.

 
I also stuck a GTX 970 into the old X58 setup, it's just slipping through on it's 550w PSU as it pulls almost exactly 500w under P95/Furmark at the wall. I'll call that build well optimized lmao. The SX8200 Pro that I had used in it but transferred to my personal Ryzen build has also returned after it experienced some file corruption which prompted me to check the old SM951 128GB. Upon getting a BTRFS scrub the drive decided it's to cease functioning entirely. I have it setup for wake on lan and I generally remote into it. I have noticed the random reboots starting to sneak in when it runs like an OBS stream for half a day but still pushing 4680mhz on single thread at 1.425v, I'll just blame it on latest 550 nvidia linux driver/nvenc which broke my suspend/wol as well. It usually isn't in the neglected tower storage corner but where it normally lives has some Haswell junk I was testing some TrueNAS Scale stuff on recently.

 
I keep looking around for worthwhile upgrades for X58 home server as well but most everything seems like an increase in power consumption or a steep decrease in available pci-e lanes. $20 used nvmes are just so tempting! LGA2011-3 stuff looks fun but I don't have a spare box of DDR4 just laying around like I do DDR3. Feature creep sets in and then IPMI would be great so I don't have to remove an nvme for a GT1030 to configure the boot order because the HBA decides it's more important after a few years and a power loss. Utilizing all of the pci-e slots and lanes right away isn't appealing. Newer built in board features and what do I really need vs. want. Splitting my VM off into a separate SFF box instead, oh look now I need a switch and 10g SFP+ sounds fun.
 
Ehhh, it's still doing it's job perfectly fine and it could probably use some larger drives again instead.

First off, what version of Cinebench are you running? There's quite a few different versions, and the scores from each of them don't line up with each other. The scores you listed sound like CB 2024 (this produces very low scores compared to the more prevalent R23, though your chip should still be doing ~750 multi core in 2024), though it can technically be something else. 
 
Second, what CPU are you actually running? The 3700XT doesn't exist, is that a typo for the 3700X or 3800XT? It won't make much of a difference, but it's still good to know. Also, is that the ASRock B550AM, and what CPU cooler are you running?
 
Third, since there's dozens of things that can cause this, going for the Hail Mary solution is probably for the best. Download HWInfo64 and launch it in "Sensors Only" mode (it'll show up in the dialog box if you want it sensors only or summary only). Once open, in the bottom right of the window there will be a button that looks like a piece of paper with a plus sign, hit that and run Cinebench. This will create a log of every sensor in the system that can be read later to see if there's something over heating, a power limit set way too low in the BIOS, the CPU not boosting correctly, etc. Send that file once you create it. 

Bottlenecks can't damage the machine

If your playing CIV6 at 240p then the bottleneck will be really bad.
If your playing hogwarts legacy at 8k then you won't have any bottleneck at all.
 
Bottlenecks exist in every machine. Sometimes they matter (depends on the games, the settings, the monitor, ect.), but usually (if the components were made at similar times) they don't. All a bottleneck means is that the performance of one of your components is limiting the performance of the others. Say your CPU can't do enough work to allow your GPU to reach its full potential or vice versa.
 
On a 4080+5900X machine you really don't need to worry about bottlenecks.

Please LTT, I beg you, do a video about bottlenecks to demonstrate "what a bottleneck is", "when it might matter" and "that bottleneck calculators are scams".

Techspot used Cinebench r23, which uses a different scoring system than Cinebench 2024, which is the version that you used.
You can download Cinebench r23 from the Microsoft store or other sources, Maxon doesn't offer it anymore though.

The iGPU in the 4th gen i3 is much better than the GeForce 210. The 210 is a display adapter from 2009 and uses a really low end GT218S chip from 2007

If you are happy with the performance of the i7-9700 I don't see any reason to upgrade now. GTA 6 release is still so far away we don't even know the exact release date let alone the system requirements.
Right now my guess would be that the i7-9700 can run GTA 6, but who knows if there will be a huge jump in CPU performance in the near future leaving the 9700 far behind.
 
GTA 5 was released for PC in 2015 and it was playable on a 2007 Core2 Quad Q6600 but not perfect as there was some pop-in when driving fast. First gen i7's like the i7-920 had no problems running the game. But that game released during a time when mainstream CPUs had been stuck on 4c/4t and 4c/8t for years.

We do not know anything about gta vi.
It’s fine in current games, can’t really tell much about the future. Sometimes you end up with a nehalem to kaby lake kinda thing where nothing changes for years. Sometimes you get generations like coffee lake and first gen ryzen which stir things up a lot. No way of really telling beyond speculation.

Blocking VP9 and AV1 with h264ify browser plugin might help a bit with video playback since the GT 520 has H.264 hardware decoding, if you have hw decoding enabled in your browser.
I personally wouldn't go with anything slower than a Q9550 2.83GHz or Q9650 3.0GHz, maybe a Q9450 2.67GHz. C2Q Q9650 3.0GHz would of course be the optimal one because it's literally two E8400's on one CPU.
I also wouldn't go with the Q6600 because it's older generation and lacks SSE4.1, and even at 3.0GHz it's only comparable to a Q9400/Q9450 2.67GHz.
 
