3500X anyone?

[porina]

 

I don't personally know about this twitter user, but they have posted stuff on unreleased CPUs in the past. I haven't been keeping count on accuracy. So, salt on standby as usual.

 

So what's a 3500X? 6c6t, no SMT. 3.6 base, 4.1 boost. This compares with the 3600 at 3.6 base, 4.2 boost. So, pretty similar but without SMT. Same TDP. No indication of price. How much cheaper would it have to be to make sense below the 3600? Personally I have uses that do not benefit at all from SMT, so this could be cheap cruncher material.

 

The link in that tweet appears to show a listing where the 3500X is priced about 70% that of 3600.

[Fasauceome]

2 minutes ago, porina said:

How much cheaper would it have to be to make sense below the 3600?

it would have to be prices around the same as a 2600, in my opinion. That way you still get it within a certain value bracket and it can be easily compared to another unit in the lineup. Since I still like to recommend the 2600 it's good to have something that would resemble a direct replacement, and once mid tier motherboards like B550 show up it would be an outright replacement in my book, for a gaming PC anyway.

[5x5]

Has to be cheaper than a 9400F imo

[Princess Luna]

5 minutes ago, porina said:

How much cheaper would it have to be to make sense below the 3600? Personally I have uses that do not benefit at all from SMT, so this could be cheap cruncher material.

In all honesty it's a pointless product... I would rather see the 3600 not include a box cooler and be another 10 bucks cheaper.

 

Just because some software may not benefit from SMT directly always have in mind the system as a whole and on what multi-tasking is concerned SMT will always be useful, I actually love how AMD has it across the board and I'm hoping for it to force Intel quit overcharging for HT on their end.

 

But either ways without knowing pricing it's hard to say if this product will be any relevant or not, having in mind we have stuff like the R5 2600 for this price point albeit I'm aware the single thread performance ain't as good.

[Jurrunio]

$125 to give the $140 9400F a run for its money?

[MeatFeastMan]

3600 is the gaming king. I think the majority of people would rather pay a small amount more for the 3600. I don't think doing a 6c 6t part is a good idea. If they read the reviews of the 3600 vs 9400f, all the reviewers say the same: 9400f has a tiny bit better gaming performance BUT the lack of threads means stutters and therefore the 3600 is absolutely worth paying more for.

 

And it'd be pitting itself against the 2600. I'd recommend taking the single core hit and getting the 2600 every day of the week over a 3500.

 

 

[porina]

2 minutes ago, Princess Luna said:

Just because some software may not benefit from SMT directly always have in mind the system as a whole and on what multi-tasking is concerned SMT will always be useful, I actually love how AMD has it across the board and I'm hoping for it to force Intel quit overcharging for HT on their end.

Personally I think the value of extra threads from HT/SMT is somewhat overplayed. At the same clock, a 6c6t part would beat a 4c8t part in almost all scenarios, so maybe see this as fitting in there. Future lower Zen 2 parts might only be 4c8t but half the L3 cache due to one CCX, although would be sweet paired up as APU.

 

2 minutes ago, Princess Luna said:

But either ways without knowing pricing it's hard to say if this product will be any relevant or not, having in mind we have stuff like the R5 2600 for this price point albeit I'm aware the single thread performance ain't as good.

In case it was missed, I edited in that the tweet linked to what appears a web store, and the pricing was about 70% that of 3600. 

[Fasauceome]

11 minutes ago, Princess Luna said:

I actually love how AMD has it across the board and I'm hoping for it to force Intel quit overcharging for HT on their end.

i mean, if AMD charges just as much for a 6 core without SMT as they would for a 4 core with SMT, not much difference.

[porina]

https://www.tomshardware.com/news/amd-ryzen-5-3500x-intel-core-i5-9400f-budget-gaming-cpu,40434.html

 

Just saw Tom's posted it too, enough for a news post? Meh, I'll let someone else repost as news if they feel like it.

[Princess Luna]

21 minutes ago, porina said:

a 6c6t part would beat a 4c8t part in almost all scenarios

Why yes, "virtual" threads will never replace cores, but I like comparing 6c/6c vs 6c/12t rather... this is not a question about cores vs threads but simply if it's worth losing SMT/HT on the SKU.

 

While your comparison should be accurate as we seen the i5 8600K is indeed faster than the i7 7700K for the most part have in mind when we talk about the i7 8700K vs the i7 9700K then there's practically no difference between the two.

 

So where do we draw the line? Being honest this is a feature that should be present always specially nowadays where silicon is good enough for it to be basically a "switch on/off" for AMD and Intel... there really aren't yields/silicon purity issues that prevents giving it to consumers and SKUs that lacks it are solely for the product segmentation.

[sreejithRJ]

in my region india the ryzen 3600 is priced equivalent of 250 dollars and the ryzen 3500 is priced on 170 dollars

so which is the value for money