Mmmfishtacos said:
Machiavellian said:
Mmmfishtacos said:
Adinnieken said:
Ashadian said:
"The CPU contained within the PS4 is running stock speeds, 1.6ghz compared to Microsofts 1.75 after an upclock - 150mhz does not make up for the different in number of CUs, and (aptly ignored by Albert) the difference in ROP units - hence why even with the 150mhz upclock, the Xbox One is still way behind in terms of raw power - To see just how little difference 150mhz makes, go into your bios and bump your clock speed up to be 150mhz faster, you will see framerates climb barely 2fps at best - most modern CPUs clock to at least 400mhz higher than stock as an average, with some people bumping that to over 900mhz.
But seriously, before you try and tell me im wrong, he is comparing 1.6 to 1.75 and claiming its 10% faster, asside from being factually wrong (10% would be 1.76ghz), it also compares the cpus on the most basic of levels, which is a stupid thing to do, even more so given that theyre APUS - compare a 2ghz celeron to a 2ghz pentium to see why.
Last but not least adding up the ddr3 and esram speeds to get a higher number is universally seen as a retarded pr stunt by those with half a brain because theoretical peaks for different ram types DO NOT COMBINE - just like if you put two 3ghz xeons in a server, it does NOT mean that the server is now running at 6ghz, its running at 3ghz with additional cores - adding ddr3 and the tiny space allotted by esram does not work, even for "on paper" results - you cannot fill the 8gb of ddr3 with the esram fast enough without the esram being bottlednecked, esram could not be used for a large majority of game resources, where as GDDR5 can be used for just about anything at the cost of slightly higher latency."
|
Actually, he doesn't ignore it. He addressed it. Each CU gets a 6% speed increase, not simply the entire GPU with the frequency increase. So while there are fewer CUs those CUs operate faster than the PS4's. When 12 CUs are operating at 6% greater speed than the PS4, that equates into a 72% speed improvement.
Early in that thread (not that post) he states near 10%. He has also stated before greater than 9%. If we want to get specific it's 9.375% No, he also talks on a deeper level, about how the the Xbox One's CPU also has a 30GB/s transfer rate between the CPU and memory, A 50% speed improvement over the PS4.
You missed the part where he also specified that the eSRAM's speed was 204GB/s. That still happens to be 16% faster than the GDDR5 memory in the PS4. Don't assume my argument is that 32MB is the same as 8GB, but I wonder what memory speed is necessary in order for a modern GPU to adequately process a 1920x1080 image without bottlenecking or taking a performance hit. The eSRAM is used both as a buffer for the DDR memory, as well as a cache to temporarily hold data that's needed quickly.
|
I think you missed the part where he got called out on all the being fud. And he got shreading for this bit of miss info too.
|
Let me ask you something. How do you know he got shredded. Without having the knowledge if the people responding actually know what they are talking about you can easily make that mistake. After reading a few comments its clear a lot of them do not know what they are talking about and most cannot even make a simple google search to be sure what they are saying is correct. This is the problem with the net is that people regurgitate information from other people who do not know what they are talking about and you get number to junk.
|
Because a lot of them do know what they are talking about. And even a quick google search will tell you that the ps4 transfers 20 GB/s down the onion buss. So that alone slames his 50 percent gain.
|
Thats exactly what I am talking about. A lot of them do not know what they are talking about. This is a pretty big misconception of the forum. Most times I see comments that are no better than your common fan site.
Yes, 20 down for PS4 compared to 30 for the X1. You do the math and tell me what you get!!
