Tesla lowers Model 3 range estimates in Europe due to extra power consumption of AMD Ryzen processor

[_Redshift_]

Although it does seem suspicious that a chip would drain 7 miles worth of energy. But I'm not a car expert.

I am in the chip industry, specifically power consumption.

Chip power consumption is the number one issue nowadays, since everyone’s chip basically works. Power and performance are huge issues.

I would not be surprised if the the range reduction happens due to power hungry semiconductors.

My Model S used to lose 12-15 miles overnight when I bought it in 2013 just sitting still. Some of that was due to BMS, but even in fair weather of spring, it still managed to lose a lot of miles. Tesla made changes to the software to not constantly communicate stuff over the wifi and the loss dropped to about 7 miles.

It still manages to lose 7 miles overnight.

 

[CarlThompson]

It doesn't matter what the maximum rating is since these chips all throttle down accordingly. So it's likely that the more modern Ryzen actually uses less power on average even if the power bursts are much higher during app launch or gameplay.

Strange that Tesla is blaming what would have to be a 250W+ increase in energy consumption while driving on a CPU that has a max sustained TDP of 45W or 54W. There's more to the story than what Tesla is saying.

 

[RipVanFinkle]

I feel like the switch to the Ryzen is due to supply shortages and not so much “increased performance”. While on the surface, it looks like a better alternative to the Atom. Dig a little deeper and you see that the chip wasn’t the best move due to increased power consumption on an EV. It’s not like a PC that’s plugged in and consumption isn’t much of a concern but heat is.

This is just like the missing data USB ports, passenger lumbar, etc. they’re just trying to spin it.

 

[CarlThompson]

Dig a little deeper and you see that the chip wasn’t the best move due to increased power consumption on an EV. It’s not like a PC that’s plugged in and consumption isn’t much of a concern but heat is.

When traveling 60 MPH at a good efficiency of 250 Wh / mile the car is using 15,000 watts. Switching to a chip that uses at most ~30 watts more that the previous chip when fully loaded is a minuscule difference compared to what the car uses to move from Point A to Point B. There is more to this story than what Tesla is saying; the decrease in range can't be attributed to the new CPU only.

 

[jjrandorin]

I said this earlier in this thread, but this specific point seems to keep being ignored:

Unless they also lower the range estimates for cars with this chip in the US, this would appear to be more about the European WLTP tests than anything else. We know already that these cars will not get EPA range, and WLTP range is farther than EPA (thus even more unattainable).

Shrug.

WLTP range was reduced. No where else that we are aware of, has tesla released a lower range estimate for the vehicle. I personally dont know what the actual differences are in WLTP vs EPA as far as the actual tests go, I just know that WLTP range is even more optimistic than EPA, which is already more optimistic than just about anyone would ever achieve.

Unless range is reduced everywhere in the world by tesla, this really is "nothing", in my opinon, since no one was ever going to get WLTP range before anyway.

 

[Gauss Guzzler]

Exactly. This is just a case of an ignorant journalist who clearly doesn't understand cars or electricity or anything misinterpreting an alleged email he allegedly saw. Darryn John should be ashamed of himself for publishing such drivel without even giving the ridiculous theory a single thought. I mean how hard is it to compare Watts to Watts and just do a simple sanity check before writing an entire article on the topic?

EPA/WLTP/etc change all the time. They test one configuration and are allowed to use it to estimate others within some limits, then they eventually have to re-test once the configuration has changed too much. They probably just had enough cell chemistry or software changes to trigger a WLTP retest but not EPA (yet).

 

[diamond.g]

It doesn't matter what the maximum rating is since these chips all throttle down accordingly. So it's likely that the more modern Ryzen actually uses less power on average even if the power bursts are much higher during app launch or gameplay.

I didn't think Zen+ was that much better on power, especially compared to Atom.

 

[stopcrazypp]

Exactly. This is just a case of an ignorant journalist who clearly doesn't understand cars or electricity or anything misinterpreting an alleged email he allegedly saw. Darryn John should be ashamed of himself for publishing such drivel without even giving the ridiculous theory a single thought. I mean how hard is it to compare Watts to Watts and just do a simple sanity check before writing an entire article on the topic?

EPA/WLTP/etc change all the time. They test one configuration and are allowed to use it to estimate others within some limits, then they eventually have to re-test once the configuration has changed too much. They probably just had enough cell chemistry or software changes to trigger a WLTP retest but not EPA (yet).

Does anyone have the text of the actual email or message to consumers? This may turn out to be another telephone game, where perhaps the wording is that Ryzen is part of it, not necessarily that it is the only change that is responsible for the difference in range.

 

[3sr+buyer]

Note that in the US, EPA economy (not range) for the 2022 RWD (138/126) is lower than for the 2021 SR+ (150/133 L-NCA or 149/132 LFP) for some reason. Doubtful that this has anything to do with Ryzen chips.

 

[Brettka7]

Note that in the US, EPA economy (not range) for the 2022 RWD (138/126) is lower than for the 2021 SR+ (150/133 L-NCA or 149/132 LFP) for some reason. Doubtful that this has anything to do with Ryzen chips.

