Does the video have a summary graph?
this is not even the most pronounced. And we now have more powerful gpus too.
No, but instead watching you can read it:Does the video have a summary graph?
My opinion is that Zen3 is again a server first CPU core. Most of the improvements and features added as per the PPR are not even remotely going to be used in Desktop.Hm, somehow hoped for more details on the changes to the cores. Unified L3$ we knew for ages, and "Zero Bubble Branch Prediction" is the only new token brand we got to hear. What exactly are the improvements for Load/Store, and the changes to the issue width of int and float engines? I guess the event was squarely targeted at gamers, and have to hope wikichip or some other specialized site will get the info I'd like to hear about Zen 3.
AMD need to build a warchest.In AMD's case the MSRP always has been an early adopter premium. Honestly I consider the $50 rise rather low for what it is (and will wait for lower street prices as always ).
Do you even know how outliers are decided? Here have a read:
"The games I play most are not outliers" attitude is completely wrong.
I haven't read the whole thread so if this is answered already then my apologies.
Do we expect the 12C-16C to be faster than the 8C? What I mean is will the 8C have one CCX this time and not suffer latency issues? I don't really need more than 8C anymore and figure why spend money on a 12C or greater if the 8C is just as fast.
"Smartly" veiled personal attacks. Typical. When I've proved that there are outliers in the performance numbers.To me it looks like your the perfect example of a outlier in this thread.
Using your displaced aggression from Intels recent failures, towards those whom are interested in discussing AMD's new offerings, isn't offering anything constructive to the conversation.
Can you really use the concept of outlier within a range of tests or only within an individual test, eg, eliminate aberrant highs and lows due to testing variability? You might be stretching the concept to advance a narrative."Smartly" veiled personal attacks. Typical. When I've proved that there are outliers in the performance numbers.
How big is the typical variability in CPU benchmarking compared to the magnitude of the difference between the outliers and median values? Of course, no testing is perfect in the real world, but I'm assuming that the data obtained is free from any variability.Can you really use the concept of outlier within a range of tests or only within an individual test, eg, eliminate aberrant highs and lows due to testing variability? You might be stretching the concept to advance a narrative.
The 3600 is $200, and the 3600XT is $240 on newegg right now. Both are just fine for what you want.I wonder if there will be any lower end skus, i.e. 5300x. I want to build my daughter a budget gaming rig, but was hoping for $200ish cpu or even less.
Without having run these tests myself, I really can't comment on variations between runs. My post though was on using outliers to eliminate certain tests outright, versus outliers within the test itself. Sorry, I can't see your use of the concept as valid.How big is the typical variability in CPU benchmarking compared to the magnitude of the difference between the outliers and median values? Of course, no testing is perfect in the real world, but I'm assuming that the data obtained is free from any variability.
I should have specified, I won't be ready to build before the new CPUS are out.The 3600 is $200, and the 3600XT is $240 on newegg right now. Both are just fine for what you want.
Well, not sure what all will show up then, or availability. So ask when they are out ???I should have specified, I won't be ready to build before the new CPUS are out.
Outliers within the test itself? You mean variability of the benchmark run? On that ground both LoL and CS:GO should be dismissed outright because they do not come with canned benchmarks(ulletical's CS:GO benchmark doesn't count as it is not representative of actual gameplay).Without having run these tests myself, I really can't comment on variations between runs. My post though was on using outliers to eliminate certain tests outright, versus outliers within the test itself. Sorry, I can't see your use of the concept as valid.
Those two values are outliers, but that doesn't mean that they are not valuable information and should just be dropped. It would be the same if you had a benchmark suite that showed a 5950X or Threadripper to be faster in almost all tests vs a similar Intel chip, except two benchmarks that use AVX512 and Intel wins handily.Without having run these tests myself, I really can't comment on variations between runs. My post though was on using outliers to eliminate certain tests outright, versus outliers within the test itself. Sorry, I can't see your use of the concept as valid.
"Smartly" veiled personal attacks. Typical. When I've proved that there are outliers in the performance numbers.
Would be the first time AMD launches all Ryzen SKUs all at once then.
No, Amd 1800X was $499, now the 5800X is only $449 so it's actually less expensive.What I don't love is that AMD has increased the price of entry for an 8-core by quite a bit.
The lack of a 65W TDP 8c/16t SKU (so 5700X) is an odd omission, especially with all the talk about improved efficiency. But AMD never launched all possible SKUs at once, so personally I fully expect it to announce and launch missing SKUs at a later date.What I don't love is that AMD has increased the price of entry for an 8-core by quite a bit. If we take AMD's original MSRP for the Ryzen 3700X ($329) and compare it to the new entry level 8-core - the 5800X at $449, that's a whopping $120 increase. People are comparing these new prices to last gen's 3800X and 3600X which fewer people purchased versus the 3700x and 3600, respectively. I was prepared to buy a Zen 3 8-core for around $350 but now at an additional hundred dollars I think I am going to hold on to my Ryzen 2700 for another 6 months to wait for discounts or perhaps a Ryzen 5700X (or 5800 non-x) will be announced
Now I'm not saying AMD shouldn't, or can't justify it. But let's call it for what it is. It's not a $50 price increase. With Zen 2 you could buy a 3600 for $199 at launch. Now with Zen 3 a 6-core is $299 at launch. That's a 50% increase in cost! And the 8-core comes out to a 30% increase in cost.
Launch prices for Zen 2: https://www.anandtech.com/show/1561...eration-desktop-ryzen-promotions-up-to-50-off
And 5600X and 5600 don't, while 5800X, 5900X and 5950X will, whereas 3800X, 3900X and 3950X didn't. ¯\_(ツ)_/¯3600 and 3600x launched the same date
No, Amd 1800X was $499, now the 5800X is only $449 so it's actually less expensive.
The 1700 was $400, I bought one.....well, it might have been a 1700x that I got,.The 1700 was $329.