Q9505, Q9500, Q9400 and Q9300 only have 6MB L2 cache while the Q9450, Q9550 and Q9650 have the full 12MB. The lower end Q8400, Q8300 and Q8200 are even worse with only 4MB cache. Though I'm not sure how much that affects real world performance but technically the 6MB and 4MB quads are a downgrade from the E8400 6MB when it comes to cache per core.
 
I still use a Q6600 @ 3.0GHz (with a Radeon HD 6570 and Ubuntu 18.04) in my HTPC and it plays 1080p YouTube videos fine when hardware decoding is enabled, and newer codecs disabled with h264ify. It is also usable without hardware decoding but the CPU usage is of course higher and there might be some dropped frames when playing 1080p60 videos. My other C2Q system with a Q9550 does a bit better and I don't remember it dropping frames even without hardware decoding

Blocking VP9 and AV1 with h264ify browser plugin might help a bit with video playback since the GT 520 has H.264 hardware decoding, if you have hw decoding enabled in your browser.
I personally wouldn't go with anything slower than a Q9550 2.83GHz or Q9650 3.0GHz, maybe a Q9450 2.67GHz. C2Q Q9650 3.0GHz would of course be the optimal one because it's literally two E8400's on one CPU.
I also wouldn't go with the Q6600 because it's older generation and lacks SSE4.1, and even at 3.0GHz it's only comparable to a Q9400/Q9450 2.67GHz.
 
Q9505, Q9500, Q9400 and Q9300 only have 6MB L2 cache while the Q9450, Q9550 and Q9650 have the full 12MB. The lower end Q8400, Q8300 and Q8200 are even worse with only 4MB cache. Though I'm not sure how much that affects real world performance but technically the 6MB and 4MB quads are a downgrade from the E8400 6MB when it comes to cache per core.
 
I still use a Q6600 @ 3.0GHz (with a Radeon HD 6570 and Ubuntu 18.04) in my HTPC and it plays 1080p YouTube videos fine when hardware decoding is enabled, and newer codecs disabled with h264ify. It is also usable without hardware decoding but the CPU usage is of course higher and there might be some dropped frames when playing 1080p60 videos. My other C2Q system with a Q9550 does a bit better and I don't remember it dropping frames even without hardware decoding

It is a dual core Excavator (one of the worst CPU architectures ever made) CPU with 15TDP, whatever you do you wont be getting much out of it.

Bottleneck calculators are bullshit
That will work fine, its a good combo.

Your need for a faster cpu/gpu varies per game, some games will push the cpu much harder than the gpu, others vice versa.

Bullshit there is the bottleneck calculator. Dont use them or trust them, that is a good pairing.

The W3175X build is slowly coming together, got the aforementioned 3175X in the post yesterday, most of my water bits should be here tomorrow if DHL is to be believed and my coldzero trays/bits are in production now 

I had decent luck overclocking my 7980XE on air, was able to get 4.4 on an AK620 so this should handle pretty well I'd reckon

I finished my X299 build few days ago.
 
CPU: Intel Core i9-10900X 10c/20t
Cooler: DeepCool Assassin III
Case: Fractal Meshify 2
Memory: 64GB Kingston DDR4 3200MHz (just 2x 32GB in dual channel for now)
Motherboard: Asus TUF X299 Mark 1
Graphics Card: Asus GeForce GTX 1080 Strix Advanced
SSD: 2TB Samsung 970 EVO Plus (OS and programs)
SSD: 2TB WD Blue SN570 (games)
HDD: 8TB Seagate Exos 7E10
Power Supply: Seasonic Focus GX-850
OS: Windows 10 LTSC 2021
 
(with a GTX 760 for testing)


 
GTX 1080 inside the case. The case has a dark tinted glass so pretty much the only visible things are the RGB LEDs on the GPU and motherboard

It is not okay. You should never use anything harder or more abrasive than plastic to clean off thermal paste - even plastic is pushing it. Paper towel or a coffee filters are all ideal for cleaning paste. Tissues or toilet paper can do in a pinch (they can leave behind large fibers, but as long as you clean off afterwards, it's okay).
 
And, if you absolutely must, you can use a plastic spudger like this as a last resort for stubborn paste:

 
Even then, you should be careful, because copper is a very soft metal - it scratches easily.
 
You cannot use a metal tool like a screwdriver to clean paste, because it will damage the surface of the cooler or CPU IHS, which is exactly what happened here.
 
Those surfaces are machined smooth and flat for a reason - you need good thermal contact, which means having the die and the cooler mount together as tightly as possible. Thermal paste is for filling in micro-scratches on the surface.
 
I hate to be the barer of bad news, your laptop's cooler is now badly damaged. It is not going to perform as well as it used to - thermal paste is way less conductive than copper, so your CPU is going to run a lot hotter now, as the thermal transfer needs to go through a thick layer of paste to reach the copper.
 
If you want to try to fix this, you could see if there are replacement heatsinks available for your laptop's model. eBay has plenty of laptops that are sold for parts, and sometimes you can find specific parts, like a cooler assembly, available for sale. That is the only way to truly fix this now: you need a brand new cooler.