This is due to the higher weight.

 

[JBT66]

This doesn’t make any sense and hopefully just Tesla marketing these cars with closer to real world range.

 

[leaftoy]

I am in the chip industry, specifically power consumption.

Chip power consumption is the number one issue nowadays, since everyone’s chip basically works. Power and performance are huge issues.

I would not be surprised if the the range reduction happens due to power hungry semiconductors.

My Model S used to lose 12-15 miles overnight when I bought it in 2013 just sitting still. Some of that was due to BMS, but even in fair weather of spring, it still managed to lose a lot of miles. Tesla made changes to the software to not constantly communicate stuff over the wifi and the loss dropped to about 7 miles.

It still manages to lose 7 miles overnight.

Can you show your work, please, how Ryzen switch results in the stated range loss for 3?

 

[_Redshift_]

Give me the design, and a few computers, and I can produce detailed analyses in a week.

 

[freeAgent]

I feel like the switch to the Ryzen is due to supply shortages and not so much “increased performance”. While on the surface, it looks like a better alternative to the Atom. Dig a little deeper and you see that the chip wasn’t the best move due to increased power consumption on an EV. It’s not like a PC that’s plugged in and consumption isn’t much of a concern but heat is.

This is just like the missing data USB ports, passenger lumbar, etc. they’re just trying to spin it.

I highly doubt it. AMD's chips are made by TSMC while Intel make their own Atom chips. Intel actually has plenty of capacity for their Atom chips (which tend to be on older process nodes that are in low demand) while AMD's chips are on high-demand TSMC nodes. This was almost certainly not a supply chain constraint-related switch.

Anyway, I've heard other people speculating that the real reason for the range change may be in the battery (i.e., Tesla put new batteries in the LR pack which have a few less kW capacity).

 

[john5520]

Even if the Ryzen is the culprit, I'll gladly take the range hit for the faster chip. They should have swapped out that atom a long time ago.

 

[stopcrazypp]

Even if the Ryzen is the culprit, I'll gladly take the range hit for the faster chip. They should have swapped out that atom a long time ago.

Yeah, even if the range hit was all from the Ryzen (which it likely isn't), I personally would gladly trade for that chip also instead of my Atom. Heck, I'd even trade my front usb ports (which I don't use other than for testing) for it.

 

[Gasaraki]

It's true. Try it yourself if you don't believe them.

Just park the car and play the Sonic or Beach Buggy games 24/7 without sleeping (bring snacks and a soda cup). You'll see that within a few weeks the car will have lost about 7 miles of range.

Or just do the math if you're not that into gaming. Assuming the Ryzen uses 5W more than the Atom, that'll increase the car's average power consumption from 15,000W all the way up to 15,005W which rounds to about 0.00% more power (yes, this is the actual calculated value). It's hard to picture just how big an increase 0.00% is, but let's just say if you drove a million miles with the Ryzen chip it would use about 0.00% more power than an Atom.
And 0.00% of 1,000,000 = 7.

The Atom processor in the car only uses 9.5 W. If the Ryzen is 45W, then it uses almost 5 times more power.

 

[terranx]

The Atom processor in the car only uses 9.5 W. If the Ryzen is 45W, then it uses almost 5 times more power.

Sure, but it's like pouring a quart vs a gallon of water into the ocean. Technically 4x more, but a meaningless difference in the overall system.

 

[Gauss Guzzler]

Yeah, maybe the Ryzen could conceivably use 35.5W more than the Atom if you're watching Netflix while posting to Tiktok and playing Beach Buggy while streaming Tidal and navigating with satellite photo maps. All while the processor is crunching an infinite loop background process.

But even in that ridiculous case it still works out to 35.5W/15,000W = 0%

 

[stopcrazypp]

Yeah, maybe the Ryzen could conceivably use 35.5W more than the Atom if you're watching Netflix while posting to Tiktok and playing Beach Buggy while streaming Tidal and navigating with satellite photo maps. All while the processor is crunching an infinite loop background process.

But even in that ridiculous case it still works out to 35.5W/15,000W = 0%

I don't think that's quite the correct math. If you look at EPA test results, page 20, the 2022 Tesla LR AWD:

UDDS cycle: 505 miles, average speed 19.59 mph, so it takes 25.8 hours. 82.067 kWh DC was used, meaning around 3200W. 35.5W/3200W = 0.0111 or 5.6 miles of range.

On Highway cycle: 475 miles, average speed 48.3 mph, 9.8 hours. 82.067 kWh DC, ~8400W. 35.5/8400W = 0.0042 or 2 miles of range.

On FTP cycle: 306 miles, average speed 21.2 mph, 14.4 hours. 76.577 kWh DC, ~5300W. 35.5/5300W = 0.0067 or 2 miles of range.
https://dis.epa.gov/otaqpub/display_file.jsp?docid=54391&flag=1

Of course, the Ryzen processor wouldn't be running at full power, so I still don't feel it'll be that much, but if it was running full bore (especially if you add in a portion of 65-90W for the discrete Navi 23 GPU) and the test cycle had very low average speeds (similar to UDDS), you can get multiple miles of range impact (not just rounding to zero).