rockfella
Ambassador of Buzz
Looking at your username you may extend the purchases for over 6 months. Rocket lake might launch by then so keep that in mind.
TheSloth
The Slowest One
Ryzen5 3600 officially doesn't support memory speed over 3200, so I was planning to buy 3200MHz sticks. And since I am never going to overclock, going with ASRock B450 Steel legend seems fine. This board was suggested here in almost all the builds with R5-3600 whenever it was in stock if overclocking was not involved. So It should enough for just gaming on stock CPU speeds. Not going to spend over 11k for motherboard in a build with last gen CPU/GPU.
Guys I need this advice before the prices go up again. Should I get this board now or hold the purchase until I have enough money splurge altogether?
omega44-xt
Gear up ...
Depends on the price. It's great for 7-7.5k if you are on a tight budget, otherwise, get Gigabyte B550M DS3H, has been going for as low as 9k at times. No point going for B450 mobos above that IMO
Save money & buy stuff together, at least the basics. Always buy CPU+mobo+RAM together IMO. If you have old PC parts, use them for some time & slowly replace stuff like PSU, GPU, cabinet & monitor.
rockfella
Ambassador of Buzz
Don't worry. You have this entire century to wait and buy.
Not sure but do you already have a Ryzen 3600 or will you buy one? If not consider Intel i3 10100 9k and GIGABYTE B460M Gaming 7.5k. Don't worry too much about RAM speeds.
Last edited by a moderator:
nac
Aspiring Novelist
Save and buy all the three (cpu, board, ram) in one go. No pointing in buying just the board now and wait for other components for months.
TheSloth
The Slowest One
I understand but I need 6c/12t processor since I will use this PC not only for gaming but other works also and it will stay minimum for around 6-7yrs. 4c/8t won't suffice my usage.
@TheSloth @bssunilreddy @omega44-xt @SaiyanGoku I wanted to post about this earlier but it slipped my mind. I recently assembled my new pc using ryzen 1200 & asrock AB350 Pro 4 with 3200MHz crucial ballistix ram. By default the system booted at 2400MHz & I selected xmp profile to 3200MHz which system then successfully applied. Everything was seemingly running fine for 1-2 days when I remembered an old post on reddit about checking the stability of ram overclock by running a prime95 blender test for at least 8 hours so I thought of giving it a try & within 20 min of starting the test it gave an error. I then spent the next day running this same test for at least 8 hours on all ram frequencies available to select in the list of options to set ram frequency while choosing xmp profile & finally the max speed at which this test ran without errors for more than 8 hours was 3066MHz.
tl;dr or In short, do not just set & forget xmp profile to your ram stated frequency/max frequency supported by your mobo after your system boots fine & seems to be running as usual. Run Prime95 blender test for at least 8 hours & only when it give no errors you should keep the ram running at that frequency.
tl;dr or In short, do not just set & forget xmp profile to your ram stated frequency/max frequency supported by your mobo after your system boots fine & seems to be running as usual. Run Prime95 blender test for at least 8 hours & only when it give no errors you should keep the ram running at that frequency.
SaiyanGoku
kamehameha!!
Userbenchmark is total BS. Don't use them for doing any comparisons or giving recommendations.
omega44-xt
Gear up ...
That's some high-level BS. An i3 10100 or even i5 10400 is inferior to R5 5600X in all respects except price.
RumbaMon19
Feel Pain.
That website (Userbenchmark) is not at all reliable. It is not only sh#t in PC CPU, But also gives wrong info about mobile processors, in some instances it said exynos 9611 to be better than sd730 and so. I think that is based on votes given by users and not on any "Benchmark".
Nerevarine
Incarnate
User based benchmarks would be doubly stupid for games because there is no way to keep rest of components same.
Ex -> Comparing i5 8400 to Ryzen 5, but i5 wala has GTX 1080 and Ryzen 5 wala has 1070. Obviously it would lead to discrepancy. User benchmarks can work if its only CPU specific benches like 7zip encryption test. But even then there are a lot of issues like someone Oced, someone underclocked etc etc..
Ex -> Comparing i5 8400 to Ryzen 5, but i5 wala has GTX 1080 and Ryzen 5 wala has 1070. Obviously it would lead to discrepancy. User benchmarks can work if its only CPU specific benches like 7zip encryption test. But even then there are a lot of issues like someone Oced, someone underclocked etc etc..
chimera201
Wise Old Owl
Userbenchmark is only good for a quick check to see if your components are working as expected. It plots a graph of your score for each component versus other users. The graph is the useful thing, the score itself is pointless. Rest of userbenchmark is useless.
SaiyanGoku
kamehameha!!
It still is useless. Run Cinebench and various tests from 3dMark instead.
Review sites aim to minimize bottlenecks in review so a 1650Ti running with an i7/R7 with latest NVMe & 32gb ram ensures that the score/review represents the best performance a 1650Ti can have in any system & then compare it with RX5500XT in a similar way ensures that comparison is fair(best performance of 1650Ti vs RX5500XT).
skeletor
Chosen of the Omnissiah
He is comparing them to 3600, not 5600X.
I doubt it as vaccine manufacturers have already claimed that vaccines are effective against new variants too.
TheSloth
The Slowest One
Yeah but after reading how several countries banned travel to UK even though UK is already among countries which have started the vaccination, I am worried that my extended wait will be extended even more now.